Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mental health service utilization by Chinese immigrants and Canadian-born Chinese in British Columbia Chen, Alice W. 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-199527.pdf [ 11.99MB ]
Metadata
JSON: 831-1.0092871.json
JSON-LD: 831-1.0092871-ld.json
RDF/XML (Pretty): 831-1.0092871-rdf.xml
RDF/JSON: 831-1.0092871-rdf.json
Turtle: 831-1.0092871-turtle.txt
N-Triples: 831-1.0092871-rdf-ntriples.txt
Original Record: 831-1.0092871-source.json
Full Text
831-1.0092871-fulltext.txt
Citation
831-1.0092871.ris

Full Text

MENTAL H E A L T H SERVICE UTILIZATION BY^HTNESE IMMIGRANTS AND CANADIAN-BORN CHINESE IN BRITISH COLUMBIA By Alice W. Chen B . S c , University of Illinois at Urbana-Champaign, 1980 M.C. , Arizona State University, 1983 M . S c , Arizona State University, 1985 A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF DOCTOR OF PHILOSOPHY in T H E F A C U L T Y OF G R A D U A T E STUDIES (Health Care and Epidemiology) > T H E UNIVERSITY OF BRITISH C O L U M B I A August 2006 © Alice W. Chen, 2006 ABSTRACT A population-based research of rates and patterns of mental health service utilization of Chinese in British Columbia was conducted by secondary analyses of three linked administrative data systems and a survey database. The three data systems comprised the federal immigration database, provincial healthcare databases and the provincial physicians register. The linkage resulted in a study population of more than 150,000 Chinese immigrants and each immigrant was matched by sex and age to a comparison subject randomly selected from the province's health plan registry. The survey database was the Canadian Community Health Survey Cycle 1.1 consisting of 18,000 respondents in British Columbia including 1084 ethnic Chinese. Results from the administrative databases showed that Chinese immigrants had much lower rates of utilization of all types of mental health services. Factors associated with mental health visits to general practitioners and psychiatrists by these Chinese immigrants included years since landing, rate of non-mental health visits, age, place of origin, education, marital status, English language skill and health service delivery area. Among a smaller cohort of 786 Chinese immigrants and 3962 comparison subjects diagnosed with severe and persistent mental illness, Chinese immigrants also had lower rates of utilization of all types of mental health services. The severely il l Chinese immigrants were also less likely than comparison subjects to receive comorbid diagnoses of less serious mental disorders. The majority of the immigrants' mental health care was provided by Chinese-speaking physicians. As a result of the practice pattern of these physicians, higher proportion of mental health care received from them was associated with lower odds of receiving diagnoses of some disorders. Results from the survey database showed that both • immigrant and Canadian-born Chinese were less likely than the non-Chinese to have contacted mental health professionals, even after adjusting for their lower probability of suffering from major depression. This research confirms the disparity in mental health service utilization concerning Chinese in British Columbia and provides evidence that both lower psychiatric morbidity and cultural barriers contribute to this disparity. Findings support Andersen's behavioural model of health care utilization and highlight the role of physician practice patterns in disparity. ii TABLE OF CONTENTS Abstract ii Table of Contents i i i List of Tables vi List of Figures xi Acknowledgements xii Dedication xiii Co-Authorship Statement xiv Chapter 1: Introduction 1 Overview of the Dissertation 1 Background 2 Rate of Mental Health Service Utilization of Immigrants and Minorities 3 Determinants of Mental Health Service Utilization: Evidence from Population Surveys 8 Determinants of Utilization Specific to Immigrants and Ethnic Minorities 16 Conceptual Framework: Behavioural Model of Health Care Utilization 19 Reasons for Current Study 25 Bibliography 30 Chapter 2: Disparities in Utilization of Mental Health Services by Chinese Immigrants in British Columbia 36 Methods 38 Results 42 Discussion 46 Conclusion 49 Bibliography 59 Chapter 3: Determinants of Mental Health Consultations by Recent Chinese Immigrants in British Columbia . 61 Methods 62 Results 67 Discussion 71 Conclusion 79 Bibliography 91 Chapter 4: Mental Health Service Utilization by Chinese Immigrants with Severe and Persistent Mental Illness in British Columbia 94 Methods 97 Results 104 Discussion 109 Conclusion 114 Bibliography 125 Chapter 5: Mental Health Consultation by Immigrant and Canadian-born Chinese in British Columbia 127 Methods 129 Results 131 Discussion 134 Conclusion ; 138 Bibliography 146 Chapter 6: Conclusion 149 Summary of Research Findings 149 Implications for Health Services Research and Policies 152 Strengths and Weaknesses of Research Design 156 Future Directions 158 Bibliography : 161 Appendix A: Data Sources in Linked Databases 165 Appendix B: Linkage Rates of Chinese Immigrants between Immigration Database and Health Database 171 Appendix C: Data Linkage Issues 174 Appendix D: Data Cleaning and Preparation in Linked Databases 179 Appendix E: Subject Selection Criteria 189 iv Appendix F: Study Variables in Linked Databases 195 Appendix G: Psychiatric Diagnoses, Fee Items and Medications 203 Appendix H : Identification of Chinese Language Ability of Physicians 207 Appendix I: Comparing Model Estimates and Bootstrap Standard Errors 210 Appendix J: Effects of Determinants on Rates of Mental Health Visits to Physicians 219 v LIST OF TABLES Table 1.1 Summary of Findings on Determinants of Mental Health Service Utilization 16 Table 1.2 Components in Andersen's Behavioural Model of Health Care Utilization 22 Table 2.1 Description of Immigrants and Comparison Subjects 50 Table 2.2a Rate and Rate Ratio of Mental Health Visits to General Practitioners and Psychiatrists by Chinese Immigrants and Comparison Subjects in 1992-2001 51 Table 2.2b Rate and Rate Ratio of Mental Health Visits to General Practitioners by Chinese Immigrants and Comparison Subjects in 1992-2001 52 Table 2.2c Rate and Rate Ratio of Visits to Psychiatrists by Chinese Immigrants and Comparison Subjects in 1992-2001 53 Table 2.3 Rate and Rate Ratio of Mental Health Hospitalizations by Chinese Immigrants and Comparison Subjects in 1992-2001 54 Table 2.4 Rate and Rate Ratio of Time in Outpatient Mental Health Care by Chinese Immigrants and Comparison Subjects in 1992-2001 55 Table 2.5a Diagnostic Categories of Mental Health Visits to General Practitioners and Psychiatrists by Chinese Immigrants and Comparison Subjects in 1992-2001 56 Table 2.5b Diagnostic Categories of Mental Health Visits to General Practitioners by Chinese Immigrants and Comparison Subjects in 1992-2001 57 Table 2.5c Diagnostic Categories of Visits to Psychiatrists by Chinese Immigrants and Comparison Subjects in 1992-2001 58 Table 3.1a Characteristics of Study Immigrants Aged Under 25 at Landing and Comparison Subjects 81 Table 3.1b Characteristics of Study Immigrants Aged 25 or Over at Landing and Comparison Subjects 82 Table 3.2a Determinants of Chinese Immigrants' Mental Health Visits to General Practitioners - Immigrants Under 25 Years Old at Landing 83 Table 3.2b Determinants of Chinese Immigrants' Mental Health Visits to General Practitioners - Immigrants Aged 25 Years or Over at Landing 84 Table 3.3a Determinants of Chinese Immigrants' Visits to Psychiatrists -Immigrants Under 25 Years Old at Landing 85 Table 3.3b Determinants of Chinese Immigrants' Visits to Psychiatrists -Immigrants Aged 25 Years or Over at Landing 86 Table 3.4 Determinants of Comparison Subjects' Mental Health Visits to General Practitioners (Multivariate Models) 87 Table 3.5 Determinants of Comparison Subjects' Visits to Psychiatrists (Multivariate Models) 88 Table 3.6 Probability of Mental Health Visits to Chinese-speaking Physicians by Chinese Immigrants 89 Table 3.7 Number of Directory Active Physicians per 100,000 Population in 1998 90 Table 4. l a Characteristics of Immigrant and Comparison Subjects Diagnosed with Severe and Persistent Mental Illness and Aged 15 or Over in 1992-2001 115 Table 4.1b Characteristics of Immigrant and Comparison Subjects Diagnosed with Severe and Persistent Mental Illness and on Pharmacare Plans A B C F G in 1996-2001 116 Table 4.2 Rates of Mental Health Service Utilization in 1992-2001 by Immigrant and Comparison Subjects Diagnosed with Severe and Persistent Mental Illness 117 Table 4.3a Diagnostic Categories of Mental Health Visits to General Practitioners & Psychiatrists of Immigrant & Comparison Subjects with Severe and Persistent Mental Illness , 118 Table 4.3b Most Responsible Diagnostic Categories of Mental Health Hospitalizations of Immigrant & Comparison Subjects with Severe and Persistent Mental Illness 119 Table 4.4 Percentage and Odds Ratio of Receiving Diagnosis of Selected Mental Health Conditions 120 Table 4.5 Rate of Mental Health Services Received from Chinese-speaking Physicians 121 Table 4.6 Diagnostic Categories of Mental Health Visits of Immigrant and Comparison Subjects to General Practitioners & Psychiatrists by Language Skill of Physician 122 Table 4.7 Odds Ratio of Receiving Diagnosis of Selected Mental Health Conditions by Proportion of Mental Health Visits to Chinese-speaking Physicians 123 Table 5.1 Characteristics of B C Respondents to CCHS Cycle 1.1 139 Table 5.2a Percentage of B C Residents Who Contacted Mental Health Professionals in Past 12 Months 140 Table 5.2b Mean Number of Mental Health Consultations of B C Residents -Of Those Who Contacted Mental Health Professionals 140 Table 5.3a Mean Depression Score of B C Residents 141 Table 5.3b Percentage of BC Residents Who Had Moderate to High Probability of Depression 141 Table 5.3c Percentage of BC Residents Who Contacted Mental Health Professionals in Past 12 Months - Of Those Who Had Moderate to High Probability Of Depression 141 Table 5.4 Mean Number of Medical Consultations of B C Residents in Past 12 Months.... 142 Table 5.5a Percentage of B C Residents Who Had Unmet Health Care Needs in Past 12 Months 143 Table 5.5b Percentage of BC Residents Who Had Unmet Mental Health Care Needs in Past 12 Months - Of Those with Unmet Health Care Needs '. 143 Table 5.6a Odds Ratio of Contacting Mental Health Professionals in Past 12 Months 144 Table 5.6b Odds Ratio of Contacting Mental Health Professionals in Past 12 Months - Of Those With Moderate to High Probability of Depression • 144 Table 5.7 Odds Ratio of Factors Associated with Mental Health Consultation Among Chinese Respondents 145 Table B. 1 Linkage Rates by Age Group and Sex 171 Table B.2 Linkage Rates by Landing Year and Sex 171 Table B.3 Linkage Rates by Marital Status and Sex 172 Table B.4 Linkage Rates by Education and Sex 172 Table B.5 Linkage Rates by Official Language Ability and Sex 172 Table B.6 Linkage Rates by Country of Birth and Sex 172 Table B.7 Linkage Rates by Country of Last Permanent Residence and Sex 173 Table D. 1 Comparing Sex in Immigration Database and Health Database 180 Table D.2 Comparing Year of Birth in Immigration Database and Health Database 181 Table D.3 Time Difference Between MSP Registration and Landing Date 182 Table D.4 M S P Visits and Health Plan Registration of Immigrant and Comparison Subjects in 1986-2001 185 Table D.5 Hospital Separations and Health Plan Registration of Immigrant and Comparison Subjects in 1986-2001 186 Table D.6 Psychiatric Prescriptions and Health Plan Registration of Immigrant and Comparison Subjects in 1986-2001 188 Table E . l Percentage of Chinese in Selected Asian Populations 190 Table H. 1 Number of Chinese-speaking Physicians Identified from CPSBC Register 208 Table 1.1 Negative Binomial Model Estimates of Rate Ratios of Mental Health Visits to General Practice and Results of Bootstrapping (Female Immigrants Under 25 at Landing) 211 ix Table 1.2 Negative Binomial Model Estimates of Rate Ratios of Visits to Psychiatry and Results of Bootstrapping (Female Immigrants Under 25 at Landing) 212 Table 1.3 Negative Binomial Model Estimates of Rate Ratios of Mental Health Visits to General Practice and Results of Bootstrapping (Male Immigrants Under 25 at Landing) . ...213 Table 1.4 Negative Binomial Model Estimates of Rate Ratios of Visits to Psychiatry and Results of Bootstrapping (Male Immigrants Under 25 at Landing) 214 Table 1.5 Negative Binomial Model Estimates of Rate Ratios of Mental Health Visits to General Practice and Results of Bootstrapping (Female Immigrants Aged 25 or Over at Landing) 215 Table 1.6 Negative Binomial Model Estimates of Rate Ratios of Visits to Psychiatry and Results of Bootstrapping (Female Immigrants Aged 25 or Over at Landing) 216 Table 1.7 Negative Binomial Model Estimates of Rate Ratios of Mental Health Visits to General Practice and Results of Bootstrapping (Male Immigrants Aged 25 or Over at Landing) 217 Table 1.8 Negative Binomial Model Estimates of Rate Ratios of Visits to Psychiatry and Results of Bootstrapping (Male Immigrants Aged 25 or Over at Landing) 218 Table J. 1 Effects of Determinants on Rate of Mental Health Visits to General Practice (Female Immigrants under 25 at Landing) 220 Table J.2 Effects of Determinants on Rate of Mental Health Visits to General Practice (Male Immigrants under 25 at Landing)... 222 Table J.3 Effects of Determinants on Rate of Mental Health Visits to General Practice (Female Immigrants Aged 25 or Over at Landing) 224 Table J.4 Effects of Determinants on Rate of Mental Health Visits to General Practice (Male Immigrants Aged 25 or Over at Landing) 227 Table J. 5 Effects of Determinants on Rate of Visits to Psychiatry (Female Immigrants under 25 at Landing) 230 Table J.6 Effects of Determinants on Rate of Visits to Psychiatry (Female Immigrants Aged 25 or Over at Landing) 231 x LIST OF FIGURES Figure 1.1 Andersen's Behavioural Model of Health Care Utilization 21 F i g u r e d Percentage of Chinese Immigrants, Immigrants and Comparison Subjects in Metropolitan Vancouver 177 xi ACKNOWLEDGMENTS The author of this thesis was supported by 1) a doctoral training award from the Canadian Institutes of Health Research 2) a Western Regional Training Centre (WRTC) studentship funded by the Canadian Health Services Research Foundation, Alberta Heritage Foundation for Medical Research and the Canadian Institutes of Health Research 3) a Research in Addictions and Mental Health Policy and Services (RAMHPS) award, funded by Canadian Institutes of Health Research Data for the linked administrative databases used in this thesis were provided by Citizenship and Immigration Canada, British Columbia's Ministry of Health and the College of Physicians and Surgeons of British Columbia. Linkage of the immigration and health databases was funded by the Canadian Population Health Initiative as part of a national immigrant health study titled "Immigrant Health and Health Care Utilization in Canada". Investigators for the national study were Marie Desmeules (Public Health Agency of Canada), Arminee Kazanjian (University of British Columbia), Doug Manuel (Institute for Clinical Evaluative Sciences), Yang Mao (Public Health Agency of Canada) and Bilkis Vissandjee (University of Montreal). The Centre for Health Services and Policy Research at the University of British Columbia performed the linkage of the databases and extracted the data required for this thesis. Data for the Canadian Community Health Survey Cycle 1.1 were provided by the British Columbia Inter-University Research Data Centre of Statistics Canada. The author would like to thank the thesis supervisor - Dr. Arminee Kazanjian -and the thesis committee - Dr. Elliot Goldner, Dr. Robert Reid and Dr. Hubert Wong -for their guidance and support in the completion of this thesis. DEDICATION This dissertation is dedicated to the memory of Dr. Kwok Chu (KC) L i 1930 - 2006 who worked tirelessly for 30 years to provide psychiatric care to British Columbians, to pioneer cross-cultural psychiatry in this province, to advocate for better mental health services to the Canadian Chinese, to educate the Chinese community about mental health issues, and to inspire a generation of mental health professionals. xiii CO-AUTHORSHIP STATEMENT This statement describes the contribution of each of the following individuals to the manuscripts in this dissertation, in accordance with the requirements of the Faculty of Graduate Studies at the University of British Columbia. Alice Chen formulated the research questions for this thesis. She also conceived the research design which took advantage of existing linked immigration and health databases, added the physician register to the linkage, and supplemented them with data from the survey database. She obtained approval to access all the data, collected the data, developed the methodology, prepared the data and performed all the analyses. She interpreted the results and wrote and revised all the manuscripts. Dr. Arminee Kazanjian was the thesis supervisor. She was also a principal investigator in the national immigrant health research study which linked the immigration and provincial health databases and facilitated access to the immigration database. She provided ongoing guidance in conducting the thesis research and interpreting the findings. She reviewed all the manuscripts and provided feedback. Dr. Hubert Wong was a member of the thesis committee. He advised on specific methodological issues and provided extensive guidance and support on all the statistical procedures and interpretations. He reviewed all the manuscripts and provided feedback. Dr. Elliot Goldner and Dr. Robert Reid were members of the thesis committee. They both provided advice on the research questions and some of the methodological issues pertaining to mental health, including the criteria for defining severe and persistent mental illness. They also reviewed the manuscripts. xiv CHAPTER 1 INTRODUCTION OVERVIEW OF T H E DISSERTATION The main subject addressed in this doctoral thesis project is the mental health service utilization of ethnic Chinese residents in British Columbia (BC), including immigrants and the Canadian-born. Issues examined include the rates of utilization, individual and contextual factors associated with utilization, differences between Chinese immigrants and the comparison group, differences between immigrant and Canadian-born Chinese and utilization among the general population as well as among those diagnosed with severe mental illness. Methods are based on secondary analyses of existing administrative and survey databases. The dissertation is structured according to the manuscript-based format approved by the Faculty of Graduate Studies at the University of British Columbia. This introductory chapter begins with the background information related to the thesis topic and a review of relevant literature and ends with the reasons and objectives of this thesis project. Chapters 2 to 5 are individual manuscripts produced from the thesis research. Chapter 2 compares the rates of utilization of four types of mental health services between Chinese immigrant and comparison group subjects in B C . Chapter 3 describes the individual and contextual variables associated with mental health visits to physicians among the Chinese immigrants. Chapter 4 examines the rates of utilization among Chinese immigrants and comparison group members diagnosed with severe and persistent mental illness and also the relationship between physicians' Chinese language skill and utilization and diagnosis. Chapter 5 reports the mental service utilization of Canadian-born Chinese and immigrant Chinese. The last chapter in this dissertation, Chapter 6, is the concluding chapter which relates Chapters 2 to 5 to each other and discusses the implications of this thesis to the field. The appendices provide additional details of methodology that are not covered in the manuscripts. 1 Chapter 1 BACKGROUND Trends in international migration have evolved in the last century. Owing to a combination of factors involving global socio-political climate and ideological and legislative changes in the West, the majority of recent immigrants to western nations no longer share common ethnic origins with their counterparts in the host countries. Consequently, many immigrant-receiving western nations witness noticeable shifts in the ethnic characteristics of their populations. Canada was the first major immigrant-receiving country to abolish racially discriminatory immigration policies. Since 1967, immigrants to Canada were selected on the basis of their education and skill and were able to sponsor relatives regardless of national origin. As a result, the ethnic makeup of the immigrant population rapidly transformed. In 1966, 87% of immigrants arriving in Canada were of European origin; only four years later, 50% came from Asia and the Caribbean (1). By 2001, only 20% of the new immigrants originated from Europe or the United States (US). 1 The dramatic change in the ethnic background of new Canadians inevitably brought about gradual changes in the ethnocultural composition of Canada's overall population. In 2001, 18% of the population in Canada were immigrants and 13% were visible minorities; only three-quarters of Canadians reported Canadian, English or French as their ethnic origin.2 Such ethnocultural diversity challenges service providers, especially health service providers. Clinical staff are frustrated by their inability to reach minority clients while ethnocultural communities criticize the health care system as being culturally insensitive. In mental health care, there have been many concerns expressed about differential utilization of mental health service by immigrants and ethnic minorities. The first section of this chapter reports the research on the rates of mental health service utilization by immigrants and minorities in western societies. The second and third sections summarize the literature on the determinants of utilization while the fourth section reviews a behavioural model of health care utilization that forms the conceptual 1 Citizenship and Immigration Canada Facts and Figures 2001 http://www.cic.gc.ca/english/pub/facts2001/limm-05.html 2 Statistics Canada Census 2001 Cat. No. 97F0024XIE2001005 and 97F0024XIE2001006 http://wwwl2.statcan.ca/engJish/census01/Products/standard/memes/DataProducts.c 2 Chapter 1 framework for this thesis. The chapter closes with a discussion of the research goals, specific aims and questions of this thesis. ' RATE OF MENTAL H E A L T H SERVICE UTILIZATION OF IMMIGRANTS AND MINORITIES Although all studies of mental health service.utilization report some differences by ethnicity or immigration status, the findings are often conflicting, most likely a result of differences in methodology and study populations. The following is a summary of some of those findings from published literature in the English language. International Studies A primary concern of US research on the subject of health care utilization is equity and disparity among racial and ethnic groups. Most research efforts are directed towards African Americans and Hispanics, the two largest visible minority groups in that country. Some of the studies also include Asians. A limitation of results from the US is that, owing to the fragmented nature of the US health service delivery system, few studies are able to provide a comprehensive picture. Nevertheless, current evidence strongly suggests that the level of mental health services received varies by racial and ethnic , group. For African Americans, findings about their rates of utilization of inpatient or outpatient mental health care range from over-representation (2, 3) to being equivalent with Whites (4-6) to under-utilization (7-10). For Hispanics, while there is some evidence that this minority group may be more likely to use some type of mental health services (2), most studies find either little difference or moderate to substantial under-utilization relative to Whites (3, 5-12). In contrast to African Americans and Hispanics, Asian Americans consistently show a lower probability of using mental health services although the amount of use may not differ once they enter the system (2, 3, 11, 13). The under-utilization persists among those presenting mental health symptoms or serious mental illness; furthermore, these Asian Americans with mental health problems are also less likely to report unmet mental health needs (8). Among patients in the public mental health system, there is evidence of ethnic differences in the probability of using various 3 Chapter 1 types of services, although contextual factors, such as region and poverty level, seem to influence the nature of the differences (14, 15). In the United Kingdom, where there is universal public health care, concern over ethnic differences places more emphasis on differential diagnosis and treatment rather than the volume or determinants of utilization. The largest visible minority groups in that country are African-Caribbeans and South Asians. It was reported that South Asians were less likely than Whites to have mental health consultations even when they met criteria for a neurotic disorder and African-Caribbeans were more likely to be diagnosed with psychotic disorders, to be admitted to hospital involuntarily and to stay longer during voluntary admissions. African-Caribbeans diagnosed with schizophrenia were also more likely to have poor outcomes even after seeking treatment. Findings on ethnic differences in the type of mental health services used were not consistent (16-19). As a nation with few indigenous ethnic minorities but a large number of immigrants, Australia directs more of its attention to the utilization of mental health services by immigrants while recognizing language and ethnic differences within the immigrant population. In a study of patients with schizophrenia or bipolar disorder in a psychiatric database, all immigrants were found to be under-represented in inpatient and outpatient services compared to the Australian-born; however, there were differences by place of origin, with southern Europeans having the highest level of utilization and Asians the lowest (20). Canadian Studies Notwithstanding the large number of immigrants residing in Canada and the official acts supporting multiculturalism, there are few published studies on the use of mental health services by ethnic minorities or immigrants in this country. Nevertheless, the limited evidence available tends to support findings from other countries that minorities and immigrants use fewer mental health services but that there is variation by ethnicity. Censuses of patients in Vancouver's community mental health system in 1988 and 1998 showed that Chinese and South Asians were under-represented in its caseloads while Latin Americans were over-represented; the percentage of Vietnamese patients 4 Chapter 1 shifted downward over the years from over-representation to at par with the general population (21, 22). An adolescent psychiatric inpatient unit in Calgary reported that there were fewer Asians admitted than the city's population demographics would indicate and the ones admitted suffered from more severe symptoms (23). A survey of the Caribbean, Vietnamese and Filipino immigrant communities in a Montreal neighbourhood revealed that the immigrants were much less likely to use health care services for psychological distress, although their rate of using general medical care was similar to the Canadian-born (24). This finding was corroborated by a national survey of Chinese aged 55 years and over in seven Canadian cities which reported that, although these older Chinese were more likely than the general population of older adults to visit general practitioners for medical problems, they had lower rates of visiting psychologists and psychiatrists (25). Only the Ontario Health Survey failed to find any significant difference between utilization by immigrants and Canadian-born (26). However, this study did not distinguish between the origin of the immigrants or their duration of residence in Canada, which might explain the difference in finding from other Canadian studies. Mental Health Status of Immigrants and Ethnic Minorities There can be two general explanations for the lower rates of utilization of mental health services by immigrants and ethnic minorities. The first is that they have fewer mental health problems and, therefore, less need for mental health services; the second is that they are not accessing mental health services for other reasons, in spite of their need. Few epidemiological studies on the prevalence of mental health disorders among immigrant and ethnic minority populations can be found and the results are conflicting. Analysis of data from the Epidemiologic Catchment Area survey in Los Angeles, which used the Diagnostic Interview Schedule as the assessment tool, reported rates ranging from comparable to lower among African Americans, Hispanics and Asians relative to Whites for most of the 16 psychiatric diagnoses covering schizophrenic, affective, anxiety and substance use disorders (27). African Americans had higher rates than Whites only in phobia and somatization. Contrary to clinical observation, African Americans did not show higher rates of schizophrenia; nor did Asians and Hispanics 5 Chapter 1 demonstrate higher rates of somatization. The National Surveys on Drug Use and Health in the US, using the short form version of the Composite International Diagnostic Interview and a screener for psychological distress, reported lower odds of mental health symptoms and serious mental illness among African Americans and Mexicans (8). Asian Americans in the study did not differ from Whites in the odds of mental health problems after adjusting for clinical and socio-demographic variables. The Chinese American Psychiatric Epidemiological Study in Los Angeles, using the University of Michigan's version of the Composite International Diagnostic Interview, confirmed that Chinese Americans experienced lower lifetime and 12-month rates of both major depression and dysthymia, compared to the results of the national population as reported in the National Comorbidity Survey which employed the same instrument (28). In contrast, the British National Survey of Ethnic Minorities reported substantially higher rates of both depression and psychosis among Caribbeans than Whites. The revised version of the Clinical Interview Schedule, the Psychosis Screening Questionnaire and the Present State Examination were used in this study. Rates of mental illness for South Asians were lower although the researcher cautioned that the performance of the measuring instruments for this cultural group was questionable (29). In Canada, the Canadian Community Health Survey (CCHS) Cycle 1.1 revealed that immigrants had lower rates of depression and alcohol dependence, as measured by the short form versions of the Composite International Diagnostic Interview (30). Specifically, recent immigrants who arrived in the previous ten years had lower rates of depression than did the Canadian-born, and those who arrived less than 30 years ago had lower rates of alcohol dependence. The length of residence in Canada was somewhat confounded by the origin of the immigrants since recent immigrants were, much more likely to come from non-European.regions which were also associated with lower rates of depression and alcohol dependence. Asian immigrants were found to have the lowest rate of depression and African immigrants the lowest rate of alcohol dependence. The low rate of depression among Asian immigrants reported by the CCHS was contradicted by a survey of older Chinese in seven Canadian cities (25). Based on the Geriatric Depression Scale, this survey reported that 17.7% of Chinese 65 years and older were mildly depressed and 6.2% were moderately to severely depressed, figures which 6 Chapter 1 were higher than the estimated prevalence of depression among the general elderly population in Canada. A secondary analysis of this survey and data from the Canadian Multicentre Osteoporosis Study confirmed that Chinese aged 55 and above scored significantly lower than their national counterparts in the mental component scale of the SF-36(31). The various epidemiological findings on prevalence of mental disorders in the immigrant and ethnocultural minority populations are far from conclusive. Considering that the findings are derived from populations who differed in age, ethnic origin or duration of stay in Canada, and are based on different measurement tools, it is not surprising that they are inconsistent. In addition, the cross-cultural relevance of psychiatric diagnostic criteria and the validity of screening instruments used in population-based surveys are controversial. Given the outstanding debate, results from such studies must be treated with reservation. At the same time, clinical observation and theoretical discourse converge to suggest that certain migrant sub-populations are at higher risk of emotional distress, a compelling assertion which cannot be ignored in the provision of services (32, 33). Access to Mental Health Services Another body of research in response to the issue of under-utilization of mental health services focuses on the accessibility of services. Much research effort has been directed towards understanding the barriers to obtaining mental health services. Barriers that have been identified in the literature include language and cultural differences, discrimination and trust in relation to health professionals, stigma and shame, lack of knowledge regarding service availability, beliefs about the nature of the problem and treatment and logistic difficulties (transportation, time off from work), as well as issues related to practice and service delivery (34-41). The use of alternative therapies and reliance on family support are much less investigated, although they may also account for the under-utilization of formal medical services. Most research on barriers to access adopts qualitative designs or community surveys. There is little direct evidence that the identified barriers actually reduce the utilization of mental health services. To establish the role of potential barriers, 7 Chapter 1 information about factors associated with utilization is needed. Knowledge about such factors, or the determinants of utilization, can assist policy makers and service planners to target sub-groups in the population most at risk of being under-served by the mental health care system, and also point to the reasons for differential utilization and provide clues for developing intervention strategies. Summary of Key Points 1. Findings on the utilization of mental health services by immigrants and ethnic minorities are conflicting, but most indicate that these minority groups use fewer mental health services than does the general population. 2. Findings on the prevalence of mental disorders among immigrant and ethnic minority populations are also inconsistent; in addition, the validity of measurement instruments in screening for mental illness among culturally minority groups has been questioned. 3. Many studies suggest that immigrants and ethnic minorities encounter barriers to accessing mental health services. 4. There is little direct evidence that the identified barriers result in reduced use of mental health services. DETERMINANTS OF MENTAL H E A L T H SERVICE UTILIZATION: EVIDENCE FROM POPULATION SURVEYS A body of research about factors associated with mental health service utilization in the general population has accumulated from population-based surveys. This section synthesizes the findings from 15 recent studies on determinants of mental health service utilization. These studies are based on eight psychiatric epidemiological surveys in five western countries. The studies and data sources are listed below. 8 Chapter 1 Country Data Source Authors U.S.A. Epidemiologic Catchment Area (ECA), 1980-1982 Shapiro et al Galloetal Marino et al National Comorbidity Survey, 1990 Katz et al Kessler et al, 1998 Kessler et al, 2001 Canada Edmonton Survey of Psychiatric Disorders, 1984-1989 Bland et al Galbaud du Fort et al Ontario Health Survey - Mental Health Supplement, 1990 Katz et al Lin et al Survey in east end of Montreal, 1992-1993 Lefebvre et al Britain National Psychiatric Morbidity Surveys of Great Britain, 1993-1994 Bebbington, Meltzer et al, 2000 Bebbington, Brugha et al, 2000 Netherlands Netherlands Mental Health Survey and Incidence Study (NEMESIS), 1996-1997 Ten Have et al Australia National Survey of Mental Health and Wellbeing, 1997 Parslow & Jorm Meadows et al The most investigated factors with respect to mental health service utilization are sex, age, ethnicity, psychiatric disorder, disability or impairment, and, to a lesser extent, marital status, socio-economic status, availability of providers, region of residence, health beliefs and attitudes and perceived need for care. Sex Most studies in the US, Canada, Britain, Australia and the Netherlands demonstrate consistently that women are more likely than men to utilize services overall for mental health reasons (7, 16, 26, 42-47). This phenomenon was not explained by differences in mental health status; three studies reported that, even among those who had 9 Chapter 1 a psychiatric diagnosis, more women than men made mental health visits (42, 43, 46). One study found that women were more likely than men to consult family physicians for mental health problem even after controlling for severity of illness (16). On the other hand, one study in the US also found that, when they utilized services, men were more likely to visit mental health specialists whereas women were more likely to use general medicine (46). The exception to the finding of sex difference was the National Comorbidity Survey in the US. After controlling for age at onset, time since onset and cohort factors, women with depressive, anxiety and addictive disorders did not differ from men in their cumulative probabilities of seeking treatment (48). Moreover, there was no difference between men and women with serious mental disorders in terms of staying in treatment (49). Age Although some evidence supports that age affects rate of mental health service utilization, the exact pattern is less consistent. Studies in the US and Great Britain found that younger populations and older populations were less likely to make mental health visits than the middle age group (7, 16, 46). However, there were also differences in the use of general medicine versus mental health specialties between the young and the old (7, 46). Among those with serious mental illness, the young were also less likely to be in stable treatment (49). In accordance with the US data, among those who met criteria for a neurotic disorder in Britain, likelihood of mental health consultation in primary care increased by age and peaked at the 45-54 age group; individuals over 65 were not surveyed in this study (16). In Australia, after controlling for other factors, age was found to have no effect on visits to general practitioners and psychiatrists and to have a negative effect on mental health visits to psychologists and other non-physicians (45). Canadian surveys included non-medical resources, such as social workers, psychologists, clergy and self-help groups, in their definition of mental health services. In contrast to findings in the US, Edmonton residents with psychiatric disorders and aged 18-44 were more likely to seek help than their older counterparts (42). Similarly, the 25-44 age group in Ontario and French-speaking adults aged 18-44 in Montreal also used the 10 Chapter 1 most services for mental health reasons, although age was no longer a significant factor in Montreal after controlling for other determinants (26, 44). In addition to age at time of survey, age at first onset of depressive, anxiety and addictive disorders was also found to affect the likelihood of seeking treatment in Edmonton and the US (43, 48). The US survey also reported that the younger cohorts were more likely to seek treatment over the course of their illness. Marital and Household Status Most studies in Canada, Britain, Australia and the Netherlands concurred that living alone or being separated, widowed or divorced was associated with increased use of mental health services (16, 26, 44, 45, 47). The exception was the Edmonton survey where the effect of being separated, widowed or divorced was no longer significant after controlling for other variables (42). Among those with serious mental illness as well as depressive, anxiety and addictive disorders in the US, marital status was not found to contribute to making initial treatment contact or to dropping out (48, 49). Socio-economic Status (SES) Evidence regarding the relationship between socio-economic indicators and mental health service utilization is sparse. Moreover, as different measures of socio-economic status have been employed, findings cannot be compared across studies. In a US survey on first mental health contact in at least six months, it was found that while those in the lowest SES quartile were less likely to use general medicine without mental health consultation, the use of general medicine and mental health specialists did not differ by SES after adjusting for other factors (7). Furthermore, among those with serious mental illness, neither employment status, income nor education was found to have an effect on the likelihood of treatment (49). In Ontario, receiving social assistance was associated with increased use of mental health services while education had no significant relationship (26). There were also some interaction effects between receiving social assistance and urbanicity on the type of services used. In Edmonton, effect of unemployment on increasing use of mental health services disappeared after controlling for other variables (42). Among those with a neurotic disorder in Britain, 11 Chapter 1 those working full-time were less likely to consult family physicians for mental health problems than those working part-time or not employed (16). In Australia, having a higher education was positively associated with making any mental health visits, although income and employment status had no effect (45). However, in the Netherlands, less education was related to higher chance of incident care use. Lower income and having paid employment were related to more ongoing use (47). Race, Ethnicity and Immigration Status Race and ethnicity are frequently studied variables in mental health service utilization. As discussed in the previous section, some racial and ethnic minorities are found to under-utilize mental health services. In Britain, South Asians were less likely than Whites to have mental health consultations even when they met criteria for a neurotic disorder (16). In the US, it was often documented that racial and ethnic minorities were less likely to use any type of mental health services, although there was variation in the magnitude of the difference among the types of services (2-12, 14, 15). In the earlier E C A study, the discrepancy between the use of mental health specialists by Whites and minorities was even more profound when other factors were controlled; on the other hand, race and ethnicity had no effect on treatment among those diagnosed with serious mental illness in the more recent national survey (7, 49). Findings regarding immigrants from these epidemiological surveys are less definitive. No significant difference was found between utilization by immigrants and utilization by Canadian-born in Ontario (26). Neither was there a significant association between whether one's usual language was English and the use of various types of mental health practitioners in Australia (45). However, results of these surveys must be interpreted with caution. As noted in the previous section, the Ontario Health Survey included immigrants of all national origins and duration of residence in Canada and the Australian National Survey of Mental Health and Wellbeing excluded those who did not speak English. 12 Chapter 1 ' Beliefs and Attitudes Few studies in the west addressed the direct relationship of health beliefs and attitudes to help-seeking behaviours. Of the studies reviewed here, only the Montreal survey measured attitudes towards mental health services (44). Results showed that attitudes did not discriminate between users and non-users. Insurance Coverage None of the studies of mental health service utilization reviewed here addressed directly the effect of health insurance coverage, but inferences were made from a study comparing utilization between the US and Ontario. Little evidence was found that the universal health coverage available in Ontario resulted in excessive use of services by those with low mental morbidity and impairment (50). In fact, Americans had higher probability of mental health service use and the difference was specific to those who had no current disorders. Availability of Service and Urban-Rural Residence There appears to be some evidence that the availability of practitioners is related i to the higher utilization of that particular type of practitioners. In New Haven in the US, where the ratio of psychiatrist to population was high, the population more often turned to mental health specialists than to general medicine when they sought help (46). It was also observed in Quebec that the higher number of psychologists per capita in that province corresponded to higher utilization of these practitioners, relative to Ontario (44). Higher utilization in urban areas may also be a reflection of the higher concentration of medical resources. This was confirmed in Ontario where living in the urban area was associated with increased use of mental health services. Urban recipients of social assistance were three times as likely as rural recipients to use mental health services despite a much smaller difference in the prevalence of disorder (26). However, among the seriously mentally il l in the US, rural residents were more likely to receive treatment (49). 13 Chapter 1 Disorder Despite the significant associations found between socio-demographic variables and service utilization, all studies point to a diagnosed mental health problem as the strongest predictor of mental health service use. This relationship prevails, notwithstanding the observations that over 70% of those who fit the criteria for a mental health diagnosis and over 50% of those with serious mental illness do not use any services and that up to one-third of those who use services do not have any diagnosis (26, 44, 46, 49, 51-53). Data from all countries generally show that those who meet criteria for a psychiatric disorder have more use of inpatient and ambulatory care than the general population. Those with a recent or incident diagnosis also have more mental health visits than those with a past diagnosis. Those having a psychiatric disorder are also more likely to consult mental health specialists while those with no disorder are more likely to use general medicine. The type of diagnosis, number of comorbidities and severity of symptoms have been reported to affect not only the proportion of survey respondents who made mental health visits, but often the type of providers consulted as well (16, 26, 42-51,54). Perceived Need and Self-rated Mental Health In addition to the evaluated need by various diagnostic instruments, the perceived need of the individuals and their perception of their own mental health are also noted by several reports to have a strong impact on the use of mental health services. In both the US and Ontario, perceived need and self-rated mental health were related to increased use of any type of services. However, relative to Ontario, the level of perceived need was higher in the US among those with no mental health disorder and excellent self-rated mental health but lower among those with affective disorder and comorbidities; as a result, self-rated mental health and having a disorder were more predictive of service utilization in Ontario (50). In Montreal as well as Australia, self-identified mental health status was similarly associated with increased use of services (44, 45). Self-perception of mental health was second only to type of disorder in predicting service utilization in Montreal. The role of 14 Chapter 1 perceived needs was specifically examined in Australia. Results suggested that perceived need had a significant relationship with subsequent use of services among those with psychiatric diagnoses (53). Disability and Impairment The determinant of disability and impairment is sometimes referred to as burden of illness. In most studies, it is measured by the number of days when activities are restricted by the illness or decreased social or daily functioning. In both the US and Ontario, disability was found to relate to use of any services but it was the weakest predictor compared to having a disorder, perceived need and self-rated mental health (50). In Australia, days out of role was predictive of use of some types of services (45). In the Netherlands, increase in limited activity days, decline in social functioning and increase in unmet care needs were related to both incident and ongoing use of services (47). Contrary to the US and Ontario, burden of illness was found to be a stronger predictor than type of disorder in the Netherlands. In Britain, disability was measured by deficits in activities of daily living; its effect on the likelihood of mental health consultations among those who had a neurotic disorder was second only to symptom severity (51). The following table summarizes the evidence for determinants of mental health service utilization according to population surveys in several countries in the western world. Comparison across studies is not possible because of methodological differences in the inclusion criteria of the study population and the type of service utilization measured. For the same reason, the absence of evidence in a particular study, does not imply that the factor has no effect on any kind of mental health service utilization. 15 Chapter 1 Table 1.1 Summary of Findings on Determinants of Mental Health Service Utilization ECA NCS Ontario Edmonton Montreal Britain Australia Netherlands (US) (US) Sex * * * * * * * Age * * * * * Socio-economic status * * * * (income, employment) Education * * Marital Status * * * * * Ethnicity/ Immigration/ Language * * Health beliefs Insurance Coverage Urbanicity * * Service * * availability Psychiatric disorder/ comorbidity/ * * * * * * * * severity Perceived Need * * * * Disability * * * * * * Significant effect reported DETERMINANTS OF UTILIZATION SPECIFIC TO IMMIGRANTS AND ETHNIC MINORITIES Even though many studies include race or ethnicity and, to a lesser extent, immigration status, as a determinant of mental health service utilization, fewer studies specifically examine determinants of such utilization among ethnic minorities or immigrants. The complexity in establishing determinants within the diverse minority groups is demonstrated in the few studies that delve into the subject and inevitably show that the effects of various determinants differ by ethnicity (10, 12, 55). In fact, a simulation of the US National Medical Expenditure Survey data concluded that African 16 Chapter 1 Americans and Hispanics would have a probability of using mental health care similar to Whites if the determinants of utilization had the same effect on them (55). Many of the factors associated with the use of mental health service utilization in minority groups are common to the general population. They include sex, age, marital status, psychiatric disorder, disability/impairment, insurance coverage, region of residence, availability of service and education (5, 10, 12, 55-57). On the other hand, other factors investigated in the literature are particularly salient to immigrants and ethnic minorities. These are place of birth, acculturation, English language ability and perceived discrimination. Place of Birth The origin of immigrants and whether an ethnic minority individual was born overseas both play a role in the use of mental health care. In addition to the Australian study that reported the significantly different rates of representation in the mental health system of immigrants by place of origin (20), a study of refugees in the US also found that Cubans had higher rates of visiting mental health specialists than Haitians (58). Furthermore, surveys of both the Mexican American and Chinese American communities reported that, within these minority communities, the US-born were more likely to use mental health services than the foreign-born (56, 59). Acculturation While place of birth was treated at times as a crude indication of acculturation, two other studies of the Mexican American and Chinese American communities in the Los Angeles area that employed more sophisticated measurements of acculturation confirmed its role in influencing mental health care utilization. By the use of scales that incorporated several domains of culturally related behaviours and preferences (such as language use, contact with ancestral country, preferences in food, media and social contacts, ethnic identity), both studies found that higher level of acculturation was associated with higher probability of use of outpatient services or any resources for mental health problems (57, 60). 17 Chapter 1 English Ability Information about the effect of English language competency is scarce and only available from Australian studies. As reported before, analyses of data from the national mental health survey did not find any significant differences whether the respondents usually spoke English or between respondents from an English-speaking country (including the Australian-born) and those born in a non-English-speaking country in the receipt of mental health care (45, 61). However, the Australian national survey had severe limitations for the purpose of studying the effect of English language ability in that those who did not speak English were excluded. On the other hand, a census conducted with health care providers in Australia reported differences in the utilization of health services by English competency (62). For instance, patients who had difficulties with English were more likely seen in inpatient than outpatient care and more frequently diagnosed with psychosis and dementia and less often with less serious disorders. They were also less likely to receive psychotherapy or psychiatric consultations or see psychologists but more likely to receive psychotropic drugs. The results suggested that those who were not fluent in English sought treatment only for more serious problems or encountered barriers to some types of mental health care. Discrimination The role of discrimination or perceived discrimination is rarely investigated and, therefore, evidence is quite weak at this point. Nevertheless, in a psychiatric epidemiological survey of Chinese Americans in the Los Angeles area, respondents who reported experiences of unfair treatment because of their language or accent were also more likely to report using informal services or seeking help from friends and relatives for mental health problems than those who did not experience such discrimination. Perceived discrimination based on race, on the other hand, was not found to be related to service use (63). 18 Chapter 1 Summary of Key Points 1. The strongest predictors of mental health service utilization are disorder and perceived need; disability is also an influential factor, indicating a certain level of equity in the mental health service systems of these countries. 2. Women have consistently been found to utilize mental health services more often than men. 3. Living alone or being divorced, separated or widowed is also consistently related to more mental health service use. 4. Ethnic minorities use fewer mental health services, although there is variation by the type of services and by ethnic group. 5. The determinants of mental health service utilization differ by ethnicity. 6. Among ethnic minorities and immigrants, some additional determinants of utilization are place of birth and acculturation. English language ability may also affect the type of services received. CONCEPTUAL FRAMEWORK: BEHAVIOURAL MODEL OF H E A L T H CARE UTILIZATION Andersen's behavioural model of health care utilization is the most frequently cited conceptual model in the literature on determinants of mental health service utilization. Whether the studies are explicitly guided by this model or not, the determinants identified can all be summed up by the components of the model. The model is developed with the purpose of assessing the equity of access in health care and to identify components where intervention can improve access (64). Andersen's behavioural model of health care utilization was first published in 1968 to explain and assess the use of health services by family units in the United States. The initial model synthesized elements of other behaviour models at the time into three components - predisposing, enabling and need - and various sub-components (65). The predisposing component consisted of factors such as age, sex, family composition, social class, ethnicity and health belief. The enabling component included the family's resources to obtain health care and the availability of health care in the community. Finally, illness.and the family's response to the illness constituted the need component. 19 Chapter 1 Andersen's postulation was that the three components were stages that led to health care use, with the predisposing factors predicting the propensity of a family to use health services, the enabling factors permitting the family to use health services and the need factors acting as the immediate determinant of service use. He further hypothesized that the importance of each component depended on the discretion exercised by families in the decision. Andersen also acknowledged interrelationships between the components. For instance, the predisposing factor of socio-economic status would directly influence the amount of resources which enabled a family to acquire health care. The model was first applied to the 1964 data on US families' use of health services. The family was the unit of analysis in that study which calculated the association between the components of the model and all types of health service use including hospitals, physician visits and dental care. Hypotheses generated from the model were generally supported. The need component was found to have the strongest correlation with the volume of services, followed by the family composition (predisposing) and family resources (enabling) subcomponents. Different subcomponents played different roles in each type of service use. Since the 1960s, the model has undergone ongoing refinement and expansion by Andersen and other health service researchers (66-68). Some of the significant changes include: 1. The unit of analysis shifts from families to individuals. 2. In addition to individual components, the social and political context pertaining to health care, such as the organization of health care, health legislation, social norms and available technologies, are recognized as impacting individuals' decision to seek care; as a result, contextual characteristics are added to the model. 3. Both perceived need and evaluated need are included in the need component. Perceived need is influenced by the social context and indicated by self-reported health status and experience of symptoms; evaluated need is reflected in professional judgment of an individual's health status and need for medical care. 20 Chapter 1 4. Besides use of formal health care, personal health practice and the process of delivering care are considered health behaviours which result from the contextual and individual components. 5. Rather than health behaviours, the final stage of the model becomes health outcomes; health outcomes include perceived and evaluated health status and consumer satisfaction. 6. With the expanded model, expected interaction among the components becomes more complex. Feedback loops are incorporated into various components of the model. Figure 1.1 represents an adapted version of the latest behavioural model of health care utilization. Figure 1.1 Andersen's Behavioural Model of Health Care Utilization (Adapted from Andersen & Davidson, Improving Access to Care in America, 2001) DETERMINANTS OF HEALTH BEHAVIOURS HEALTH BEHAVIOURS HEALTH OUTCOMES Individual Characteristics Predisposing Enabling Need Contextual Characteristics Predisposing Enabling Need Personal Health Practices Process of Medical Care Use of Health Services Perceived Health Status Evaluated Health Status Consumer Satisfaction Examples of predisposing, enabling and need factors in the individual and contextual domains are organized by their subcomponents in Table 1.2. 21 Chapter 1 Table 1.2 Components in Andersen's Behavioural Model of Health Care Utilization INDIVIDUAL CHARACTERISTICS Predisposing Enabling Need Demographic (Individual) Age Sex Mafital Status Financing Income and savings Health insurance Perceived Need Perceived health status and response (influenced by ethnicity, education and health beliefs) Social Social Class Occupation Education Ethnicity Organization Regular source of care Nature of source of care Means of transportation Evaluated Need Professional assessment of health status (influenced by technology, clinical guidelines and practice patterns) Health Beliefs Values Attitude Knowledge CONTEXTUAL CHARACTERISTICS Predisposing Enabling Need Demographic (Community) Age Sex Marital Status Financing Per capita income Price of medical care Method of compensating providers Environmental Need Quality of housing, water and air Injury and death rates Social Employment level Crime rate Educational level Ethnic composition Proportion of recent immigrants Organization Distribution of facilities Ratios of physicians and hospital beds Hours and location of service Outreach programs Population Health Indices Mortality rates Morbidity rates Disability days Health Beliefs Community values Cultural norms Political perspectives 22 Chapter 1 Versatility is an advantage of this behavioural model; different components or subcomponents can be introduced that are pertinent to a specific population in each study. The model has been applied to underprivileged populations; the Behavioural Model for Vulnerable Populations adds "vulnerable domains" to the individual components of predisposing, enabling and need factors. Under these vulnerable domains are common concerns that affect the health behaviours of these populations, such as homelessness or access to transportation. Mental illness and substance abuse, while need factors when predicting the use of mental health and substance abuse services, are considered predisposing factors in relation to general health behaviour. In addition to ethnicity which is included in the original model, predisposing factors in this adapted model that are pertinent to this thesis and that have been supported by the literature review are country of birth, immigration, acculturation and literacy (69). The purpose of the health service utilization model is not only to predict or explain health service use, but also to assess the equity of access to health services in order to develop policies that promote equitable access. Different concepts of access are proposed. Potential access is the presence of enabling resources and realized access is the actual use of health services. Effective access takes place when use of health services improves health outcomes and efficient access is when maximum improvement is achieved at minimum expense (67, 68). Andersen defines equitable access as occurring when demographic variables such as age and sex, which correspond to illness levels, and the need variables account for most of the variance in utilization. Conversely, there is inequitable access when factors such as ethnicity, health beliefs and income determine the use of health care (67, 68). Access can be measured in terms of health care utilization or consumer satisfaction (64, 70). Since the behavioural model of health services utilization postulates that these two constructs are the consequences of a causal sequence of interrelated determinants, there are process indicators and outcome indicators for evaluating access. The process indicators correspond to variables that affect potential access, primarily factors in the predisposing and enabling components of the model. Examples in the contextual domain include the volume and distribution of health 23 Chapter 1 resources, such as the number of physicians and hospital beds, hours of operation and waiting times; examples of process indicators in the individual domain are knowledge of health services information, income and health insurance. Outcome indicators, on the other hand, correspond to variables associated with realized access, elements in the health behaviour and health outcome components. Examples of outcome indicators include type and amount of health services used, purpose of use, attitude about a doctor's visit and adherence to doctor's advice. To achieve the goal of equitable access, Andersen and his associates also propose the concept of "mutability" as the criterion for selecting factors that policy-makers should target for change (64, 66, 67). Mutability refers "to the extent to which a given component can actually be altered to influence the distribution of health services." (Andersen & Newman, p. 118) For an intervention to be effective in improving access, the relevant factor must play an important role in distribution of services and the factor must be amenable to change by health related policies. Most of the predisposing variables, such as demographics and social structure, are considered immutable. These indicators serve to highlight populations for whom access may differ but they are not amenable to change and are not likely targets to address inequity. An exception in the predisposing domain is the factor of health beliefs which is regarded to be of medium mutability. High mutability factors include most elements in the enabling component, such as health system variables. Need variables are not formally classified but Andersen suggests that perceived need may be altered by health education and changes in financial incentives whereas evaluated need can be altered by clinical practice guidelines (67). Elucidating the factors associated with mental health service utilization by Chinese immigrants and Canadian-born Chinese will contribute towards assessing whether access to mental health services by this underserved group is equitable. Although insurance coverage is not an issue in Canada for most medical services, it is acknowledged that there are still non-financial barriers to health care (34). In accordance with the behavioural model, factors that will be examined in the current thesis include the individual predisposing variables of sex, age, education, marital status, length of . residence in Canada and English ability, the contextual variables of the socio-economic 24 Chapter 1 status of the community and availability of Chinese-speaking physicians and the need factors of psychiatric diagnosis. While many of these factors are "immutable", knowledge of their relationship to service utilization will draw attention to sub-populations that are at risk of being underserved. The inequity may still be compensated by strengthening enabling resources in the health care environment, such as targeted outreach programmes or culturally responsive therapies. REASONS FOR CURRENT STUDY Limitations of Existing Research Evidence There are two usual sources of data in studies of mental health service utilization: surveys and administrative databases. A population-based survey is a common tool for collecting information about mental health service utilization of the general population. Notwithstanding the inherent risk for errors in self-reported data, the advantage of surveys is that their results can be generalized to the population since the data are derived from representative samples randomly drawn from known population bases. Administrative databases eliminate the risk of recall errors but can under-estimate utilization where comprehensive databases are not available and subjects may obtain care from multiple systems. Similarly, in the absence of a database that captures the entire population, results from administrative database analysis cannot be generalized to members of the population not included in the database. Presumably because of the difficulties in sampling small subgroups in the population, information about minorities' utilization of mental health services relies more often on administrative databases. When survey designs are employed, they tend to be confined to selected areas with high concentrations of minority residents, or adopt purposive sampling methods. As a result, generalizing the findings and obtaining a comprehensive understanding of a minority population's use of mental health services is not possible. Findings from different insurance databases or different surveys are not comparable because they cover different populations. Another approach to researching minorities' utilization of mental health services targets consumers within the mental health system. Since the total population eligible for services is unknown in this type of study design, these studies either describe only the 25 Chapter 1 types of services used by individuals already accessing the system or estimate the rate of utilization by using ethnic composition in census data as the denominator. ; • : . Research Gaps In the absence of definitive standards about the appropriate level of mental health service utilization, the state of affairs observed in the general population will have to serve as the norm against which equity and disparity are measured. In this regard, a body of research findings has accumulated to support the contention that certain minority groups are under-served by the mental health system relative to the majority population; however, there are still tremendous gaps in knowledge, especially in Canada. For mental health services to become equitable, there must be reliable information about where the shortfalls in services are. The gaps in knowledge can be summarized as follow: 1. There are few population-based studies that address the question of minorities' utilization of mental health services. Most of the existing research on the subject targets only a segment of the minority population as a result of limitations in data sources. 2. There are no population-based studies in Canada on mental health service utilization by any minority group. The only exception is the survey of older Chinese which is subject to the limitations of survey methodology. Much of the published literature is generated in other countries, the results of which may not apply in Canada because of differences in health care systems, immigration history and population demographics. 3. At the same time, few studies of mental health service utilization by minority groups attempt to distinguish between the effects of ethnicity versus immigration experience, perhaps a reflection of the significant overlap -between the ethnic minority population and the immigrant population where these studies are conducted. Immigrants and/or ethnic minorities are often treated as a monolithic group; however, such aggregation obliterates differences among subgroups of the study population and, in fact, confounds the results. 26 Chapter 1 4. Notwithstanding the evidence pointing to barriers to mental health care, little is known as to the extent to which the barriers affect the utilization of such care. Purpose of Current Thesis Research The current thesis research is undertaken as a first step towards addressing these gaps in knowledge. Owing to issues of data access, the scope of the study is limited only to British Columbia at present; however, the study can be replicated in other provinces and with other ethnic groups. The design of this study overcomes some of the limitations cited previously by: • Focussing on one specific ethnic group, thereby avoiding the confounding effects of ethnicity • Examining the difference between the Canadian-born and the foreign-born within this one ethnic group, clarifying the effects of immigration experience and ethnocultural origin • Using administrative databases as data sources for at least the immigrant population. The comprehensive databases available for this study capture not only most of the immigrants in the study population, but also most of their health care utilization, allowing conclusions to be drawn about the population. Administrative data for the Canadian-born are not available at this time, but data for this population are derived from a population-based survey. Chinese is selected as the study group since it represents the largest visible minority and immigrant group in British Columbia (BC). In 2001, Chinese comprised almost 10% of the people in BC, more than 40% of the visible minority population3 and 35% of the immigrants arriving in Vancouver that year came from Chinese origins4. The purpose of the thesis research is to confirm reports that Chinese under-utilize mental health services in this province relative to the general population and to identify general patterns of utilization, as well as factors associated with utilization. Some of the specific research questions are: 3 Statistics Canada 2001 Census 4 Citizenship and Immigration Canada, Facts and Figures 2001 http://www.cic.gc.ca/english/pub/facts2001/4van-02.html 27 Chapter 1 1. Do Chinese immigrants in Canada use fewer mental health services than the general population? 2. Do ethnic disparities in utilization exist among those with serious mental illness? 3. How does the utilization of mental health services differ between native-born minorities and immigrant minorities? 4. What are the factors that influence mental health service utilization among immigrants and ethnic minorities? 5. What are the possible explanations for ethnic disparities in utilization? Based on existing evidence, it is expected that Chinese immigrants and Canadian-born Chinese will have lower rates of overall mental health service utilization than the general population. The disparity is expected to be relatively smaller, however, among those with serious mental illness, since there are fewer alternative treatments for serious conditions. If language and cultural barriers contribute to the lower levels of service utilization, Canadian-born Chinese will be more likely to receive mental health services than immigrants. The magnitude of any difference can also be quantified in this study. Outcomes and ramifications of this study include: • Information about the actual rates of utilization of different types of mental health care by Chinese immigrants in BC, and the relative rates compared to the general population. This information is crucial to determine if there are disparities in utilization. • Information about predisposing, enabling and need factors associated with mental health service utilization by Chinese in B C . This information allows researchers and health policy makers to assess if there is equity in access to health care, according to Andersen's principles (67, 68). In the process, there will also be better understanding of sub-populations who are most at risk of under-utilization and the reasons for under-utilization, which is crucial to developing strategies and programs to remedy the discrepancies. Findings from this thesis, described in the following chapters, will contribute to the mosaic of knowledge about minorities' utilization of health care and provide a basis 28 Chapter 1 for health policy makers and service planners to implement effective and equitable mental health services and policies. Chapter 1 BIBLIOGRAPHY 1. Forging Our Legacy: Canadian Citizenship and Immigration, 1900-1977. Ottawa: Citizenship and Immigration Canada: 2000. 2. Cheung FK, Snowden LR. Community mental health and ethnic minority populations. Community Ment Health J 1990; 26: 277-291. 3. Snowden LR, Cheung FK. Use of inpatient mental health services by members of ethnic minority groups. Am Psychol 1990; 45: 347-355. 4. Cooper-Patrick L , Gallo JJ, Powe NR, Steinwachs D M , Eaton WW, Ford DE. Mental health service utilization by African Americans and Whites: The Baltimore Epidemiologic Catchment Area follow-up. Med Care 1999; 37: 1034-1045. 5. Padgett D K , Patrick C, Burns BJ, Schlesinger HJ. Ethnicity and the use of outpatient mental health services in a national insured population. Am J Public Health 1994; 84: 222-226. 6. Temkin-Greener H , Clark KT. Ethnicity, gender, and utilization of mental health services in a Medicaid population. Soc Sci Med 1988; 26: 989-996. 7. Gallo JJ, Marino S, Ford D, Anthony JC. Filters on the pathway to mental health care, II. Sociodemographic factors. Psychol Med 1995; 25: 1149-1160. 8. Harris K M , Edlund MJ , Larson S. Racial and ethnic differences in the mental health problems and use of mental health care. Med Care 2005; 43: 775-784. 9. Lasser K E , Himmelstein D U , Woolhandler SJ, McCormick D, Bor D H . Do minorities in the United States receive fewer mental health services than whites? Int J Health Serv 2002; 32: 567-578. 10. Padgett DK, Patrick C, Burns BJ, Schlesinger HJ. Ethnic differences in use of inpatient mental health services by blacks; whites and Hispanics in a national insured population. Health Serv Res 1994; 29: 135-153. 11. Meinhardt K , Vega W. A method for estimating underutilization of mental health services by ethnic groups. Hosp Community Psychiatry 1987; 38: 1186-1190. 12. Wells K B , Golding JM, Hough RL, Burnam M A , Karno M . Factors affecting the probability of use of general and medical health and social/community services for Mexican Americans and non-Hispanic whites. Med Care 1988; 26: 441-452. 13. Leong FTL. Asian Americans' differential patterns of utilization of inpatient and outpatient public mental health services in Hawaii. J Community Psychol 1994; 22: 82-96. 30 Chapter 1 -14. Chow JC-C, Jaffee K, Snowden L. Racial/ethnic disparities in the use of mental health services in poverty areas. Am J Public Health 2003; 93: 792-797. 15. Hu T-W, Snowden LR, Jerrell JM, Nguyen TD. Ethnic populations in public mental health: Services choice and level of use. Am J Public Health 1991; 81: 429-434. 16. Bebbington PE, Meltzer H , Brugha TS et al. Unequal access and unmet need: neurotic disorders and the use of primary care services. Psychol Med 2000; 30: 1359-1367. 17. Bhugra D, Leff J, Mallett R, Der G, Corridan B, Rudge S. Incidence and outcome of schizophrenia in Whites, African-Caribbeans and Asians in London. Psychol Med 1997;27:791-798. 18. Commander M J , Odell SM, Surtees PG, Sashidharan SP. Characteristics of patients and patterns of psychiatric service use in ethnic minorities. Int J Soc Psychiatry 2003; 49:216-224. 19. Crowley JJ, Simmons S. Mental health, race and ethnicity: a retropecstive study of the care of ethnic minorities and whites in a psychiatric unit. J Adv Nurs 1992; 17: 1078-1087. 20. Klimidis S, McKenzie DP, Lewis J, Minas EH. Continuity of contact with psychiatric services: immigrant and Australian-born patients. Soc Psychiatry Psychiatr Epidemiol 2000; 35: 554-563. 21. Peters R. Increasing the responsiveness of mainstream mental health services to ethnocultural minorities. In: Health and Cultures Vol. II Programs, Services and Care. Edited by Masi R, Mensah L , McLeod K A . Oakville: Mosaic Press: 1993: 187-203. 22. Intercultural Committee. Multiculturalism and Mental Health: Developing Culturally Competent Systems of Care. Vancouver: Greater Vancouver Mental Health Service: 1999. 23. Roberts N , Crockford D. Psychiatric admissions of Asian Canadians to an adolescent inpatient unit. Can J Psychiatry 1997; 42: 847-851. 24. Kirmayer LJ , Galbaud du Fort G, Young A, Weinfeld M , Lasry J-C Pathways and Barriers to Mental Health Care in an Urban Multicultural Milieu: An Epidemiological and Ethnographic Study. Culture & Mental Health Research Unit Report 6 (Part 1). Montreal: Culture & Mental Health Research Unit, Sir Mortimer B. Davis - Jewish General Hospital: 1996. 25. Lai DWL, Tsang KT, Chappell N L , Lai D C Y , Chau SBY. Health and Well Being of Older Chinese in Canada. Calgary: Faculty of Social Work, University of Calgary: 2003. 31 Chapter 1 26. Lin E, Goering P, Offord DR, Campbell D, Boyle M H . The use of mental health services in Ontario: epidemiologic findings. Can J Psychiatry 1996; 41: 572-577. 27. Zhang A Y , Snowden LR. Ethnic characteristics of mental disorders in five U.S. communities. Cultural Diversity and Ethnic Minority Psychology 1999; 5: 134-146. 28. Takeuchi DT, Chung R C - Y , Lin K - M et al. Lifetime and twelve-month prevalence rates of major depressive episodes and dysthymia among Chinese Americans in Los Angeles. Am J Psychiatry 1998; 155: 1407-1414. 29. Nazroo JY. Rethinking the relationship between ethnicity and mental health: the British Fourth National Survey of Ethnic Minorities. Soc Psychiatry Psychiatr Epidemiol 1998; 3.3: 145-148. 30. Al i J. Mental Health of Canada's Immigrants. Health Reports 82-003 13 Supplement. Statistics Canada: 2002. 31. Lai DWL. Health status of older Chinese in Canada. Can J Public Health 2004; 95: 193-197. 32. Bhugra D. Migration and mental health. Acta Psychiatr Scand 2004; 109: 243-258. 33. Review of the Literature on Migrant Mental Health. Canadian Task Force on Mental Health Issues Affecting Immigrants arid Refugees: 1988. 34. Bowen S. Access to Health Services for Underserved Populations in Canada. "Certain Circumstances": Issues in Equity and Responsiveness in Access to Health Care in Canada. Ottawa, Ontario: Health Canada: 2000. 35. Green G, Bradby H , Chan A, Lee M , Eldridge K. Is the English National Health Service meeting the needs of mentally distressed Chinese women? J Health Serv Res Policy 2002; 7: 216-221. 36. Leong FTL, Lau ASL. Barriers to providing effective mental health services to Asian Americans. Mental Health Services Research 2001; 3: 201-214. 37. L i HZ, Browne AJ. Defining mental illness and accessing mental health services: Perspectives of Asian Canadians. Can J Commun Ment Health 2000; 19: 143-159. 38. L i P-L, Logan S, Yee L, Ng S. Barriers to meeting the mental health needs of the Chinese community. J Public Health Med 1999; 21: 74-80. 39. Miranda J, Lawson W, Escobar J. Ethnic minorities. Mental Health Services Research 2002;4:231-237. 40. Phan T. Investigating the use of services for Vietnamese with mental illness. J Community Health 2000; 25: 411-425. 32 Chapter 1 41. Takeuchi DT, Leaf PJ, Kuo H-S. Ethnic differences in the perception of barriers to help-seeking. Soc Psychiatry Psychiatr Epidemiol 1988; 23: 273-280. 42. Bland RC, Newman SC, Orn H . Help-seeking for psychiatric disorders. Can J Psychiatry 1997; 42: 935-942. 43. Galbaud du Fort G, Newman SC, Boothroyd LJ , Bland RC. Treatment seeking for depression: role of depressive symptoms and comorbid psychiatric diagnoses. J Affect Disord 1999; 52: 31-40. 44. Lefebvre J, Lesage A, Cyr M , Toupin J, Fournier L . Factors related to utilization of services for mental health reasons in Montreal, Canada. Soc Psychiatry Psychiatr Epidemiol 1998; 33: 291-298. 45. Parslow RA, Jorm AF. Who uses mental health services in Australia? An analysis of data from the National Survey of Mental Health and Wellbeing. Aust N Z J Psychiatry 2000; 34: 997-1008. 46. Shapiro S, Skinner EA, Kessler L G et al. Utilization of health and mental health services: three Epidemiologic Catchment Area sites. Arch Gen Psychiatry 1984; 41: 971-978. 47. ten Have M , Vollebergh W, Bijl R V , de Graaf R. Predictors of incident care service utilisation for mental health problems in the Dutch general population. Soc Psychiatry Psychiatr Epidemiol 2001; 36: 141-149. 48. Kessler RC, Olfson M , Berglund PA. Patterns and predictors of treatment contact after first onset of psychiatric disorders. Am J Psychiatry 1998; 155: 62-69. 49. Kessler RC, Berglund PA, Bruce M L et al. The prevalence and correlates of untreated serious mental illness. Health Serv Res 2001; 36: 987-1007. 50. Katz SJ, Kessler RC, Frank RG, Leaf P, Lin E, Edlund M . The use of outpatient mental health services in the United States and Ontario: the impact of mental morbidity and perceived need for care. Am J Public Health 1997; 87: 1136-1143. 51. Bebbington PE, Brugha TS, Meltzer H et al. Neurotic disorders and the receipt of psychiatric treatment. Psychol Med 2000; 30: 1369-1376. 52. Manderscheid RW, Rae DS, Narrow WE, Locke BZ, Regier DA. Congruence of service utilization estimates from the Epidemiologic Catchment Area Project and other sources. Arch Gen Psychiatry 1993; 50: 108-114. 53. Meadows G, Burgess P, Fossey E, Harvey C. Perceived need for mental health care, findings from the Australian National Survey of Mental Health and Well-being. Psychol Med 2000; 30: 645-656. 33 Chapter 1 54. Marino S, Gallo JJ, Ford D, Anthony JC. Filters on the pathway to mental health care, I. Incident mental disorders. Psychol Med 1995; 25: 1135-1148. 55. Freimen MP, Cunningham PJ. Use of health care for the treatment of mental problems among racial/ethnic subpopulations. Med Care Res Rev 1997; 54: 80-100. 56. Vega W A , Kolody B, Aguilar-Gaxiola S, Catalano R. Gaps in service utilization by Mexican Americans with mental health problems. Am J Psychiatry 1999; 156: 928-934. 57. Wells K B , Golding JM, Hough RL, Burnam M A , Karno M . Acculturation and the probability of use of health services by Mexican Americans. Health Serv Res 1989; 24: 237-257. 58. Portes A, Kyle D, Eaton WW. Mental illness and help-seeking behavior among Mariel Cuban and Haitian refugees in south Florida. J Health Soc Behav 1992; 33: 283-298. 59. Ying Y-W, Miller LS. Help-seeking behavior and attitude of Chinese Americans regarding psychological problems. Am J Community Psychol 1992; 20: 549-556. 60. Tabora B L , Flaskerud JH. Mental health beliefs, practices, and knowledge of Chinese American immigrant women. Issues Ment Health Nurs 1997; 18: 173-189. 61. Pirkis J, Burgess P, Meadows G, Dunt D. Access to Australian mental health care by people from non-English-speaking backgrounds. Aust N Z J Psychiatry 2001; 35: 174-182. 62. Stuart GW, Minas HI, Klimidis S, O'Connell S. English language ability and mental health service utilisation: a census. Aust N Z J Psychiatry 1996; 30: 270-277. 63. Spencer M S , Chen J. Effect of discrimination on mental health service utilization among Chinese Americans. Am J Public Health 2004; 94: 809-814. 64. Aday L A , Andersen R. A framework for the study of access to medical care. Health Serv Res 1974; 9: 208-220. 65. Andersen R. A Behavioral Model of Families Use of Health Services. Center for Health Administration Studies, Research Series 25. The University of Chicago: 1968. 66. Andersen R, Newman JF. Societal and individual determinants of medical care utilization in the United States. Milbank Mem Fund Q 1973; 51: 95-124. 67. Andersen R M . Revisiting the behavioral model and access to medical care: Does it matter? J Health Soc Behav 1995; 36: 1-10. 34 Chapter 1 68. Andersen R M , Davidson PL. Improving access to care in America. In. Changing the US Health Care System. Edited by Rice TH, Kominski GF. San Francisco: Jossey-Bass: 2001: 3-30. 69. Gelberg L , Andersen R M , Leake B D . The behavioral model for vulnerable populations: application to medical care use and outcomes for homeless people. Health Serv Res 2000; 34: 1273-1302. 70. Aday L A , Andersen R, Fleming GV. Health Care in the U.S.: Equitable for Whom? Beverly Hills: Sage Publications: 1980 35 CHAPTER 2 DISPARITIES IN UTILIZATION OF MENTAL H E A L T H SERVICES BY CHINESE IMMIGRANTS IN BRITISH COLUMBIA 1 Trends in international migration have evolved in the last century. Owing to a combination of factors involving global socio-political climate and ideological and legislative changes in the West, the majority of recent immigrants to western nations no longer share common ethnic origins with the established residents in the host countries., Canada is one such country. Since changes in its immigration policies in 1967, the ethnic makeup of the immigrant population in Canada has rapidly transformed. While the early settlers in Canada were predominantly of European origin, Asia and other continents became the major contributors of new immigrants by the end of the 20 t h century. This is especially true in British Columbia (BC), the westernmost province. With its proximity to Asia, it is not surprising that 74% of the immigrants arriving in 2001 came from Asian countries. Among these new Canadians, Chinese comprised the largest ethnocultural group, with People's Republic of China, Taiwan and Hong Kong contributing 35% of the 34,000 new immigrants to Vancouver.2 The dramatic shift in the ethnic background of newcomers consequently brought about gradual changes in the overall ethnocultural composition of the host society as well. In 2001, 26% of the 3.9 million residents in B C were immigrants; 24% of these immigrants were born in a Chinese territory and 60% of them immigrated within the previous decade.3 Such ethnocultural diversity challenges service providers, especially in the health / sector. Professional staff are frustrated by their inability to adequately serve minority clients while ethnocultural communities often criticize the health care system as being culturally insensitive, i n mental health care, there have been many concerns expressed by mental health service providers and immigrant advocates about the disparities in utilization of mental health services between newcomers, especially the Chinese immigrants, and the rest of the population. 1 A version of this chapter will be submitted for publication in a journal. 2 Citizenship and Immigration Canada Facts and Figures 2001 http://www.cic.gc.ca/english/pub/facts2001/4van-02.html 3 Statistics Canada Census 2001 Cat. No. 97F0009XCB2001002 http://wwwl2.statcan.ca/english/cens us01/Products/standard/themes/DataProducts.cfm?S=l 36 Chapter 2 Despite the growing concern, knowledge about the use of mental health services by Chinese immigrants is limited. Some reports on all immigrants' use of general health care indicate that new immigrants are healthier than longer-term immigrants and use less health care (1-4). However, the health status of the new immigrants declines after arrival and their use of health care increases accordingly, such that there are no statistically significant differences between longer-term immigrants and the native-born. On the other hand, even when the overall use of medical services is not different between immigrants and non-immigrants, the use of health services for psychological distress is much lower among immigrants (5). In terms of Chinese immigrants' use of mental health services, few studies can be found in the published literature. The little evidence available suggests that immigrants and Asians, including Chinese, under-utilize mental health services relative to their need compared to other groups in the population. For instance, Asian Americans consistently showed a lower probability of using mental health services although the level of use might not differ once they entered the system (6-10). Among Australian patients with schizophrenia or bipolar disorder, immigrants were under-represented in inpatient and outpatient services compared to the Australian-born and Asian immigrants had the lowest level of utilization (11). In Canada, censuses of patients in Vancouver's community mental health centres' caseloads in 1988 and 1998 showed that Chinese and South Asians were under-represented (12,13). Similarly, an adolescent inpatient psychiatric unit in Calgary reported that there were fewer Asians admitted than the city's population demographics would indicate, and the ones admitted suffered from more severe disorders (14). A national survey of Chinese aged 55 years and over in seven Canadian cities also reported that older Chinese were less likely than older adults in the general population to visit psychologists and psychiatrists (15). However, most of the Canadian data were either collected from specific treatment facilities or based on survey design; therefore, results were either limited in generalizability or subject to response biases. In light of the growing size of the Chinese immigrant population in B C and the concern over equitable access to services, this study aims to provide population-based 37 Chapter 2 evidence on the current state of disparities in mental health service utilization by this minority group. The specific objectives of the study are to: • describe the rates of utilization by Chinese immigrants of mental health services, including individual consultations with general practitioners, consultations with psychiatrists, hospitalizations and community mental health • compare the rates of utilization between Chinese immigrants and sex and age-matched comparison group members METHODS Study Design This study adopted a retrospective observational cohort design. Subjects were identified from a national immigration database and from the provincial health plan registry. Their utilization data for physician, hospital and community mental health services for the study period of 1992-2001 were extracted from population-based provincial health administrative databases. Data Sources Two administrative data systems provided the data for this study: Landed Immigrant Data System (LIDS) The database from Citizenship and Immigration Canada (CIC) consists of individual information on all immigrants who landed in B C between 1985 and 2000. BO Linked Health Data (BCLHD) The B C L H D is a linked database of most administrative health records in the province. Databases used in this study were the Medical Services Plan (MSP) Registry, MSP Payments, Hospital Separations, Vital Statistics-Death and the Mental Health Services files. MSP Registry: The Medical Services Plan is the publicly-funded provincial health insurance plan which covers all BC residents, including legal immigrants. In this study, registration in the health plan was taken as indication of residency in the province. Time periods when an individual was not registered in the plan were not included in the analyses. 38 Chapter 2 Vital Statistics Death: The date of death was used to determine the end point of participation for subjects who died. MSP Payments: The MSP Payments file consists of all claims made by health practitioners for services insured by the health plan. Since most general practitioners in B C are compensated on a fee for service basis, the Payments file contains records of most services received by B C residents from general practitioners (16). A caveat in using this database to estimate psychiatric visits is that only 70% of services provided by psychiatrists are compensated this way and are expected to be documented in this file. While the actual rate of psychiatric visits calculated from this database may under-estimate the true rate of contacts between subjects and psychiatrists, the rate ratio between immigrants and comparison group members is expected to provide an accurate representation of the relative utilization of the two study populations since there is no indication that psychiatrists who treat immigrants and non-immigrants are compensated differently. The remainder of psychiatric services not covered by fee for service is paid through salary or sessional arrangements and their utilization is reflected in the Mental Health Services database described below. Hospital Separations: The Hospital Separations file contains records of all separations from acute care and extended care facilities in B C , including day procedures. Mental Health Services: The Mental Health Services file contains records of most activations in BC's mental health system. Each activation represents an episode of care in one of the system's programs and may contain numerous contacts over a lengthy period of time. A client may have simultaneous activations in different programs. Information on individual contacts is not currently available in this database. The database includes services in outpatient mental health programs as well as services provided through the few inpatient facilities under the jurisdiction of the mental health system, such as the province's only tertiary care psychiatric hospital. The immigration and health databases were linked at an individual level by an independent research centre as part of a national immigrant health study, using probabilistic linkage methods (17). The linkage allowed utilization records of individual subjects to be extracted and studied. Identifiable personal information was removed before the linked data were released for analysis. 39 Chapter 2 Subjects The study immigrants were identified from the immigration database. They comprised immigrants from China, Taiwan, Hong Kong or Macau who landed in B C in 1985-2000 and who were registered in the provincial health plan for some time during the study period of 1992-2001. A total of 152,184 immigrants were identified. Each linked immigrant was matched by sex, year of birth and area of residence to a comparison subject who was randomly selected from the health plan registry, excluding those in the immigration database. To minimize the bias that could occur when comparing individuals observed during different points in time, only the utilization records in years when both the immigrant and comparison subject in a matched pair were registered were included in the analysis. This resulted in the exclusion of 3211 (2.1%) pairs of immigrant and comparison subjects because they had no common years of health plan registration. Hence, the final study population consisted of 148,973 pairs of immigrants and comparison group members, or 297,946 individual subjects. Outcome Variables Four types of outcomes were examined in this study. Two related to physician services: the number of mental health visits to general practitioners and the number of visits to psychiatrists. The other two were number of mental health hospitalizations and length of outpatient mental health care activations. A mental health visit to a general practitioner was defined as all the inpatient and outpatient services for which a general practitioner submitted payment claims for a patient in one day where the primary diagnosis or service item was related to mental health. A visit to a psychiatrist included all inpatient and outpatient services claimed by a psychiatrist for a patient in one day. A mental health hospitalization was either an inpatient episode in the Mental Health Services system or a discharge record in the Hospital Separations database which met one of the following criteria: (a) the most responsible diagnosis was psychiatric, (b) the patient service code was psychiatry, (c) the most responsible physician specialty was psychiatry, or (d) electroconvulsive therapy was performed during the stay. Finally, use 40 Chapter 2 of outpatient mental health care was measured by the number of years of an activation in the Mental Health Services database where the location of the care was an outpatient setting. Only utilization records and time which occurred while the individual was participating in the study were included. That is, only utilization in years when both the matched subjects were registered in the health plan was counted. Data Analysis The rates of all service utilization outcomes in immigrants and comparison subjects were estimated as the total number of events or time in care divided by the total time registered in the health plan registry. In addition to rates for the total population, the estimation was also stratified by sex and age group at entry to study to accommodate sex and age variation in service utilization. Disparity for each outcome was measured in relative terms by a rate ratio, obtained by dividing the utilization rate in the immigrant group by the rate in the comparison group. The 95% confidence interval for each estimate was determined by bootstrapping with 1000 replications. Bootstrapping computes the variance of the data by resampling the original data with replacement and the bootstrap distribution of the estimated parameter approximates the sampling distribution (18). The bootstrap method took into account the matched subjects design in this study by resampling the matched pairs as a unit. Age-standardized rates for each outcome were calculated for the entire comparison group and also stratified by sex. Since the comparison members were matched by age to the immigrants, they represented a slightly younger population than the general population in BC. Age-standardized rates estimated the utilization of the non-immigrant population in the province and age-standardized rate ratios approximated the actual difference between the Chinese immigrant and the non-immigrant populations. The comparison rates were directly standardized to the 1996 census population of B C . 4 Rates for the Chinese immigrants were not standardized since the entire population of recent Chinese immigrants was included in the study. 4 Age standardization of comparison group rates was carried out by five age groups: 0-14, 15-24, 25-44, 45-64 and 65 or above. 41 Chapter 2 The diagnostic categories of mental health visits to general practitioners and psychiatrists were also tabulated to provide an overview of the reasons for mental health consultations and differences between immigrants and comparison group members. The frequency of each diagnostic category, the percentage of the total number of visits and the number of individuals involved in each category were calculated. The 95% confidence interval of the percentage of each diagnostic category was estimated by treating the multiple visits of each individual as a cluster. RESULTS Description of Subjects A total of 148,973 pairs of Chinese immigrants and comparison subjects matched by sex and age were included in this study. Their characteristics are shown in Table 2.1. This population was almost evenly divided between males and females and consisted mostly of younger adults. Over 80% of the immigrants arrived after 1991 and the average length of time in the study was over five years. Rate of Mental Health Visits to Physicians Table 2.2a shows the combined number of mental health visits per 100 person-years made by Chinese immigrants and comparison subjects to general practitioners and psychiatrists in 1992-2001. Immigrants and the comparison group shared some similarities in utilization by sex and age. For both groups, men made fewer visits than women and the under 15 age groups had the lowest rate of visits. However, immigrants of every sex and age group had lower rates of mental health visits than the comparison group. Overall, immigrants made only 19% as many physician visits for mental health reasons as the comparison group in the study. The rate ratios between immigrants and the comparison group were lowest for the 15-24 and 25-44 age groups among women and for the 25-44 and 45-64 age groups among men. Rates of mental health visits to general practitioners displayed a similar pattern by sex and age, as shown in Table 2.2b. The contrast between immigrants and comparison subjects was greater among men than among women, with immigrant men making only 17% as many visits as men in the comparison group while the corresponding figure for 42 Chapter 2 women was 24%. Among immigrants, the rate of mental health visits to general practitioners was highest in the older age groups. Immigrant women 25 and over had five to six times as many visits as the under 15 group although the differences among the adult women 25 and over were not significant. Adult immigrant men also had higher rates than the under 15 group, with the highest rate found in the elderly; rate for men 65 and over was about four times that for children and 50% higher than youth and younger adults. In contrast, the most frequent users among men and women in the comparison group were found in the 25-44 age group. As a result, the difference between immigrants and the comparison group was greater for women in the 15-24 and 25-44 age groups and for men in the 25-44 and 45-64 age groups; the difference was relatively smaller in the oldest and youngest age groups for both men and women. Visits to psychiatrists presented a different pattern, as shown in Table 2.2c. In general, both immigrant and comparison group women were more likely to have psychiatric visits than men, with the exception of the youngest age group under 15 and, for immigrants, the age group of 15-24 as well. At an overall rate ratio of 0.14, immigrants of both sexes and all age groups had lower rates of psychiatric visits than the comparison group and the relative level of disparity did not differ between sexes. There was no consistent pattern of age group difference among the immigrants, with young immigrant males 15-24 years of age showing the highest rate. The comparison group demonstrated a pattern similar to that for general practice visits, with the 25-44 and 45-64 groups having the highest rates. In summary, Chinese immigrants were much less likely than the comparison subjects to be treated for mental health conditions by physicians. The relative disparity was greater in psychiatric visits than in visits to general practitioners. With regard to the latter, the disparity was also greater among men than women. Differences in level of disparity were also observed among age groups, with disparities generally lower among the youngest and the oldest subjects and higher among the young and middle-aged adults. Rate of Mental Health Hospitalization Use of mental health services in acute care and inpatient psychiatric facilities was quite infrequent, as shown in Table 2.3. Overall, Chinese immigrants were much less 43 Chapter 2 likely than the comparison group to be hospitalized for psychiatric reasons, at a rate ratio of only 0.14. The relative disparity was greater for men, who had a rate ratio of 0.10, than for women, with a rate ratio of 0.17. Immigrant women were more likely than men to experience mental health hospitalization while there wasTao significant difference between comparison group women and men. Understandably, the under 15 groups for both sexes among immigrants and the comparison group had the lowest rates of hospitalization in general, with the exception being the 45-64 group of immigrant males who had a rate lower than the children. The highest rates were found in the 65 and over groups among comparison group men and women and also among immigrant men; however, for immigrant women, the 15-24 had a higher rate, albeit not statistically significant, than the older age groups. Comparing immigrant and comparison group members, the lowest rate ratios for both men and women were found in the older age groups, although the results were not always statistically significant. Among women, the 65 and over group had the lowest rate ratio, indicating greatest disparity among elderly women; among men, the lowest ratio was in the 45-64 age group. In summary, Chinese immigrants were much less likely than the comparison group to be hospitalized for mental health reasons. The relative difference was greater among men than women and among older adults than younger individuals. Rate of Outpatient Mental Health Care The use of outpatient care in community mental health centres was also rare, as illustrated in Table 2.4. Chinese immigrants were again far less likely than the comparison group to use these services, the overall rate ratio of time engaged in care being 0.13. There was no difference in relative disparity between men and women. Both immigrant and comparison group women were more likely than men to be using services in mental health programs, even though the difference between immigrant women and men did not reach statistical significance. Among immigrant women, the most frequent users were the 25-44 age group whereas, among immigrant men, the 15-24 age group had the highest rate. Among comparison group women and men, the highest utilization was found respectively in those aged 45-64 and those aged 25-44. Age differences in rate ratios were not statistically significant. 44 Chapter 2 In summary, Chinese immigrants were much less likely than comparison group members to receive services from community mental health centres. The level of relative disparity was similar for men and women and there were no statistically significant age differences. Reasons for Mental Health Consultation Table 2.5a summarizes, by diagnostic categories, all the mental health visits made to general practitioners and psychiatrists between 1992 and 2001 in the years when both the immigrant and comparison group member in a matched pair were registered in the health plan. It corroborates the previous findings on rates of utilization that immigrants had far fewer mental health consultations than the comparison group and it illustrates the differences in reasons for consultations between the two groups. For Chinese immigrants, 35% of their overall mental health consultations were for anxiety/depression, a category unique to BC's health plan and covers a variety of non-specific depressive and anxiety symptoms. Other major mental health reasons for which the immigrants visited general practitioners or psychiatrists were neurotic disorders and depressive disorder not elsewhere classified. While these three diagnoses also accounted for the majority of mental health consultations by the comparison group, 17% of the comparison subjects' visits were for drug dependence, a condition which comprised less than 2% of immigrant visits. Differences between practice pattern of general practitioners and psychiatrists are apparent from Tables 2.5b and 2.5c, which show the diagnostic categories of mental health visits to these two specialty types separately. Most of the depression/anxiety and drug dependence conditions were treated by general practitioners. Among immigrants, almost half of their mental health consultations with general practice were for anxiety/depression while the top reason for the comparison group's visits to primary care was drug dependence. With respect to psychiatric visits, the main reasons for consultation, among immigrants, were affective psychoses, neurotic disorders, schizophrenic psychoses and depressive disorder not elsewhere classified. In the comparison group, both neurotic disorders and depressive disorder not elsewhere 45 Chapter 2 classified remained the top reasons for visits to psychiatrists, followed by affective psychoses. DISCUSSION This study is the first comprehensive report of mental health service utilization by a population of Chinese immigrants in a Canadian province. The findings confirm previous reports and anecdotal evidence that Chinese immigrants are under-represented in the mental health service system relative to non-immigrants. Data from this study also yield quantitative evidence of the substantial disparities in utilization of four types of mental health services between Chinese immigrants and the general population in BC. During 1992-2001, Chinese immigrants were only 13% to 21% as likely as age-matched comparison group members to use formal mental health services. Disparities were lowest in services provided by general practitioners and higher in psychiatric services, hospitalization and outpatient community mental health care. With respect to general practice and hospital care, disparities were also higher among men than women. Presuming the level of emotional disturbance is the same between Chinese immigrants and the rest of the population, this discrepancy raises the alarm that there are gross inequities in services. However, the alternative argument is that Chinese immigrants have lower rates of mental disorders, at least as defined by current psychiatric diagnostic schemes. This argument is supported by evidence from epidemiological surveys which indicate lower prevalence of mental disorders in the source countries of Chinese immigrants - Taiwan, Hong Kong and China - than in most other countries and also lower prevalence of depression and dysthymia among Chinese Americans in Los Angeles (19-22). It is conceivable that, even burdened with the stresses of migration and resettlement in a new country, their resilience to psychological difficulties protected them from the higher prevalence of emotional problems found in the host population. This study also finds smaller disparities between immigrants and the comparison subjects among women, youths and young adults and the elderly, indicating that these groups have relatively higher rates of mental health service utilization than the other Chinese immigrants. This is consistent with the literature suggesting that these are the most vulnerable sub-groups among immigrants (23). Conflicts in gender roles and 46 Chapter 2 changes to traditional family patterns affect women adversely. The emotional demands during the developmental stage of adolescence and young adulthood are further compounded by cultural adjustments and identity crises. Decreased ability to adapt, change in family roles and lower levels of English competency and mobility combine to place the elderly immigrants at higher risk. On the other hand, it cannot be ruled out that Chinese irrimigrants encounter barriers to mental health services. Although direct financial barriers are not expected to exist in the health services examined here, there may be indirect financial barriers (e.g. child care costs, transportation, unpaid time off work), language barriers, information barriers, cultural or education barriers. Cultural barriers include differences in health beliefs and values, racism and ethnocentrism in the health system and differences in effectiveness of prescribed treatment. Misdiagnosis and inappropriate treatment are also likely consequences when there are cultural and language differences (24). Thus, even when Chinese immigrants experience emotional difficulties, they may fail to use available mental health services because they do not recognize their problems as health issues, not know where to seek help, refrain from seeking help for fear of the stigma in their community or find the recommended therapies not helpful. The disparity is smallest in visits to general practitioners presumably because there are fewest barriers to primary care. Chinese immigrants living in metropolitan Vancouver have access to many general practitioners who share with them a common cultural heritage and language, reducing some of the potential barriers. The fact that disparity is also smaller among women and elderly may be explained by the fact that these sub-populations have more opportunities to interface with the health care system because of other medical concerns and, therefore, are more likely to have their mental health problems recognized and treated. However, the relative "advantage" of these groups in primary care does not extend into specialist services, as shown in this study. Immigrants and the general population differ not only in the rates of mental health service utilization, but also in the reasons for utilization. Although the diagnostic patterns in the payment claims were influenced, to some extent, by the diagnostic preferences of physicians, they indicated nevertheless that irrimigrants in this study used general practice mental health services primarily for depressive, anxiety and neurotic 47 Chapter 2 symptoms, while over 40% of their visits to specialist mental health services were for more serious psychotic disorders. In contrast, the top reason that the comparison group received mental health services from general practice was drug dependence, a condition which was quite rare among the immigrants. Moreover, only 24% of the comparison group's visits to psychiatrists were for psychotic conditions whereas a higher proportion were for neurotic and depressive disorders. These differences shed light on some other factors that contribute to the disparity in mental health service utilization between immigrants and the comparison group. One is that the lower rate among immigrants of service utilization for substance-related problems accounts for a sizeable portion of the overall disparity in utilization between immigrants and the comparison group, particularly in general practice. The second factor is that immigrants tend to use specialist services for more serious but rare psychiatric conditions, rather than for the more common but less serious disorders. Since the less serious disorders such as depressive and neurotic conditions comprise the majority of the visits to psychiatrists, the disparity in psychiatric service utilization between immigrants and the comparison group is further magnified. There are limitations to this study. Foremost is that this study is based on databases of service utilization records. While these databases provide a comprehensive picture of utilization, they do not offer any information about the "need" for services, that is, the prevalence of mental health difficulties in the population that would have benefited from intervention. Hence, the present study does not offer conclusive evidence as to the extent to which reduced need or barriers to access contribute to the disparities observed. Neither does the database give any indication of the quality of the service received. As discussed above, proper diagnosis and treatment may be hampered by language and cultural barriers. While the rate of utilization is one measure of disparity, the effectiveness of the intervention and the satisfaction of the patients with the experience are equally important dimensions of equity and access in health services. Another limitation of the databases used in this study is that they contain only records of utilization of formal medical care covered by the province's health plan. Many alternative therapies are not included and it is not known the extent to which the immigrants rely on these other forms of intervention. Nevertheless, since costs of these 48 Chapter 2 alternative treatments would have to be borne by the immigrants themselves, it is not very likely that the volume of these treatments would have compensated for the disparity in use of formal care observed in this study. - Further investigative efforts must continue to clarify the patterns and determinants of mental health service utilization. Such efforts will elucidate the existence of gaps and barriers in services. The health service system can then develop appropriate strategies to address the mental health needs of a rapidly growing minority immigrant population. CONCLUSION This study presents strong population-based evidence that Chinese immigrants in B C use all inpatient and outpatient mental health services at a much lower rate than does the general population, although the magnitude of the disparity varies by sex and age: The extent to which the differences are explained by lower prevalence of mental disorders still needs to be established. Furthermore, immigrants are less likely to use general practice for substance use-related problems and less likely to use specialist services for less serious psychiatric conditions. Future research needs to clarify the reasons for these differences and identify any gaps and barriers in mental health services. \ \ 49 Chapter 2 Table 2.1 Description of Immigrant and Comparison Subjects Sex Female Male 51.05% 48.95% Age Group at Entry to Study 0-14 15-24 25-44 45-64 65+ 18.52% 13.55% 43.39% 18.37% 6.17% Landing Year of Immigrant 1985-1988 1989-1991 1992-1994 1995-1997 1998-2000 5.28% 12.7% 29.86% 34.74% 17.42% Mean Length of Time in Study (Years) Immigrant Comparison 5.23 5.41 Total Number = 148,973 pairs 50 Chapter 2 Table 2.2a Rate and Rate Ratio of Mental Health Visits to General Practitioners and Psychiatrists by Chinese Immigrants and Comparison Subjects in 1992-2001 FEMALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group Comparison) Under 15 13188 5.33 (4.60,6.52) 19.35 (17.81 ,21.12) 0.28 (0.24,0.36) 15 to 24 10449 12.84 (11.61 .14.48) 66.72 (61.46,74.50) 0.19 (0.17,0.22) 25 to 44 34931 20.94 (20.11,21.97) 113.13 (108.37,119.31) 0.19 (0.17,0.20) 45 to 64 12760 21.45 (19.96,23.58) 89.46 (83.08,96.88) 0.24 (0.22,0.27) 65 and over 4730 21.96 (19.49,24.78) 71.59 (65.10,78.99) 0.31 (0.26,0.36) ALL FEMALE Crude 76058 17.28 (16.71,17.91) 83.28 (80.49,86.20) 0.21 (0.20,0.22) Age-standard ized 2 78.46 (76.08,81.35) 0.22 (0.21 ,0.23) MALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group Comparison) Under 15 14405 5.41 (4.80,6.25) 22.57 (20.85,24.70) 0.24 (0.21 ,0.28) 15 to 24 9733 12.00 (9.83,15.07) 44.60 (39.38,53.33) 0.27 (0.21 ,0.35) 25 to 44 29709 10.44 (9.72,11.29) 86.57 (80.42,93.76) 0.12 (0.11 ,0.13) 45 to 64 14611 10.14 (9.42,10.92) 65.70 (60.82,71.81) 0.15 (0.14,0.17) 65 and over 4457 14.52 (13.00,17.27) 51.49 (46.31 ,58.51) 0.28 (0.24,0.36) ALL MALE Crude 72915 9.82 (9.33,10.42) 61.43 (58.95,64.86) 0.16 (0.15,0.17) Age-standardized2 59.27 (56.75,62.28) 0.17 (0.15,0.18) TOTAL POPULATION Crude 148973 13.65 (13.23,14.06) 72.67 (70.83,74.86) 0.19 (0.18,0.20) Age-standardized2 69.21 (67.25,71.01) 0.20 (0.19,0.20) 1 Rate is number of visits per 100 person years 2 Rate of the comparison group is age-standardized to the 1996 census population in BC 51 Chapter 2 Table 2.2b > Rate and Rate Ratio of Mental Health Visits to General Practitioners by Chinese Immigrants and Comparison Subjects in 1992-2001 1 Rate is number of visits per 100 person years 2 Rate of the comparison group is age-standardized to the 1996 census population in BC • 52 FEMALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group Comparison} Under 15 13188 3.08 (2.76,3.80) 10.46 (9.76,11.35) 0.29 (0.26,0.36) 15 to 24 10449 9.29 (8.71,9.89) 48.02 (43.99,53.42) 0.19 (0.17,0.22) 25 to 44 \ 34931 16.55 (16.06,17.04) 77.01 (73.57,81.59) 0.21 (0.20,0.23) 45 to 64 12760 17.14 (16.31,17.95) 58.79 (54.80,65.78) 0.29 (0.26,0.32) 65'and over 4730 17.96 (16.31 ,20.27) 51.31 (47.83,55.30) 0.35 (0.31 ,0.40) ALL FEMALE Crude 76058 13.41 (13.08,13.74) 56.34 (54.33,58.74) 0.24 (0.23,0.25) Age-standardized2 53.14 (51.34,55.26) 0.25 (0.24,0.26) MALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group ComDarison} Under 15 14405 2.81 (2.52,3.21) 8.86 (8.00,10.01) 0.32 (0.27,0.37) 15 to 24 9733 7.00 (5.95,9.14) 29.99 (25.58,38.07) 0.23 (0.18,0.32) 25 to 44 29709 8.01 (7.45,8.59) 62.86 (57.83,69.90) 0.13 (0.11 ,0.14) 45 to 64 14611 8.51 (8.02,9.01) 45.57 (41.32,51.56) 0.19 (0.16.0.21) 65 and over 4457 11.93 (10.79,13.95) 34.54 (31.38,41.23) 0.35 (0.29,0.41) ALL MALE Crude ' 72915 7.19 (6.87,7.54) 42.12 (39.65,44.89) 0.17 (0.16,0.18) Age-standardized2 40.38 (38.24,43.06) 0.18 (0.16,0.19) TOTAL POPULATION Crude 148973 10.38 (10.15,10.62) 49.44 (47.92,51.18) 0.21 (0.20,0.22) Age-standardized2 46.95 (45.49,48.50) 0.22 (0.21 ,0.23) Chapter 2 Table 2.2c Rate and Rate Ratio of Visits to Psychiatrists by Chinese Immigrants and Comparison Subjects in 1992-2001 FEMALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group Comparison) Under 15 13188 2.25 (1.73,3.08) 8.89 (7.72,10.35) .0.25 (0.19,0.35) 15 to 24 10449 3.56 (2.75,4.99) 18.70 . (15.79,21.87) 0.19 (0.14,0.27) 25 to 44 34931 4.39 (3.79,5.12) 36.12 (33.30 38.97) 0.12 (0.10,0.15) 45 to 64 12760 4.31 (3.23,6.12) 30.67 (27.54,35.42) 0.14 (0.10,0.20) 65 and over 4730 4.01 (3.04,5.43) 20.28 (16.51 ,25.46) 0.20 (0.14,0.29) ALL FEMALE Crude 76058 ' 3.86 (3.43,4.35) 26.94 (25.26,28.39) 0.14 (0.13,0.16) Age-standardized2 25.32 (23.88 26.80) 0.15 (0.14,0.18) MALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group Comparison) Under 15 14405 2.60 (2.12,3.16) 13.71 (12.27,15.35) 0.19 (0.15,0.24) 15 to 24 9733 5.00 (3.58,7.21) 14.61 (12.28,17.28) 0^34 • (0.23,0.51) 25 to 44 29709 2.43 (2.00,2.98) 23.72 (21.80,25.77) 0.10 (0.08,0.13) 45 to 64 14611 1.63 (1.19,2.17) 20.13 (17.84,22.75) 0.08 (0.06,0.11) 65 and over 4457 2.59 (1.67,4.55) 16.94 (13.58,21.85) 0.15 (0.10,0.27) ALL MALE Crude 72915 2.63 (2.30,3.02) 19.31 (18.24,20.49) 0.14 (0.12,0.16) Age-standardized2 18.89 (17.96,20.03) 0.14 (0.12,0.16) TOTAL POPULATION Crude 148973 3.26 (2.99,3.59) 23.24 (22.33,24.22) 0.14 (0.13,0.15) Age-standardized2 22.26 (21.37,23.32) 0.15 (0.13,0.16) 1 Rate is number of visits per 100 person years 2 Rate of the comparison group is age-standardized to the 1996 census population in BC 53 Chapter 2 Table 2.3 Rate and Rate Ratio of Mental Health Hospitalizations by Chinese Immigrants and Comparison Subjects in 1992-2001 FEMALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group ComDarisonV Under 15 13188 0.09 (0.06.0.14) 0.37 (0.31,0.47) 0.24 (0.15 0.38) 15 to 24 10449 0.20 (0.14,0.26) 0.91 (0.73,1.16) 0.22 (0.14,0.30) 25 to 44 34931 0.13 (0.10,0.15) 0.79 (0.70,0.91) 0.16 (0.12,0.21) 45 to 64 12760 0.11 (0.07,0.20) 0.68 (0.57,0.84) 0.16 (0.10,0.29) 65 and over 4730 0.12 (0.08,0.18) 1.45 (1.11 ,1.97) 0.08 (0.05,0.13) ALL FEMALE Crude 76058 0.13 (0.11,0.15) 0.76 (0.70,0.83) 0.17 (0.14,0.20) Age-standardized2 0.80 (0.73,0.89) 0.16 (0.13,0.19) MALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% C| flmmiarant vs 95% C| Age Group - • Comparison) Under 15 14405 0.04 (0.02,0.07) 0.28 (0.22,0.37) 0.15 (0.08,0.27) 15 to 24 9733 0.11 (0.08,0.17) 0.77 (0.62,1.04) 0.15 (0.09,0.24) 25 to 44 29709 0.08 (0.06,0.12) 0.79 (0.69,0.91) . 0.11 (0.08,0.15) 45 to 64 14611 0.02 (0.01 ,0.04) 0.57 (0.46,0.74) 0.04 (0.02,0.07) 65 and over 4457 0.15 (0.09,0.23) 1.46 (0.85,2.77) 0.10 (0.04,0.21) ALL MALE Crude 72915 0.07 (0.06,0.09) 0.68 (0.62,0.77) 0.10 (0.08,0.13) Age-standardized2 0.71 (0.62,0.85) 0.10 (0.07,0.12) TOTAL POPULATION Crude 148973 0.10 (0.09,0.11) 0.72 (0.67 0.77) 0.14 (0.12,0.16) Age-standardized2 0.75 (0.70,0.84) 0.13 (0.11 ,0.15) 1 Rate is number of hospital discharges per 100 person years 2 Rate of the comparison group is age-standardized to the 1996 census population in BC 54 Chapter 2 Table 2.4 Rate and Rate Ratio of Time In Outpatient Mental Health Care by Chinese Immigrants and Comparison Subjects in 1992-2001 FEMALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group Comparison) Under 15 13188 0.07 (0.05,0.10) 0.48 (0.41,0.56) 0.15 (0.10,0.22) 15 to 24 10449 0.08 (0.05,0.14) 0.70 (0.57,0.84) 0.12 (0.07,0.21) 25 to 44 34931 0.16 (0.12,0.19) 0.88 (0.79,0.98) 0.18 (0.13,0.22) 45 to 64 12760 0.08 (0.05,0.14) 1.09 (0.93,1.26) 0.07 (0.04,0.14) 65 and over 4730 0.06 (0.01,0.21) 0.83 (0.65,1.12) 0.07 (0.01,0.23) ALL FEMALE Crude 76058 0.11 (0.09,0.13) 0.82 (0.77 0.88) 0.13 (0.11 ,0.17) Age-standardized2 0.82 (0.76,0.88) 0.13 (0.11 ,0.17) MALE N (pairs) Immigrant Comparison Rate Ratio Rate1 95% CI Rate1 95% CI (Immiarant vs 95% CI Age Group „ ComDarison) Under 15 14405 0.09 (0.06,0.12) 0.66 (0.57,0.75) 0.14 (0.10,0.19) 15 to 24 9733 0.12 (0.08,0.21) 0.53. (0.43,0.67) 0.23 (0.14,0.40) 25 to 44 29709 0.10 (0.07,0.14) 0.82 (0.72,0.92) 0.12 (0.08,0.17) 45 to 64 14611 0.03 (0.01 ,0.08) 0.58 (0.47,0^69) 0.05 (0.02,0.14) 65 and over 4457 0.03 (0.00,0.09) 0.55 (0.40,0.74) 0.05 (0.01 ,0.17) ALL MALE Crude 72915 0.08 (0.06,0.10) 0.68 (0.63,0.74) 0.12 (0.09,0.15) Age-standardized2 0.66 (0.61,0.71) 0.12 (0.09,0.15) TOTAL POPULATION Crude 148973 0.10 (0.08,0.11) 0.75 (0.71 ,0.79) 0.13 (0.11,0.15) Age-standardized2 0.74 (0.70,0.78) 0.13 (0.11,0.15) 1 Rate is number of years in care per 100 person years 2 Rate of the comparison group is age-standardized to the 1996 census population in BC 55 Table 2.5a y Diagnostic Categories of Mental Health Visits to General Practitioners and Psychiatrists by Chinese Immigrants and Comparison Subjects in 1992-2001 ej T3 Immigrant Comparison |-t #of # of Visits % 95% CI #of # of Visits % 95% CI Code Diagnostic Category Subiects1 Subiects1 50B2 Anxiety/Depression 17471 37782 35.16% (34.16%, 36.17%) 28481 99318 16.48% (15.95% 17.02%) 300 Neurotic Disorders 8668 21333 19.85% (18.96%, 20.75%) 19394 91981 15.26% (14.61% 15.92%) 311 Depressive Disorder NEC 3867 12334 11.48% (10.72%, 12.24%) 20346 128708 21.36% (20.64% 22.08%) 308 Acute Reaction to Stress 4092 7429 6.91% (6.57%, 7.26%) 11866 29126 4.83% (4.63% 5.04%) 296 Affective Psychoses 742 6554 6.10% (5.40%, 6.80%) 2917 30455 5.05% (4.65% 5.46%) 295 Schizophrenic Psychoses 396 4936 4.59% (3.76%, 5.43%) 1342 20002 3.32% (2.97% 3.66%) 307 Special Symptoms or Syndromes NEC 2484 4085 3.80% (3.54%, 4.06%) 3932 10858 1.80% (1.57% 2.03%) 309 Adjustment Reaction 1541 3358 3.13% (2.82%, 3.43%) 7344 30954 5.14% (4.76% 5.52%) 304 Drug Dependence 136 1648 1.53% (0.70%, 2.36%) 2723 103921 17.25% (15.55% 18.95%) 298 Other Nonorganic Psychoses 241 1439 1.34% (0.94%, 1.74%) 674 2354 0.39% (0.33% 0.45%) 314 Hyperkinetic Syndrome of Childhood 166 905 0.84% (0.61%, 1.07%) 950 5373 0.89% (0.76% 1.02%) 313 Disturbance of Emotions - Childhood and Adolescence 165 726 0.68% (0.48%, 0.87%) 710 4518 0.75% (0.62% 0.88%) 301 Personality Disorders 335 653 0.61% (0.51%, 0.71%) 1851 9974 1.66% (1.44% 1.87%) 290 Senile & Presenile Organic Psychotic Conditions 135 640 0.60% (0.34%, 0.85%) 594 3339 0.55% (0.45% 0.65%) 302 Sexual Deviations & Disorders 390 614 0.57% (0.48%, 0.66%) 939 2244 0.37% (0.31% 0.44%) 293 Transient Organic Psychotic Conditions 91 517 0.48% (0.30%, 0.67%) 392 1991 0.33% (0.28% 0.38%) 306 Physiological Malfunction Arising From Mental Factors 270 504 0.47% (0.37%, 0.57%) 749 2337 0.39% (0.30% 0.48%) 297 Paranoid States 127 471 0.44% (0.29%, 0.59%) 358 1318 0.22% (0.16% 0.28%) Non-Psychiatric Diagnoses3 78 410 0.38% (0.17%, 0.59%) 747 4032 0.67% (0.53% 0.81%) 312 Disturbance of Conduct NEC 213 309 0.29% (0.23%, 0.34%) 1375 2866 0.48% (0.43% 0.52%) 310 Nonpsychotic Mental Disorders - Organic Brain Damage 78 219 0.20% (0.12%, 0.28%) 560 1932 0.32% (0.26% 0.38%) 303 Alcohol Dependence Syndrome 108 156 0.15% (0.11%, 0.18%) 1976 7429 1.23% (1.11% 1.35%) 294 Other Organic Psychotic Conditions (Chronic) 42 121 0.11% (0.06%, 0.17%) 306 1311 0.22% (0.16% 0.27%) 305 Nondependent Abuse of Drugs 75 84 0.08% (0.06%, 0.10%) 660 5008 0.83% (0.46% 1.20%) 292 Drug Psychoses 36 79 0.07% (0.01%,0.14%) 153 290 0.05% (0.04% 0.06%) 04A2 General Psychiatric Examination 63 70 0.07% (0.05%, 0.08%) 40 43 0.01% (0.00% 0.01%) 299 Psychoses With Origin Specific to Childhood 22 61 0.06% (0.02%, 0.09%) 151 626 0.10% (0.07% 0.13%) 291 Alcoholic Psychoses 9 9 0.01% (0.00%, 0.01%) 127 291 0.05% (0.03% 0.06%) TOTAL 4 30756 107446 100% 60171 602599 100% 11ndividual subjects may be treated for more than one diagnosis. The total number of subjects does not equal the sum of subjects for the diagnostic categories. 2 MSP diagnostic codes; all the others are ICD-9 codes 3 Includes developmental delays (315), psychic factors associated with other diseases (316), mental retardation (317-318), other conditions of brain and the nervous system (348-349), general symptoms (780), other family circumstances (V61) a\ 4 16 missing diagnoses among comparison subjects were not included in the total count Table 2.5b Diagnostic Categories of Mental Health Visits to General Practitioners by Chinese Immigrants and Comparison Subjects in 1992-2001 Immigrant Comparison Code Diagnostic Category #of # of Visits % 95% CI #of # of Visits % 95% CI Subjects1 Subjects1 50B2 Anxiety/Depression 17452 37636 46.02% (45.11%, 46.94%) 28165 90398 22.03% (21.26%, 22.81%) 300 Neurotic Disorders 8189 16111 19.70% (19.09%, 20.31%) 17350 47534 11.59% (11.14% ,12.03%) 311 Depressive Disorder NEC 3648 8655 10.58% (10.09%, 11.08%) 18992 83219 20.28% (19.54%, 21.02%) 308 Acute Reaction to Stress 4073 7300 8.93% (8.53%, 9.32%) 11647 26971 6.57% (6.28%, 6.86%) 307 Special Symptoms or Syndromes NEC 2439 3747 4.58% (4.33%, 4.83%) 3575 6913 1.68% (1.56%, 1.81%) 309 Adjustment Reaction 1147 1747 2.14% (1.97%, 2.30%) 5396 13032 3.18% (2.87%, 3.49%) 304 Drug Dependence 131 1425 1.74% (0.81%, 2.67%) 2680 102659 25.02% (22.80%, 27.25%) 295 Schizophrenic Psychoses 285 1217 1.49% (1.18%, 1.79%) 910 6241 1.52% (1.29%, 1.75%) 302 Sexual Deviations & Disorders 380 564 0.69% (0.60%, 0.78%) 866 1533 0.37% (0.32%, 0.43%) 301 Personality Disorders 312 528 0.65% (0.55%, 0.74%) 1346 4274 1.04% (0.91%,1.17%) 306 Physiological Malfunction Arising From Mental Factors 266 488 0.60% (0.46%, 0.73%) 705 1432 0.35% (0.30%, 0.40%) 290 Senile & Presenile Organic Psychotic Conditions 108 458 0.56% (0.25%, 0.87%) 405 1642 0.40% (0.31%, 0.49%) 298 Other Nonorganic Psychoses 141 330 0.40% (0.30%, 0.50%) 511 1187 0.29% (0.24%, 0.34%) 312 Disturbance of Conduct NEC 207 281 0.34% (0.29%, 0.40%) 1293 2391 0.58% (0.53%, 0.64%) 296 Affective Psychoses 96 218 0.27% (0.18%, 0.35%) 1003 3798 0.93% (0.81%, 1.04%) 297 Paranoid States 82 189 0.23% (0.14%, 0.32%) 273 743 0.18% (0.11%, 0.25%) 314 Hyperkinetic Syndrome of Childhood 121 164 0.20% (0.16%, 0.24%) 639 1780 0.43% (0.38%, 0.49%) 303 Alcohol Dependence Syndrome 107 155 0.19% (0.14%, 0.24%) 1918 6893 1.68% (1.53%, 1.83%) 313 Disturbance of Emotions - Childhood and Adolescence 70 134 0.16% (0.09%, 0.24%) 236 366 0.09% (0.07%, 0.11%) 310 Nonpsychotic Mental Disorders - Organic Brain Damage 66 120 0.15% (0.10%, 0.19%) 463 1191 0.29% (0.23%, 0.35%) 305 Nondependent Abuse of Drugs 74 83 0.10% (0.08%, 0.13%) 600 4674 1.14% (0.59% ,1.68%) 04A2 General Psychiatric Examination 63 70 0.09% (0.06%, 0.11%) 40 43 0.01% (0.01%, 0.01%) 294 Other Organic Psychotic Conditions (Chronic) 28 56 0.07% (0.02%, 0.12%) 209 698 0.17% (0.11%, 0.23%) 292 Drug Psychoses 31 41 0.05% (0.02%, 0.08%) 107 164 0.04% (0.03%, 0.05%) 293 Transient Organic Psychotic Conditions 23 27 0.03% (0.02%, 0.05%) 83 140 0.03% (0.02%, 0.05%) 299 Psychoses With Origin Specific to Childhood 14 16 0.02% (0.01%, 0.03%) 63 97 0.02% (0.02%, 0.03%) 291 Alcoholic Psychoses 8 8 0.01% (0.00%, 0.02%) 97 192 0.05% (0.03%, 0.06%) Non-Psychiatric Diagnoses3 6 6 0.01% (0.00%, 0.01%) 86 90 0.02% (0.02%, 0.03%) TOTAL 30395 81774 100% 58508 410295 100% 1 Individual subjects may be treated for more than one diagnosis. The total number of subjects does not equal the sum of subjects for the diagnostic categories. 2 MSP diagnostic codes; all the others are ICD-9 codes 3 Includes diagnostic categories such as general symptoms (780), poisoning (965,967,977), toxic effects of alcohol and other substances (980,989) Table 2.5c Diagnostic Categories of Visits to Psychiatrists by Chinese Immigrants and Comparison Subjects in 1992-2001 Immigrant ' Comparison #of # of Visits % 95% CI #of # of Visits % 95% CI Code Diagnostic Category Subjects1 Subjects1 296 Affective Psychoses 690 6336 24.68% (21.99%, 27.38%) 2262 26657 13.86% (12.75% 14.97%) 300 Neurotic Disorders 704 5222 20.34% (17.19%,23.50%) 3430 44447 23.11% (21.47% 24.75%) 295 Schizophrenic Psychoses 217 3719 14.49% (11.72%, 17.25%) 837 13761 7.16% (6.32% 7.99%) 311 Depressive Disorder NEC 404 3679 14.33% (11.69%, 16.97%) 3756 45489 23.65% (22.14% 25.17%) 309 Adjustment Reaction 425 1611 6.28% (5.11%, 7.44%) 2199 17922 9.32% (8.36% 10.28%) 298 Other Nonorganic Psychoses 114 1109 4.32% (2.70%, 5.94%) 200 1167 0.61% (0.46% 0.76%) 314 Hyperkinetic Syndrome of Childhood 72 741 2.89% (1.95%, 3.82%) 402 3593 1.87% (1.49% 2.24%) 313 Disturbance of Emotions - Childhood and Adolescence 97 592 2.31% (1.52%, 3.09%) 487 4152 2.16% (1.75% 2.56%) 293 Transient Organic Psychotic Conditions 70 490 1.91% (1.14%,2.68%) 313 1851 0.96% (0.80% 1.13%) Non-Psychiatric Diagnoses3 72 404 1.57% (0.71%, 2.43%) 664 3942 2.05% (1.62% 2.48%) 307 Special Symptoms or Syndromes NEC 50 338 1.32% (0.63%, 2.01%) 409 3945 2.05% (1!41% 2.69%) 297 Paranoid States 50 282 1.10% (0.55%, 1.65%) 105 575 0.30% (0.19% 0.41%) 304 Drug Dependence 9 223 0.87% (0.00%, 1.75%) 139 1262 0.66% (0.41% 0.91%) 290 Senile & Presenile Organic Psychotic Conditions 33 182 0.71% (0.32%, 1.10%) 252 1697 0.88% ' (0.65% 1.12%) 50B2 Anxiety/Depression 29 146 0.57% (0.30%, 0.83%) 682 8920 4.64% (3.71% 5.56%) 308 Acute Reaction to Stress 24 129 0.50% (0.15%, 0.85%) 282 2155 1.12% (0.84% 1.40%) 301 Personality Disorders 28 125 0.49% (0.22%, 0.75%) 589 5700 2.96% (2.36% 3.57%) 310 Nonpsychotic Mental Disorders - Organic Brain Damage 13 99 0.39% (0.09%, 0.68%) 100 741 0.39% (0.25% 0.52%) 294 Other Organic Psychotic Conditions (Chronic) 14 65 0.25% (0.07%, 0.44%) 102 613 0.32% (0.19% 0.44%) 302 Sexual Deviations & Disorders 10 50 0.19% (0.00%, 0.41%) 87 711 0.37% (0.22% 0.52%) 299 Psychoses With Origin Specific to Childhood 9 45 0.18% (0.03%, 0.33%) 92 529 0.28% (0.18% 0.37%) 292 Drug Psychoses 5 38 0.15% (0.00%, 0.39%) 50 126 0.07% (0.04% 0.09%) 312 Disturbance of Conduct NEC 6 28 0.11% (0.00%, 0.24%) 103 475 0.25% (0.15% 0.34%) 306 Physiological Malfunction Arising From Mental Factors 4 16 0.06% (0.00%, 0.14%) 44 905 0.47% (0.21% 0.73%) 291 Alcoholic Psychoses 1 1 0.00% (0.00%, 0.01%) 30 99 0.05% (0.02% 0.08%). 303 Alcohol Dependence Syndrome 1 1 0.00% (0.00%, 0.01%) 130 536 0.28% (0.09% ,0.47%) 305 Nondependent Abuse of Drugs 1 1 0.00% (0.00%, 0.01%) 67 334 0.17% (0.10% .0.25%) TOTAL* 2266 25672 100% 11388 192304 100% 11ndividual subjects may be treated for more than one diagnosis. The total number of subjects does not equal the sum of subjects for the diagnostic categories. 2 MSP diagnostic codes; all the others are ICD-9 codes 3 Includes developmental delays (315), psychic factors associated with other diseases (316), mental retardation (317-318), other conditions of brain and the nervous system (348-349), general symptoms (780), other family circumstances (V61) . 4 1 6 missing diagnoses among comparison subjects were not included in the total count Chapter 2 BIBLIOGRAPHY 1. Chen J, Ng E, Wilkins R. The health of Canada's immigrants in 1994-95. Health Rep 1996; 7: 33-45. 2. Kliewer E, Kazanjian A . The Health Status and Medical Services Utilization of Recent Immigrants to Manitoba and British Columbia: A Pilot Study. Vancouver: B C Office of Health Technology Assessment, Centre for Health Services and Policy Research, University of British Columbia: 2000. 3. Leclere FB, Jensen L , Biddlecom A E . Health care utilization, family context, and adaptation among immigrants to the United States. J Health Soc Behav 1994; 35: 370-384. 4. Newbold B. Health status and health care of immigrants in Canada: a longitudinal analysis. J Health Serv Res Policy 2005; 10: 77-83a. 5. Kirmayer LJ , Galbaud du Fort G, Young A, Weinfeld M , Lasry J-C. Pathways and Barriers to Mental Health Care in an Urban Multicultural Milieu: An Epidemiological and Ethnographic Study. Culture & Mental Health Research Unit Report 6 (Part 1). Montreal: Culture & Mental Health Research Unit, Sir Mortimer B. Davis - Jewish General Hospital: 1996. 6. Cheung F K , Snowden LR. Community mental health and ethnic minority populations. Community Ment Health J 1990; 26: 277-291. 7. Leong FTL. Asian Americans' differential patterns of utilization of inpatient and outpatient public mental health servicesin Hawaii. J Community Psychol 1994; 22: 82-96. 8. Matsucka JK, Breaux C, Ryujin D H . National utilization of mental health services by Asian Americans/Pacific Islanders. J Community Psychol 1997; 25: 141-145. 9. Meinhardt K , Vega W. A method for estimating underutilization of mental health services by ethnic groups. Hosp Community Psychiatry 1987; 38: 1186-1190. 10. Snowden LR, Cheung F K . Use of inpatient mental health services by members of ethnic minority groups. Am Psychol 1990; 45: 347-355. 11. Klimidis S, McKenzie DP, Lewis J, Minas IH. Continuity of contact with psychiatric services: immigrant and Australian-born patients. Soc Psychiatry Psychiatr Epidemiol 2000;35:554-563. 12. Peters R. Increasing the responsiveness of mainstream mental health services to ethnocultural minorities. In: Health and Cultures Vol . II Programs, Services and Care. Edited by Masi R, Mensah L , McLeod K A . Oakville: Mosaic Press: 1993: 187-203: 59 Chapter 2 13. IntercuJtural Committee. Multiculturalism and Mental Health: Developing Culturally Competent Systems of Care. Vancouver: Greater Vancouver Mental Health Service: 1999. 14. Roberts N , Crockford D. Psychiatric admissions of Asian Canadians to an adolescent inpatient unit. Can J Psychiatry 1997; 42: 847-851. 15. Lai D W L , Tsang K T , Chappell N L , Lai D C Y , Chau S B Y . Health and Well Being of Older Chinese in Canada. Calgary: Faculty of Social Work, University of Calgary: 2003. 16. Kazanjian A , Reid RJ, Pagliccia N , Apland L , Wood L . Issues in Physician Resources Planning in B.C.: Key Determinants of Supply and Distribution, 1991-96. H H R U 00:02. Vancouver: Centre for Health Services and Policy Research, University of British Columbia: 2000. 17. Newcombe H B . Handbook of Record Linkage: Methods for Health and Statistical Studies, Administration, and Business. New York: Oxford Medical Publications: 1988. 18. Hesterberg T, Moore DS, Monaghan S, Clipson A , Epstein R. Bootstrap methods and permutation tests. In: Introduction to the Practice of Statistics. Edited by Moore DS, McCabe GP. New York: W. H . Freeman: 2005: 14-1-14-70. 19. Chen C-N, Wong J, Lee N , Chan-Ho M-W, Lau JT-F, Fung M . The Shatin Community Mental Health Survey in Hong Kong. Arch Gen Psychiatry 1993; 50: 125-133. 20. Hwu H-G, Yeh E-K, Chang L - Y . Prevalence of psychiatric disorders in Taiwan defined by the Chinese Diagnostic Interview Schedule. Acta Psychiatr Scand 1989; 79: 136-147. 21. WHO World Mental Health Survey Consortium. Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. J A M A 2004; 291:. 2581-2590. 22. Takeuchi DT, Chung R C - Y , Lin K - M et al. Lifetime and twelve-month prevalence rates of major depressive episodes and dysthymia among Chinese Americans in Los Angeles. A m J Psychiatry 1998; 155: 1407-1414. 23. Review of the Literature on Migrant Mental Health. Canadian Task Force on Mental Health Issues Affecting Immigrants and Refugees: 1988. 24. Bowen S. Access to Health Services for Underserved Populations in Canada. "Certain Circumstances": Issues in Equity and Responsiveness in Access to Health Care in Canada. Ottawa, Ontario: Health Canada: 2000. 60 CHAPTER 3 DETERMINANTS OF MENTAL H E A L T H CONSULTATIONS BY RECENT CHINESE IMMIGRANTS IN BRITISH COLUMBIA 1 The province of British Columbia (BC), located on the west coast of Canada, has always been a popular destination for immigrants from Asia. Since changes in Canada's immigration policies in 1967, when national origin was eliminated as a selection criterion, Asians gradually comprised the majority of immigrants arriving in this province, and Chinese immigrants constituted a growing proportion of residents in BC. In 2001, 26% of the 3.9 million residents were immigrants; 24% of these immigrants were born in a Chinese territory and 60% of them came within the previous decade.2 In that same year, the People's Republic of China, Taiwan and Hong Kong contributed 35% of the 34,000 newcomers to Vancouver.3 Such ethnocultural diversity presents challenges to the provision of health care, especially in mental health. Several studies have documented under-utilization of mental health services by Asian and Chinese minorities in the United States, Australia and Canada relative to other groups. Asians, including immigrants, were less likely to use inpatient and outpatient mental health services (1-6). Chinese in Vancouver were under-represented in the community mental health system and older Chinese men and women in seven Canadian cities, most of whom were immigrants, were less likely than the general population to see psychiatrists and psychologists (7-9). Disparities in utilization raise the concern that there may be inequities in access to mental health services, especially in relation to recent immigrants. A variety of factors, including the lack of English skills, different health beliefs and the lack of knowledge about the health system, have been suggested as barriers that hinder immigrants from receiving needed mental health interventions. Little is known about the factors associated with mental health service utilization among Chinese immigrants in the Canadian context. According to published literature, 1 A version of this chapter will be submitted for publication in a journal. 2 Statistics Canada Census 2001 Cat. No. 97F0009XCB2001002 http://wwwl2.statcan.ca/english/census01/Pro^ 3 Citizenship and Immigration Canada Facts and Figures 2001 http://www.cic.gc.ca/english/pub/facts200 l/4van-02.html 61 Chapter 3 some factors associated with mental health service utilization among minority groups are common to the general population, including sex, age, marital status, psychiatric disorder, disability or impairment, insurance coverage, region of residence, availability of service and education (10-15). Other factors are more salient to immigrants and ethnic minorities, such as place of birth, acculturation, English language ability and perceived discrimination (2, 16-18). However, most studies are conducted outside Canada where the health care system and immigration patterns differ substantially. In light of the size of the Chinese immigrant population in this province, this study aims to fill in some of the gaps in information about the determinants of this minority group's use of mental health services. The specific objective is to examine the individual and contextual factors associated with mental health visits to general practitioners and psychiatrists by recent Chinese immigrants in BC. METHODS Study Design This study adopted a retrospective observational cohort design. Study immigrants were identified from a national immigration database and comparison group members from the provincial health plan registry. Their utilization data for the study period of 1992-2001 were extracted from provincial health administrative databases. Data Sources Data for this study came from three administrative data systems: Landed Immigrant Data System (LIDS) The database provided by Citizenship and Immigration Canada (CIC) consists of individual information on all immigrants who landed in B C between 1985 and 2000. Data available to this study included basic demographic information, landing date, place of origin, educational level, marital status and ability to communicate in English or French. BC Linked Health Data (BCLHD) The B C L H D is a linked database of most administrative health records in the province. Databases used in this study were the provincial health plan registry and the 62 Chapter 3 physician payments records. Since all B C residents are covered by the provincial health plan after three months of residency and most physicians are reimbursed on a fee for service basis, the database contains records of most physician services received by B C residents. College of Physicians and Surgeons of B C (CPSBC) Register The register of physicians in B C from 1985-2001 was used to determine the supply of general practitioners and psychiatrists in each health jurisdiction in B C and to identify Chinese-speaking physicians. Data available from this register included name, address, specialties, place of graduation and self-reported language skills. As part of a national immigrant health research project, the LIDS and B C L H D were linked by an independent health services research centre, using probabilistic linkage methods (19). The list of Chinese-speaking physicians was generated based on the surname, place of graduation and self-reported language skills in the CPSBC register, supplemented by a telephone verification procedure. The list was then linked to the physician payments files deterministically, using each physician's unique billing number, to identify the services provided by Chinese-speaking physicians. Subjects Study immigrants were those who landed in B C between 1985 and 2000 from China, Taiwan, Hong Kong or Macau and whose records were successfully linked to the province's health database. Each linked immigrant was matched by sex, age and health service jurisdiction to a comparison subject randomly selected from the province's health plan registry in the year of landing and who was not in the immigration database. Only subjects who were registered in the health plan for at least one day during the study period of 1992-2001 were included in the analyses. Hence, the final study population consisted of 152,184 immigrants and 151,303 comparison subjects. Study Variables Outcomes Two types of outcomes were examined in this study: mental health visits to general practitioners and visits to psychiatrists. A mental health visit to a general 63 Chapter 3 practitioner was defined as all the services for which a general practitioner submitted payment claims for a patient in one day where the diagnosis or the service was related to mental health. A visit to a psychiatrist included all services claimed by a psychiatrist for a patient in one day. Determinants The determinants of mental health service utilization for immigrants in this study were age, years since landing, rate of non-mental health visits to general practitioners, mean socio-economic quintile of residence, place of origin, last health service delivery area (HSDA) of residence, educational level, marital status and English skill. Age was the age of the immigrant during the study period when they became eligible for the study, i.e., when they had landed and were registered in the health plan. Years since landing was the interval from landing date to the mid-point of their eligible registration period, , representing the "mean" length of time in Canada of the immigrant while participating in the study. The rate of non-mental health visits was calculated individually by dividing the total number of visits to general practice of each immigrant, minus the mental health visits, by the total registration time. Place of origin was the place of last permanent residence, either China, Taiwan, Hong Kong of Macau. Health service delivery areas are official jurisdictions in B C for planning and delivering services and an individual's H S D A was assigned according to the last address documented in the health plan registry. Only HSDAs where there were sizeable populations of Chinese immigrants were individually identified in the analyses; areas with small number of immigrants were aggregated in the "Other" category, which also included 0.4% of immigrants with missing address information. With the exception of this category and South Vancouver Island, all of the health areas were located within the greater Vancouver area. The socio-economic quintile is an indicator of the socio-economic status of a'community assessed according to the income of the residents reported in the census. The socio-economic quintile has a value of 1 to 5, with 1 representing the lowest socio-economic status and 5 the highest. As the quintile for a neighbourhood might change over censuses and immigrants might move from year to year, the mean of all the documented socio-economic quintiles in the health plan registry during the study period was used as a continuous variable to reflect the average socio-economic status of each immigrant's 64 , . . . . . . . . . / Chapter 3 environment during this time. Socio-economic information was missing for about 0.5% of the immigrants. Information on educational level, marital status and English skill was available only at the time of landing. Education was categorized into four levels: 0 to 9 years of schooling, 10 to 12 years of schooling, post-secondary education but no university degrees and at least one university degree. Marital status consisted of three categories: single, married and all others, including the separated, divorced and widowed. English skill was a self-reported dichotomous variable indicating whether the immigrant was able to communicate in English or not. For the comparison group, the only determinants studied were age, mean socio-economic quintile of residence, rate of non-mental health visits to general practitioners and H S D A because the other variables were either not relevant or the information was not available for this group. Physician Variables Another group of variables was generated to examine the pattern of physician utilization by immigrants but was not entered into the regression models. These involved the language ability of the physicians and the supply of physicians. Services provided by Chinese-speaking general practitioners and psychiatrists were identified from the physician payments files and the percentage of mental health visits made by Chinese immigrants to Chinese-speaking physicians was calculated to gain an understanding of the language competency of the physicians responsible for mental health care of these immigrants. In addition, the number of physicians per 100,000 population in 1998 was calculated for each H S D A and for the following four types of physicians as an indicator of supply: general practitioners, psychiatrists, Chinese-speaking general practitioners and Chinese-speaking psychiatrists. Data Analysis The association of each determinant to consultations with general practitioners or psychiatrists was estimated as a rate ratio using a generalized linear regression model assuming a negative binomial distribution for the outcome, offset by the participation time of each subject. The negative binomial distribution was chosen to take into account ( • ' ' - • • • " • ' ' " 65 Chapter 3 the overdispersion of the data caused by the large proportion of subjects with no mental health visits and the variation in individual rates (20). Bootstrapping of the estimated regression coefficients in the multivariate models was used to assess the sensitivity of the standard errors to the negative binomial model assumption. Since the bootstrap standard errors generally were similar to those from the original regression models, only the results from the original models were reported here. A l l analyses were conducted using SAS9.1. Analyses for immigrants and the comparison group were performed separately in order to understand the effects of the determinants in each group. Previous studies had shown consistently that women overall had higher rates of health care utilization than men; therefore, analyses were stratified by sex. Furthermore, since three of the determinants - educational level, marital status and English competency - were either not meaningful among young subjects or were more likely to change in value over the study period, separate analyses were also done by two age groups with those three determinants excluded from the models for immigrants younger than 25 at landing. The effect of each determinant on an outcome was examined in a univariate model and then in a multivariate model which contained all the determinants for that sex and age group. Additional models involving subgroups of variables were used to address specific questions of interest, for example, to identify the combination of variables that confounded the effects of place of origin and education. Eight separate sets of regression results are reported in this study for the immigrant population: the univariate model and the multivariate model of all the determinants for each of the two age groups (ages 0 to 24 at landing and ages 25 or older at landing), for males and females and for visits to general practitioners and visits to psychiatrists. Only the multivariate models are presented for the comparison group, which serve to highlight differences between immigrants and non-immigrants in the role each determinant plays in service utilization. The supply of physicians was not entered into the regression analyses since the unit on which this variable was calculated - health service delivery area - was already in the models. The relationship of this variable to the effect of health service delivery areas on mental health visits was manually reviewed by the author. The proportions of mental 66 Chapter 3 health visits by Chinese immigrants to Chinese-speaking general practitioners and psychiatrists were estimated by logistic regression with multiple visits by the same subject treated as a cluster. RESULTS Description of Subjects Tables 3.1a and 3. lb summarize the demographic characteristics of the immigrants and the comparison group in this study. Of the 152,184 immigrants, one-third were younger than 25 when they landed. Among the older immigrants, there were slightly more women than men. The average age of young female immigrants was 14 and young males 13; the average ages in the older age group were 43 and 44 respectively. More than 80% of the immigrants landed during the study period, in 1992 or after. About half of the immigrants came from Hong Kong and less than 1% from Macau; more than a third of the young immigrants came from Taiwan while the older immigrants were more likely than the younger ones to arrive from China. Vancouver was the most popular residence for this cohort of immigrants, followed by Fraser North and Richmond. Among older immigrants, men had slightly higher levels of education than women (31% vs 20% with university degrees) and were more likely to report speaking English (56% vs 46%). Women were more likely than men to be separated, divorced or widowed (9% vs 2%). The comparison group members were similar to immigrants in most demographic variables. Although they were initially matched to the immigrants by health jurisdiction, they were more likely than immigrants to reside in Fraser South and Other areas of B C and less likely to be in Richmond by the end of the study period. Immigrants' Determinants of Mental Health Consultations With General Practitioners Tables 3.2a and 3.2b present the rate ratios (RR) and the 95% confidence intervals from both the univariate and multivariate models of each determinant for each sex and age sub-group of immigrants. After adjusting for all the variables, the factors that had major contribution to the number of young immigrants' mental health consultations with general practitioners were 67 Chapter 3 age, years since landing, rate of non-mental health visits, place of origin and health service delivery area (HSDA). The effects of many determinants were reduced between the univariate and multivariate models. For young immigrants, mental health consultations with general practitioners increased substantially with age (RR= 1.08-1.10), number of years since landing (RR=1.07-1.13) and non-mental health visits (RR=1.19-1.20). Young females from Hong Kong had 20% more visits than those from Taiwan while young males from China had 90% more visits than young Taiwanese males. Both young males and females in Richmond were less likely than their counterparts in Vancouver to visit their general practitioners for mental health problems (RR=0.76-0.80) but those in South Vancouver Island and other parts of B C were more likely to do so (RR=1.32-2.90 and 1.55-1.92). For older immigrants, the determinants that had significant contribution in the multivariate models were years since landing, rate of non-mental health visits, place of origin, HSDA, education, marital status and English competency. The effects of some of these determinants were altered in direction and magnitude between the univariate and multivariate models. Both years since landing and rate of non-mental health visits to general practitioners were associated with the rate of mental health visits (RR=1.07-1.08 and 1.18-1.23). Women from China had slightly more visits than women from Taiwan while women from Macau had fewer visits; on the other hand, men from China and Hong Kong had 45% and 18% more visits than men from Taiwan while men from Macau were not significantly different from Taiwanese men. Both men and women in Richmond had lower rates of mental health visits than those in Vancouver (RR=0.71-0.74) while those in South Vancouver Island had higher rates (RR=1.33-1.47); men in North Shore and other regions of B C also had higher rates than men in Vancouver (RR=1.25 and 1.29). while women in North Shore had lower rate than women in Vancouver (RR=0.80). Both men and women with only 0-9 years of schooling had lower rates of mental health visits than those with university degrees (RR.^0.86-0.89). In terms of marital status, unmarried women (single and separated/divorced/widowed) were slightly less likely than married women to visit general practitioners for mental health problems (RR=0.92-0.94) while single men had more visits than married men (RR=1.22). The ability to communicate in 68 Chapter 3 English had a slightly positive effect on mental health consultations only among men (RR=1.10) but had no effect in women. Immigrants' Determinants of Consultations with Psychiatrists The rate ratios and confidence intervals of determinants of psychiatric visits are presented in Tables 3.3a and 3.3b. There were fewer significant findings owing to the less frequent occurrence of psychiatric visits and larger variance. Among young female immigrants, determinants that had significant impact on S • rate of psychiatric visits in the multivariate model were age, years since landing, rate of non-mental health visits and place of origin. The effects of most determinants were attenuated compared to the univariate models. Among young male immigrants, the only determinants that had statistically significant impact were age and rate of non-mental health visits. For both young female and male immigrants, age (RR=1.04-1.05) and rate of non-mental health visits to general practitioners (RR=1.09-1.13) were associated with increased visits to psychiatrists. Among young females, years since landing also had a positive effect on psychiatric visits (RR=1.15). Young females from Hong Kong made more visits than those from Taiwan (RR=1.64) while those from Macau made far fewer (RR=0.07). Young immigrants who lived in Fraser South also had a lower rate of psychiatric visits than those in Vancouver (RR=0.43-0.49). Among older female immigrants, factors that influenced rate of psychiatric visits in the multivariate model were years since landing, rate of non-mental health visits, place of origin, HSDA, marital status and English skill. The effect of education,,while statistically significant in the univariate model, was reversed and reduced to non-significance. Among older immigrant men, determinants that had a significant effect on psychiatric visits in the multivariate model were rate of non-mental health visits, place of origin, H S D A and marital status. Among older immigrants, rate of non-mental health visits to general practitioners was associated with increased psychiatric visits in both men and women (RR=1.15-1.26) while years since landing had a positive effect only in women (RR=1.13). Both men and women from China and Hong Kong had more psychiatric visits than those from Taiwan (RR=1.76-2.44 and 2.08-2.68). Both men and women in Fraser North had fewer visits 69 Chapter 3 than those in Vancouver (RR-0.54-0.64). Men in North Shore and Richmond also showed lower rates of visits compared to men in Vancouver (RR=0.30 and 0.40) while women in other regions of BC had a much lower rate than women in Vancouver (RR=0.14). Single immigrants of both sexes were more likely to visit psychiatrists than their married counterparts (RR=2.30-5.39). Women who spoke English had a lower rate of psychiatric visits than those who did not (RR=0.72). Comparison Subjects' Determinants of Mental Health Consultations Results of the multivariate analyses for comparison subjects are presented in Tables 3.4 and 3.5. For young comparison group members, age, rate of non-mental health visits, mean socio-economic quintile of residence and HSDA influenced the rate of mental health visits to both general practitioners and psychiatrists. Among older comparison subjects, rate of non-mental health visits, mean socio-economic quintile of residence and HSDA were associated with consultations with general practitioners; only non-mental health visits to general practice and HSDA affected the rate of psychiatric consultations. While the pattern of the effects of determinants was largely similar between immigrants and the comparison group, especially in visits to general practitioners, the magnitude of the effects of some of the determinants showed variation between the two groups. Probability of Visiting Chinese-speaking Physicians Table 3.6 shows the estimated probability that a mental health visit for a Chinese immigrant was provided by a Chinese-speaking general practitioner or psychiatrist. Almost 90% of mental health visits to general practitioners by Chinese immigrants were made to Chinese-speaking physicians, with the probability being higher among the older immigrants. A similar age pattern was observed for psychiatric visits, although only 67% of overall psychiatric services were received from Chinese-speaking physicians. Supply of Physicians The supply of general practitioners and psychiatrists and Chinese-speaking general practitioners and psychiatrists is shown in Table 3.7 as the number of physicians 70 Chapter 3 per 100,000 population in 1998 by HSDA. Not surprisingly, Vancouver had the highest number of physicians in all categories. South Vancouver Island was in second place for all general practitioners and psychiatrists regardless of language skills. When ability to speak Chinese was considered, Richmond was second to Vancouver in terms of supply of general practitioners; both North Shore and Richmond came after Vancouver for psychiatrists. DISCUSSION Complementing reports that Chinese immigrants as a group are less likely to receive mental health services, this study shows the appreciable and interesting variation in utilization within this population and provides information on some of the factors that influence utilization. The results support assertions of Andersen's behavioural model of health care utilization that both individual and contextual factors affect health care utilization. Andersen's behavioural model postulates that health care utilization is explained by predisposing, enabling and need factors and that equity of access occurs when need factors, rather than predisposing or enabling factors, determine utilization (21-24). Most determinants examined in this study, such as length of time in Canada, place of origin, educational level and marital status, fall under predisposing factors in the individual domain. Some of the factors may play dual roles in this model. For example, English competency may be a predisposing factor in that it is a proxy for health beliefs and acculturation, but English language is also an enabling characteristic because it is frequently a necessary medium to access many forms of care in B C . Age and sex, while individual demographic characteristics, are closely related to risk for certain health conditions and are considered indicators of need. The utilization of general medical services for other than mental health reasons is a process outcome in Andersen's model; in this study on mental health service utilization, it can be considered a predisposing, enabling or need factor depending on the mechanism by which this factor influences use of mental health services, as will be discussed later. In addition to individual characteristics, Andersen's model also recognizes determinants in the social and political context of health care access. The socio-economic quintile of residence in this study, 71 Chapter 3 associated with employment levels and health norms in a neighbourhood, is a predisposing factor in the contextual domain, even though it probably has strong correlation with individual socio-economic status. Some of the contextual enabling factors, such as the availability of health services in a community and the organization of health care, are represented in this study by the health service delivery areas (HSDA) in BC. Two factors consistently emerge in this study as having strong impact on Chinese immigrants' mental health consultations with general practitioners and psychiatrists: years since landing and rate of non-mental health visits. Years since landing: The length of time an immigrant has been in Canada is associated with a substantial increase in the number of mental health consultations of most Chinese immigrants. Both increased need and reduced language and cultural barriers likely converge to bring about the enhanced use of mental health services over time in Canada. The literature suggests that immigrants are at heightened risk of psychological distress not only immediately after arrival but several years later (25). However, unlike the newcomers, immigrants who have settled in their host country have also acquired more language, health beliefs and knowledge about health services to seek help. The finding supports the literature and highlights that, while newcomers are potentially the more under-served group, the needs of immigrants who have settled in this country should not be overlooked. The finding may also be explained by cohort difference, that is, immigrants who arrived in earlier years were higher users of mental health services for some yet unidentified reasons, even after controlling for characteristics which noticeably shifted among the cohorts, such as place of origin and education. Rate of non-mental health visits: In Canada, where specialist care is accessed by referral from the primary care provider, utilization of the two forms of physician services is not independent of each other. On the other hand, there is no peremptory relationship between consultations for mental health and non-mental health conditions. In this study, the rate of non-mental health visits to general practitioners is associated with a dramatic increase in the rate of mental health visits to both general practitioners and psychiatrists. The implications of this finding for access to care depend on the reason for this relationship. The first possible reason is that immigrants with other chronic health 72 Chapter 3 conditions are more likely to develop mental health problems. In this case, the rate of general visits is an indication of need for mental health care, according to Andersen's model, and is useful in predicting high users. A second possible explanation is that individuals who see their general practitioners frequently are more likely to have comorbid conditions, including mental disorders, diagnosed. Non-mental health visit functions thus as an enabling factor in Andersen's model. This scenario raises concern about the mental health needs of those who do not have regular contact with medical professionals. Finally, a third explanation is that certain individuals are simply more inclined to seek medical help, for physical as well as psychological issues. Non-mental health use of primary care becomes a predisposing factor in this case. This explanation also draws attention to the possibility that some immigrants experience barriers, whether cultural, informational or otherwise, to all forms of medical care. The stronger relationship between mental health and non-mental health consultations among the Chinese immigrants relative to comparison subjects suggests that Chinese immigrants' help-seeking behaviour for mental health difficulties is more dependent on having comorbid medical diagnoses. The possibility of under-diagnosis and under-treatment of psychiatric problems in otherwise physically healthy individuals may thus be higher in this population. Other individual predisposing factors, such as age, place of origin, education, marital status and English skill, affect only sub-populations of Chinese immigrants in this study. Age: Age emerges as a strong determinant only among the younger immigrants and the effect is likely due to the very low rate of mental health service utilization among children (26). The age pattern is also observed among the comparison population. Place of origin: Despite sharing a common ethnocultural heritage, the Chinese immigrants differ in their mental health service utilization according to the geo-political regions they migrated from. Some of these differences by place of origin were confounded by several other determinants in the model; place of origin was one of the determinants whose effect was most attenuated between the univariate and multivariate models. For instance, among young females, the low rate of non-mental health visits by those from Taiwan, the older age of those from China and the longer stay in Canada of 73 Chapter 3 those from Hong Kong accounted for much of the difference by place of origin on general practitioner visits. With all these other factors adjusted in the multivariate model, the effect of being from China no longer reached statistical significance while the effect of being from Hong Kong was also reduced substantially. With regard to psychiatric visits by young females, the concentration of immigrants from China and Hong Kong in Vancouver and the larger proportion of Taiwanese immigrants living in Fraser South, two health areas with contrasting utilization rates, also explained some of the effects of place of origin. Nevertheless, controlling for the other determinants did not eliminate all effects by place of origin. In general, those from Macau and Taiwan had the lowest rates of mental health visits to physicians. Immigrants from China and Hong Kong, on the other hand, showed higher rates of mental health visits, with the exception of young men from Hong Kong. These; discrepancies may reflect subtle differences in socio-economic conditions, cultural orientation or immigration environment among Chinese immigrants from different origins. The relationship of these "sub-cultural" differences to risk levels and propensity to access care will have to be explored in future studies. Education: Education is found to have little effect on mental health service utilization in this study. This is consistent with other research studies which are also inconclusive about the effects of education (27-29). Even though having university degrees was often associated with lower rates of utilization in the univariate models, it was found that place of origin, rate of non-mental health visits and years since landing together accounted for a large portion of the univariate effect of education. Those who had lower levels of education were also more likely to be from Hong Kong than Taiwan. Higher levels of education were also associated with more recent immigration and fewer medical visits to the doctor. Most differences by education dissipated after adjusting for the other determinants. In fact, men and women with university degrees became slightly more likely to have mental health consultations than those with lower levels of education. University-educated immigrants may have greater need for care as there is some evidence in the literature that immigrants with higher levels of education encounter more difficulties in adjustment (25). 74 Chapter 3 Marital Status: Married women are minimally more likely than single or separated/divorced/widowed women to receive mental health care from their general practitioners while the opposite is true for married men. Although the effect among women is small, this finding contradicts other research results on the general population that separated/divorced/widowed individuals are at greater risk of using mental health services, even after controlling for sex (27-31). Further perusal of the immigration records reveals that the majority of the immigrants who lost their spouses arrived as sponsored parents/grandparents. While it is possible that, in the Chinese cultural context, these immigrants enjoy greater family support than their equivalent group in general population surveys, barriers such as social isolation and lack of transportation can also underlie their lower utilization of care. On the other hand, the finding about marital status draws attention to the vulnerability of married immigrant women and more attention is required to promote mental health in this group, which comprises three-quarters of the adult female Chinese immigrants to BC. With respect to psychiatric services, single people seem to have much higher rates of service utilization than their married counterparts, in both men and women. This is probably due to the fact that those who engage in psychiatric treatment generally suffer from more severe symptoms and are more likely to delay marriage because of role impairments. English skill: The findings from this study do not support arguments that language barrier is a major reason for lower mental health care use among Chinese immigrants. The effect of English competency was observed only among men for general practitioner visits and among women for psychiatric services. In men, speaking English increased the rate of utilization and, in women, the effect was reversed. Several other issues have to be taken into consideration in assessing the effect of English skill. The first is that, since the Chinese immigrants in B C receive care primarily from general practitioners who speak their language, their English skills are less relevant to access to care than in other environments. In this context, the reason men who speak English have higher rates of visits may instead be that they have higher aspirations for their life in Canada and, consequently, experience greater pressure or disillusionment. The second issue is that English skill was self-reported by immigrants as they entered Canada and is an attribute most likely to change after arrival. The higher probability of misclassification of this i 75 Chapter 3 variable would have led to diminished values of most relationships. Finally, several other determinants among this cohort, including place of origin, education, marital status and years since landing were related to English skill and would have acted as confounders. It is not clear why women who speak English have lower rates of psychiatric visits, even though psychiatric services are less likely to be obtained in the Chinese language. It may be that women with English skills are able to adapt to their new life faster and are less prone to developing serious mental health difficulties. The relationship between English competency and mental health service utilization among Chinese immigrants in B C appears to be much more complex than can be explained by this study. Socio-economic quintile of residence has little impact on the rate of mental health consultations. Under the universal health care policy in Canada, socio-economic status of an individual and the community should have no bearing on health care access. In fact, data from a Canadian health survey showed that individuals from low-income households were more likely to be frequent users of primary care although those with high income were more likely to visit specialists (32). Analyses of administrative databases in Manitoba also reported that individual use of psychiatrists increased with the income level of the neighbourhood (33). In this study, socio-economic quintile shows a similar negative effect in the univariate models for young immigrants and general practitioner visits, but that effect largely disappears after adjusting for other factors. Age, rate of non-mental health visits and place of origin account for the effect of socio-economic environment among these young immigrants. In contrast, socio-economic quintile plays a more influential role among comparison subjects. The comparison group's pattern mirrors the relationship between general physician utilization and household income reported in the Canadian studies. Higher income of the neighbourhood predicts fewer mental health visits to general practice for both comparison age groups but is associated with markedly more psychiatric visits among young comparison members. One explanation for the weak effect of socio-economic quintile in the Chinese immigrant population is that the neighbourhood's average income is less reflective of an immigrant's individual socio-economic status, since an immigrant's choice of residence is influenced by a different set of factors, such as proximity to ethnic community resources and transitional occupational and employment status. Thus, i f the explanatory 76 Chapter 3 power of the neighbourhood's income with regard to individual health care utilization lies in its association with the individual's socio-economic status and resources for health, this contextual component is expected to be less predictive of the outcome in the immigrant than in the non-immigrant population. Health service delivery area (HSDA) indicates directly the organization of health services in a community and represents a contextual variable in Andersen's model. There are marked differences in utilization by H S D A in this study. Despite variations among the sub-groups, a general observation with respect to general practitioner visits is that immigrants in South Vancouver Island and Other areas of B C have higher rates than Vancouver whereas those in Richmond have lower rates. With respect to psychiatric services, the picture is less clear. Factors associated with distance from the metropolitan area may increase the mental health risk of residents in South Vancouver Island and Other regions, thereby increasing their mental health consultations with general practitioners. The lack of psychiatric specialists in the remote Other area, however, keeps the rate of psychiatric visits low. For Chinese immigrants, the lack of ethnic community support and resources in these outlying regions presents an added risk. Resource availability has been reported in the literature to be associated with increased utilization (27, 31, 34). In the case of Chinese immigrants, since Chinese-speaking physicians provide a major proportion of care, both the general availability of physicians and availability of Chinese-speaking physicians are pertinent. However, the present study does not find a clear relationship between resource availability and utilization. For instance, even though Vancouver has the most number of general practitioners per population, its immigrants do not have the highest rate of mental health consultations. In terms of Chinese-speaking general practitioners, South Vancouver Island and Other health service areas of BC, with very few Chinese physicians, have immigrants with higher rates of service utilization, whereas, Richmond, with relatively high concentration of Chinese physicians; has very low rates of utilization. The data are more supportive of a relationship between supply of psychiatrists and utilization. Vancouver, with the most number of general and Chinese-speaking psychiatrists, also has the highest rate of psychiatric visits among the older immigrants while the Other areas, 77 Chapter 3 with few psychiatric specialists, have generally low rates. The relationship, however, does not hold among young immigrants. There are two limitations to this study with regard to establishing a relationship between physician resources and utilization. First, the supply of physicians in this study represents a count of all actively registered physicians, including those in post-graduate training; it does not take into account the work pattern of individual physicians and is not likely an accurate reflection of the actual amount of physician time available for clinical care. Second, since most of the immigrants in this study are located in metropolitan Vancouver, their access is not restricted to resources in their own health jurisdiction. Furthermore, 18% of the immigrants have changed HSDAs during the study period. Hence, there may be very little true difference in terms of geographic access to physicians among the majority of immigrants in this study. A preliminary analysis of the data reveals that, in 1998, 27% of all the visits by Chinese immigrants to Chinese-speaking general practitioners and psychiatrists were outside of the immigrant's HSDA. While this study does not demonstrate that higher number of Chinese-speaking physicians in the population necessarily increases utilization by Chinese immigrants, the language ability of physicians may be related to health care utilization in different ways. The probabilities of Chinese immigrants being treated by Chinese-speaking general practitioners and psychiatrists for mental health difficulties were 0.88 and 0.67 respectively, while the corresponding figures for the comparison population were 0.11 and 0.08. This is an indication that Chinese immigrants and the rest of the population receive care from different clusters of physicians with at most mild overlap. Therefore, an alternative explanation to Chinese immigrants' under-utilization of mental health services can be that their physicians have different practice patterns. It has yet to be ruled out that these physicians are less likely to diagnose mental health problems or less likely to refer them to specialists for psychiatric care. There is evidence that, in addition to patient characteristics, physicians' inclination towards providing counselling also influences their detection of mental health problems (35). These hypotheses will have to be explored in future studies. A limitation of this study is that the data are drawn from service utilization records. While these databases provide a comprehensive picture of utilization, 78 Chapter 3 identification of mental health visits is only approximate and they do not offer information about the prevalence of mental health difficulties in the community. As such, these databases alone cannot provide direct answers about equity of access according to Andersen's behavioural model. Nevertheless, the findings have identified sub-groups of Chinese immigrants who are more at risk of under-utilization. In seeking to improve access to mental health services, the next step is to understand the extent of need, the reasons for lower rates of utilization and strategies to reduce the disparity. CONCLUSION Number of years since landing and rate of medical visits to general practitioners are consistently strong predictors of mental health visits to general practitioners and psychiatrists by Chinese immigrants. Other individual factors contributing to mental health visits include age, place of origin of the immigrants, educational level, marital status and English skill. The health service area of an immigrant's residence also influences the rate of mental health visits, although there is no evidence from this study that the relationship is explained by supply of physicians. Findings from this study complement Andersen's behavioural model by confirming the effect of some factors and introducing others that are specific to the immigrant population and to mental health services. In addition to the predisposing factors of age, ethnicity, education and marital status that are in the original model, this study provides evidence that length of residence in Canada, place of origin and English competency are also pertinent factors in the immigrant population. Socio-economic environment is not found to have impact on mental health care utilization in this study. Other contextual factors, such as the organization and delivery of health care, apparently play a role although the specific elements that exert the influence is not yet clear. This study also shows that use of general medical services, while an outcome in itself, can be considered a contributing variable with respect to use of mental health services, but again, the precise nature of that relationship is not known. These determinants of utilization highlight the fact that Chinese immigrants are not homogeneous in relation to mental health services; some sub-groups are at higher risk of experiencing mental health difficulties or barriers to services. Research that treats 79 Chapter 3 immigrants or ethnic minorities as one entity is inadequate in describing the utilization pattern or identifying the under-served groups. Utilization of mental health services is undeniably a complex behaviour that results from dynamic interplay of a myriad of factors. Findings from this study begin to unravel some of this complexity. Research in the future will have to elucidate some of the other factors in the dynamic, such as need for care, health beliefs and alternative resources. Quantifying the prevalence of mental health problems in different sub-populations will clarify i f the variation in utilization is due to differences in morbidity. Besides, identifying the risk factors in each sub-population is crucial to targeted efforts to promote mental health among immigrants For instance, understanding the stresses encountered by more established immigrants versus newcomers, the mental health risks of married women and older adults, the relationship between employment status and education credentials and mental health outcome will contribute to policies and programs that enhance the settlement and integration of immigrants. The support systems and informal resources of immigrants may also explain the variation in utilization among immigrants from different origins and visa categories; knowledge in this regard is informative to immigration policies as well as mental health services. Furthermore, insight into how immigrants of various demographic backgrounds perceive mental, health symptoms and what they believe are effective remedies will shed light on the nuances of the cultural barriers to providing mental health care. Another area of research that is hitherto untapped is the utilization of traditional Chinese medicine; probing into this domain will help gauge the respective roles of formal medical care and alternative therapies in mental health interventions among the immigrants and their sub-populations. Finally, further knowledge is necessary to understand how contextual factors, such as the delivery of care and the availability of culturally responsive services, affect immigrants' access to mental health care. This study demonstrates that Chinese immigrants receive care predominantly from physicians who speak their native language. Questions that remain to be answered are whether increasing the number of Chinese-speaking practitioners is a solution to the challenge of cultural barriers and whether there is a relationship between the demographic characteristics of practitioners and their practice patterns. 80 Chapter 3 Table 3.1a Characteristics of Study Immigrants Aged Under 25 at Landing and Comparison Subjects Total Number Mean Age at Entry to Study (Years) Mean Years in Study Mean Socio-economic Quintile Health Service Delivery Area (HSDA) Vancouver Richmond North Shore/Coast Garibaldi Fraser North Fraser South South Vancouver Island Other Year of Landing 1985-1988 1989-1991 1992-1994 1995-1997 1998-2000 Immigrant Female 25633 13.97 5.59 2.99 39.58% 22.67% 3.37% 23.76% 7.75% 1.42% 1.46% 6.30% 12.89% 29.85% 35.22% 15.74% Male 25778 13.21 5.56 3.01 37.90% 22.82% 3.70% 24.40% 8.46% 1.31% 1.41% 5.99% 12.79% 29.93% 35.41% 15.88% Comparison Female 25474 13.7 6.14 2.93 34.72% 16.22% 4.76% 23.15% 12.95% 2.04% 6.16% Male 25652 12.94 6.02 2.97 34.21% 16.93% 4.82% 23.00% 12.95% 1.91% 6.18% Place of Origin Hong Kong China Taiwan Macau 48.50% 17.13% 33.48% 0.89% 48.40% 14.75% 35.91% 0.95% 81 Chapter 3 Table 3.1b Characteristics of Study Immigrants Aged 25 or Oyer at Landing and Comparison Subjects Total Number Mean Age at Entry to Study (Years) Mean Years in Study Mean Socio-economic Quintile Health Service Delivery Area (HSDA) Vancouver Richmond North Shore/Coast Garibaldi Fraser North Fraser South South Vancouver Island Other Year of Landing 1985-1988 1989-1991 1992-1994 1995-1997 1998-2000 Immigrant Comparison Female 51896 43.01 5.54 2.79 41.23% 23.01% 2.97% 23.46% 6.20% 1.64% 1.48% 5.37% 11.81% 29.16% 34.84% 18.83% Male 48877 44.36 5.58 2.80 40.94% 22.87% 2.89% 23.88% 6.32% 1.62% 1.49% 5.82% 14.18% 29.74% 32.91% 17.35% Female 51691 42.67 6.09 2.84 36.39% 16.81% 4.44% 22.30% 11.28% 2.38% 6.40% Male 48486 44 6.11 2.85 36.37% 16.40% 4.37% 22.58% 11.52% 2.37% 6.40% Place of Origin Hong Kong China Taiwan Macau 49.88% 26.49% 22.85% 0.78% 49.84% 25.08% 24.25% 0.83% Education 0-9 years 10-12 years Non-University University Degree Marital Status Single Married Separated/Divorced/Widowed 26.60% 24.66% 29.07% 19.67% 16.23% 74.58% 9.19% 18.40% 21.60% 28.92% 31.08% 16.23% 81.81% 1.96% English Ability English No English 45.65% 54.35% 55.60% 44.40% 82 Table 3.2a Determinants of Chinese Immigrants' Mental Health Visits to General Practitioners - Immigrants Under 25 Years Old at Landing Female Male Univariate Multivariate Univariate Multivariate Rate Ratio 4 95% CI Rate Ratio4 95% CI Rate Ratio 4 95% CI Rate Ratio4 95% CI Age 1 1.13 (1.12,1.13) 1.10 (1.09,1.11) 1.08 (1.08,1.09) 1.08 (1.07,1.08) Years Since Landing 1 1.18 (1.16,1.21) 1.07 (1.05,1.15) 1.18 (1.15,1.20) 1.13 (1.11,1.16) Rate of Non-Mental Health Visi ts 2 1.26 (1.25,1.28) 1.20 (1.19,1.22) 1.15 (1.13,1.17) 1.19 (1.17,1.21) Mean Socio-economic Quinti le 3 0.87 (0.84,0.90) 1.01 (0.98,1.04) 0.83 (0.80,0.86) 0.92 (0.88,0.96) Place of Origin China 2.19 (1.96,2.45) 1.10 (0.97,1.24) 2.59 (2.26,2.98) 1.89 (1.62,2.22) Hong Kong 1.66 (1.52,1.80) 1.20 (1.09,1.30) 1.53 (1.37,1.69) 0.99 (0.89,1.11) Macau 1.00 (0.66,1.52) 0.79 (0.52,1.19) 1.88 (1.21 ,2.94) 1.01 (0.64,1.58) Taiwan 1.00 1.00 1.00 1.00 Health Service Delivery Area (HSDA) Fraser North 0.85 (0.77,0.93) 1.03 (0.94,1.13) 0.78 (0.69,0.87) 1.10 (0.98,1.24) Fraser South 0.74 (0.64,0.86) 1.01 (0.87,1.18) 0.53 (0.44,0.64) 0.94 (0.77,1.14) North Shore 0.58 (0.46,0.73) 0.83 (0.67,1.03) 0.60 (0.46,0.78) 1.01 (0.78,1.32) Richmond 0.66 (0.60,0.73) 0.76 (0.69,0.83) 0.63 (0.56,0.71) 0.80 (0.71,0.90) South Vancouver Island 1.37 (1.01 ,1.86) 1.32 (0.99,1.76) 2.27 (1.57,3.29) 2.90 (2.04,4.13) Other, 1.21 (0.88,1.66) 1.92 (1.41,2.62) 1.62 (1.10,2.40) 1.55 (1.03,2.32) Vancouver 1.00 1.00 1.00 1.00 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year1 3 Rate ratio is change per unit (range = 1 to 5) 4 Rate ratio in bold type indicates a result that is significantly different from 1 at 5% level Table 3.2b Determinants of Chinese Immigrants' Mental Health Visits to General Practitioners - Immigrants Aged 25 Years or Over at Landing Female Male Univariate Rate Ratio4 95% CI Multivariate Pate Ratio4 95% c i Univariate Rate Ratio4 95% CI Multivariate Rate Ratio4 95% CI Age 1 1.00 (1.00,1.00) 1.00 (1.00,1.00) 1.01 (1.01,1.01) 0.99 (0.99,1.00) Years Since Landing 1 1.10 (1.09,1.11) 1.08 (1.07,1.09) 1.13 (1.11,1.14) 1.07 (1.05,1.08) Rate of Non-Mental Health Visits 2 1.18 (1.17,1.18) 1.18 (1.17,1.18) 1.23 (1.22,1.24) 1.23 (1.22,1.24) Mean Socio-economic Quintile3 0.97 (0.96,0.99) 1.02 (1.00,1.03) 0.96 (0.94,0.98) 1.01 (0.98,1.03) Place of Origin China 1.34 (1.26,1.41) 1.11 (1.05,1.18) 1.56 (1.44,1.68) 1.45 (1.33,1.58) Hong Kong 1.26 (1.20,1.32) 1.01 (0.96,1.07) 1.45 (1.36,1.55) 1.18 (1.10,1.27) Macau 1.02 (0.82,1.26) 0.80 (0.64,0.98) 1.05 (0.79,1.40) 0.94 (0.70,1.25) Taiwan 1.00 1.00 1.00 1.00 Hearth Service Delivery Area (HSDA) Fraser North 0.99 (0.95,1.04) 1.05 (1.00,1.10) 0.88 (0.82,0.94) 0.94 (0.88,1.01) Fraser South 0.92 (0.84,0.99) 1.06 (0.98,1.15) 0.85 (0.77,0.95) 1.05 (0.94,1.17) North Shore 0.77 (0.69,0.87) 0.80 (0.71,0.90) 0.99 (0.85,1.16) 1.25 (1.07,1.45) Richmond 0.77 (0.73,0.81) 0.74 (0.71,0.78) 0.72 (0.67,0.77) 0.71 (0.66,0.76) South Vancouver Island 1.13 (0.97,1.32) 1.33 (1.15,1.53) 1.37 (1.12,1.67) 1.47 (1.22,1.77) Other 0.91 (0.77,1.09) 0.98 (0.82,1.18) 1.06 (0.83,1.34) 1.29 (1.01 ,1.64) Vancouver 1.00 1.00 1.00 1.00 Education 0-9 Years 1.12 (1.06,1.19) 0.86 (0.80,0.92) 1.17 (1.09,1.27) 0.89 (0.81 ,0.97) 10-12 Years 1.18 (1.12,1.26) 1.03 (0.97,1.09) 1.17 (1.09,1.26) 0.98 (0.90,1.06) Non-University 1.09 (1.03,1.15) 0.97 (0.92,1.03) 1.27 (1.19,1.36) 1.09 (1.02,1.17) University Degree 1.00 1.00 1.00 1.00 Marital Status Single 0.88 (0.84,0.93) 0.94 (0.89,0.99) 1.19 (1.11,1.27) 1.22 (1.14,1.32) Separated/Divorced/Widowed 0.97 (0.91 ,1.04) 0.92 (0.85,0.98) 1.33 (1.11,1.60) 1.10 (0.92,1.32) Married 1.00 1.00 1.00 1.00 English Skill English 0.97 (0.93,1.00) 1.02 (0.98,1.07) 1.07 (1.02,1.13) 1.10 (1.04,1.17) No English 1.00 1.00 1.00 1.00 n 00 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 4 Rate ratio in bold type indicates a result that is significantly different from 1 at 5% level Table 3.3a Determinants of Chinese Immigrants' Visits to Psychiatrists - Immigrants Under 25 Years Old at Landing Female Male Univariate Multivariate Univariate Multivariate Rate Ratio 4 95% CI Rate Ratio4 95% CI Rate Ratio 4 95% CI Rate Ratio4 95% CI Age 1 1.08 (1.05,1.10) 1.04 (1.01,1.07) 1.05 (1.02 1.08) 1.05 (1.01,1.08) Years Since Landing 1 1.26 (1.16,1.36) 1.15 (1.05,1.25) 1.16 (1.06 1.26) 1.09 (0.98,1.20) Rate of Non-Mental Health Visits 2 1.15 (1.08,1.23) 1.09 (1.02,1.16) 1.12 (1.04 1.21) 1.13 (1.04,1.23) Mean Socio-economic Quinti le 3 0.89 (0.78,1.01) 0.95 (0.82,1.10) 0.98 (0.86 1.12) 0.97 (0.83,1.13) Place of Origin China 1.97 (1.21,3.18) 1.27 (0.72,2.26) 1.25 (0.75 2.09) 1.11 (0.62,1.99) Hong Kong 2.74 (1.91,3.95) 1.64 (1.10,2.45) 1.21 (0.84 1.75) 0.97 (0.63,1.50) Macau 0.10 (0.01,0.89) 0.07 (0.01,0.68) 0.32 (0.05 1.92) 0.27 (0.04,1.59) Taiwan 1.00 1.00 1.00 1.00 Health Service Delivery Area (HSDA) Fraser North 0.56 (0.37,0.86) 0.72 (0.47,1.12) 0.80 (0.52 1.22) 1.04 (0.65,1.65) Fraser South 0.23 (0.12,0.45) 0.43 (0.22,0.88) 0.37 (0.20 0.70) 0.49 (0.24,1.00) North Shore 0.36 (0.14,0.91) 0.63 (0.25,1.61) 2.07 (0.86 4.99) 2.17 (0.87,5.42) Richmond 0.55 (0.36,0.85) 0.67 (0.44,1.04) 0.87 (0.57 1.34) 1.00 (0.65,1.55) South Vancouver Island 1.70 (0.44,6.62) 1.88 (0.49,7.17) 2.19 (0.52 9.16) 2.54 (0.60,10.65 Other 0.51 (0.13,2.06) 1.07 (0.23,5.00) 1.53 (0.36 6.46) 1.16 (0.22,6.21) Vancouver 1,00 1.00 1.00 1.00 n •s CD >-t 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 4 Rate ratio in bold type indicates a result that is significantly different from 1 at 5% level oo Table 3.3b Determinants of Chinese Immigrants' Visits to Psychiatrists - Immigrants Aged 25 Years or Over at Landing Female Male Univariate Rate Ratio4 95% CI Multivariate Rate Ratio4 95% CI Univariate Rate Ratio4 95% CI Multivariate Rate Ratio4 95% CI Age' 1.00 (0.99 1.01) 0.99. (0.98,1.00) 0.99 (0.98 1.00) 1.00 (0.98 1.01) Years Since Landing1 1.21 (1.14 1.29) 1.13 (1.06,1.21) 1.07 (1.00 1.15) 1.02 (0.95 1.10) Rate of Non-Mental Health Visits 2 1.14 (1.10 1.17) 1.15 (1.12,1.19) 1.16 (1.10 1.22) 1.26 (1.19 1.34) Mean Socio-economic Quintile3 0.91 (0.82 1.00) 0.97 (0.88,1.08) 0.96 (0.84 1.10) 1.05 (0.91 1.22) Place of Origin China 3.14 (2.26 4.37) 2.44 (1.67,3.54) 1.31 (0.83 2.06) 1.76 (1.05 2.94) Hong Kong 4.15 (3.09 5.56) 2.68 (1.87,3.82) 1.49 (1.00 2.21) 2.08 (1.36 3.17) . Macau 2.15 (0.59 7.84) 2.97 (0.81 ,10.87) 1.76 (0.32 9.76) 1.26 (0.25 6.37) Taiwan 1.00 1.00 1.00 1.00 Health Service Delivery Area (HSDA) Fraser North 0.53 (0.40 0.72) 0.64 (0.48,0.86) 0.57 (0.38 0.85) 0.54 (0.36 0.80) Fraser South 0.36 (0.22 0.59) 0.73 (0.43,1.26) 0.51 (0.26 0.99) 0.96 (0.49 1.85) North Shore 0.73 (0.37 1.45) 1.33 (0.64,2.77) 0.31 (0.12 0.81) 0.30 (0.11 0.79) Richmond 0.82 (0.61 1.10) 0.83 (0.62,1.12) 0.50 (0.34 0.75) 0.40 (0.27 0.61) South Vancouver Island 0.59 (0.24 1.47) 1.08 (0.44,2.61) 0.45 (0.13 1.57) 0.68 (0.21 2.23) Other 0.04 (0.01 0.17) 0.14 (0.03,0.59) 0.24 (0.06 0.97) 0.22 (0.04 1.07) Vancouver 1.00 1.00 1.00 1.00 Education 0-9 Years 1.67 (1.18 2.36) 0.73 (0.47,1.13) 0.85 (0.53 1.34) 0.97 (0.56 1.68) 10-12 Years 1.74 (1.22 2.47) 0.83 (0.57,1.22) 0.72 (0.46 1.11) 0.78 (0.50 1.24) Non-University 1.52 (1.08 2.13) 0.83 (0.58,1.19) 0.95 (0.63 1.42) 0.98 (0.66 1.46) University Degree 1.00 1.00 1.00 1.00 Marital status Single 2.01 (1.47 2.75) 2.30 (1.66,3.20) 3.43 (2.27 5.16) 5.39 (3.45 8.44) Separated/Divorced/Widowed 1.14 (0.76 1.70) 0.96 (0.60,1.53) 0.57 (0.18 1.75) 0.35 (0.11 1.09) Married 1.00 1.00 1.00 1.00 English Skill English 1.00 (0.79 1.26) 0.72 (0.54,0.95) 1.26 (0.92 1.73) 0.92 (0.63 1.34) No English 1.00 1.00 1.00 1.00 o -8. i-i 00 0\ Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 4 Rate ratio in bold type indicates a result that is significantly different from 1 at 5% level Table 3.4 Determinants of Comparison Subjects' Mental Health Visits to General Practitioners (Multivariate Models) Subjects Matched with Immigrants Under 25 Years Old at Landing Subject Matched with Immigrants Aged 25 Years or Over at Landing Female Male Female Male Rate Ratio4 95% CI Rate Ratio4 '95% CI Rate Ratio4 95% CI Rate Ratio4 95% CI Age 1 1.14 (1.13,1.14) 1.12 (1.11,1.13) 0.99 (0.99,0.99) 0.98 (0.98,0.98) Rate of Non-Mental Health Visits 2 1.17 (1.16,1.18) 1.14 (1.12,1.16) 1.07 (1.07,1.08) 1.10 (1.10,1.11) Mean Socio-economic Quinti le 3 0.95 (0.93,0.97) 0.87 (0.85,0.90) 0.87 (0.86,0.88) 0.82 (0.80,0.83) Health Service Delivery Area (HSDA) Fraser North 1.04 (0.97,1.12) 0.99 (0.90,1.09) 1.05 (1.00,1.09) 0.75 (0.71 ,0.80) Fraser South 1.01 (0.92,1.10)^ 1.05 (0.93,1.19) 1.21 (1.15,1.28) 0.83 (0.77,0.89) North Shore 1.04 (0.90,1.19) 0.81 (0.67,0.97) 1:02 (0.94,1.11) 0.78 (0.70,0.88) Richmond 0.78 (0.71,0.84) 0.80 (0.72,0.90) 0.87 (0.83,0.91) 0.65 (0.61 ,0.70) South Vancouver Island 1.29 (1.07,1.55) 1.05 (0.80,1.38) 1.16 (1.04,1.29) 1.66 (1.43,1.92) Other 1.47 (1.31,1.64) 1.44 (1.23,1.69) 1.18 (1.10,1.26) 0.86 (0.78,0.94) Vancouver 1.00 1.00 1.00 r 1.00 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 4 Rate ratio in bold type indicates a result that is significantly different from 1 at 5% level Table 3.5 Determinants of Comparison Subjects' Visits to Psychiatrists (Multivariate Models) Subjects Matched with Immigrants Under 25 Years Old at Landing Subject Matched with Immigrants Aged 25 Years or Over at Landing Female Male Female Male Rate Ratio4 95% CI Rate Ratio4 95% CI Rate Ratio4 95% CI Rate Ratio4 95% CI Age 1 1.09 (1.07,1.11) 1.03 (1.01,1.04) 0.99 (0.98,0.99) 0.99 (0.98,0.99) Rate of Non-Mental Health Visits 2 1.08 (1.06,1.11) 1.04 (1.01 ,1.08) 1.05 (1.04,1.06) 1.06 (1.05,1.08) Mean Socio-economic Quintile 3 1.21 (1.13,1.30) 1.15 (1.07,1.24) 1.00 (0.96,1.05) 1.01 (0.96,1.07) Health Service Delivery Area (HSDA) Fraser North 0.71 (0.57,0.89) 0.50 (0.39,0.63) 0.46 (0.39,0.53) 0.51 (0.43,0.61) Fraser South 0.66 (0.50,0.87) 0.64 (0.48,0.85) 0.57 (0.47,0.70) 0.49 (0.39,0.61) North Shore 1.00 (0.66,1.50) 0.97 (0.63,1.48) 0.68 (0.51,0.91) 0.88 (0.63,1.22) Richmond 0.84 (0.65,1.08) 1.00 (0.77,1.29) 0.72 (0.61,0.85) 0.70 (0.58 0.85) South Vancouver Island 0.60 (0.33,1.10) 1.18 (0.62,2.23) 0.91 (0.63,1.32) 0.78 (0.51,1.20) Other 0.73 (0.51 ,1.05) 0.75 (0.51 ,1.11) 0.56 (0.44,0.72) 0.44 (0.33,0.58) Vancouver 1.00 1.00 1.00 1.00 o 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 4 Rate ratio in bold type indicates a result that is significantly different from 1 at 5% level oo oo Table 3.6 Probability of Mental Health Visits to Chinese-speaking Physicians by Chinese Immigrants General Practitioners Psychiatrists No. of Visits Probability 95% CI No. of Visits Probability 95% CI Female Under 25 at landing 9999 0.82 (0.79,0.85) 5166 0.59 (0.52,0.66) Aged 25 or over at landing 48707 ,0.91 (0.90,0.91) 12102 0.72 (0.66,0.78) Male Under 25 at landing 7078 0.72 (0.60,0.81) 5508 0.56 (0.47,0.65) Aged 25 or over at landing 23672 0.88 (0.86,0.90) 6003 0.74 (0.68,0.79) ALL 89456 0.88 (0.86,0.89) 28779 0.67 (0.64,0.71) 00 Chapter 3 Table 3.7 Number of Directory Active Physicians per 100, Health Service Delivery Area GP Vancouver 278.8 Richmond 98.3 North Shore/Coast Garibaldi 116.8 Fraser North 96.4 Fraser South 78.5 South Vancouver Island 161.4 Other 102.6 BC TOTAL 140.5 Physician numbers are from College of Physicians Population estimates are from BC STATS, Service D Population in 1998 Chinese Psychiatrist Chinese GP Psychiatrist 41.3 39.2 2.4 7.7 27.9 0.9 12.4 4.2 1.1 10.2 10.8 0.4 6.0 9.8 0.0 18.8 4.5 0.0 5.9 2.3 0.0 14.2 11.8 0.6 and Surgeons of British Columbia BC, BC Ministry of Management Services. 90 Chapter 3 BIBLIOGRAPHY 1. Cheung F K , Snowden LR. Community mental health and ethnic minority populations. Community Ment Health J 1990; 26: 277-291. 2. Klimidis S, McKenzie DP, Lewis J, Minas JU. Continuity of contact with psychiatric services: immigrant and Australian-born patients. Soc Psychiatry Psychiatr Epidemiol 2000; 35: 554-563. 3. Leong FTL. Asian Americans' differential patterns of utilization of inpatient and outpatient public mental health services in Hawaii. J Community Psychol 1994; 22: 82-96. 4. Meinhardt K , Vega W. A method for estimating underutilization of mental health services by ethnic groups. Hosp Community Psychiatry 1987; 38: 1186-1190. 5. Roberts N , Crockford D. Psychiatric admissions of Asian Canadians to an adolescent inpatient unit. Can J Psychiatry 1997; 42: 847-851. 6. Snowden LR, Cheung F K . Use of inpatient mental health services by members of ethnic minority groups. A m Psychol 1990; 45: 347-355. 7. Intercultural Committee. Multiculturalism and Mental Health: Developing Culturally Competent Systems of Care. Vancouver: Greater Vancouver Mental Health Service: 1999. 8. Lai D W L , Tsang K T , Chappell N L , Lai D C Y , Chau SBY. Health and Well Being of Older Chinese in Canada. Calgary: Faculty of Social Work, University of Calgary: 2003. 9. Peters R. Increasing the responsiveness of mainstream mental health services to ethnocultural minorities. In: Health and Cultures V o l : II Programs, Services and Care. Edited by Masi R, Mensah L , McLeod K A . Oakville: Mosaic Press: 1993: 187-203. 10. Freimen MP, Cunningham PJ. Use of health care for the treatment of mental problems among racial/ethnic subpopulations. Med Care Res Rev 1997; 54: 80-100. 11. Padgett D K , Patrick C, Burns BJ, Schlesinger HJ. Ethnicity and the use of outpatient mental health services in a national insured population. A m J Public Health 1994; 84: 222-226. 12. Padgett D K , Patrick C, Burns BJ, Schlesinger HJ. Ethnic differences in use of inpatient mental health services by blacks, whites and Hispanics in a national insured population. Health Serv Res 1994; 29: 135-153. 13. Vega WA, Kolody B, Aguilar-Gaxiola S, Catalano R. Gaps in service utilization by Mexican Americans with mental health problems. A m J Psychiatry 1999; 156: 928-934. 91 Chapter 3 14. Wells K B , Golding J M , Hough R L , Burnam M A , Karno M . Factors affecting the probability of use of general and medical health and social/community services for Mexican Americans and non-Hispanic whites. Med Care 1988; 26: 441-452. 15. Wells K B , Golding J M , Hough RL, Burnam M A , Karno M . Acculturation and the probability of use of health services by Mexican Americans. Health Serv Res 1989; 24:237-257. 16. Spencer M S , Chen J. Effect of discrimination on mental health service utilization among Chinese Americans. Am J Public Health 2004; 94: 809-814. 17. Stuart GW, Minas HI, Klimidis S, O'Connell S. English language ability and mental health service utilisation: a census. Aust N Z J Psychiatry 1996; 30: 270-277. 18. Tabora B L , Flaskerud JH. Mental health beliefs, practices, and knowledge of Chinese American immigrant women. Issues Ment Health Nurs 1997; 18: 173-189. 19. Newcombe HB. Handbook of Record Linkage: Methods for Health and Statistical Studies, Administration, and Business. New York: Oxford Medical Publications: 1988. 20. Gardner W, Mulvey EP, Shaw EC. Regression analyses of counts and rates: Poisson, overdispersed Poisson, and negative binomial models. Psychol Bull 1995; 118: 392-404. 21. Andersen R. A Behavioral Model of Families Use of Health Services. Center for Health Administration Studies, Research Series 25. The University of Chicago: 1968. 22. Andersen R, Newman JF. Societal and individual determinants of medical care utilization in the United States. Milbank Mem Fund Q 1973; 51: 95-124! 23. Andersen R M . Revisiting the behavioral model and access to medical care: Does it matter? J Health Soc Behav 1995; 36: 1-10. 24. Andersen R M , Davidson PL. Improving access to care in America. In: Changing the US Health Care System. Edited by Rice T H , Kominski GF. San Francisco: Jossey-Bass: 2001: 3-30. 25. Review of the Literature on Migrant Mental Health. Canadian Task Force on Mental Health Issues Affecting Immigrants and Refugees: 1988. 26. Chen A W . (Unpublished manuscript) Disparities in Utilization of Mental Health Services by Chinese Immigrants in British Columbia. University of British Columbia: 2006 27. Lin E, Goering P, Offord DR, Campbell D, Boyle M H . The use of mental health services in Ontario: epidemiologic findings. Can J Psychiatry 1996; 41: 572-577. 92 Chapter 3 28. Parslow R A , Jorm A F . Who uses mental health services in Australia? An analysis of data from the National Survey of Mental Health and Wellbeing. Aust N Z J Psychiatry 2000; 34: 997-1008. , 29. ten Have M , Vollebergh W, Bijl R V , de Graaf R. Predictors of incident care service utilisation for mental health problems in' the Dutch general population. Soc Psychiatry Psychiatr Epidemiol 2001; 36: 141-149. 30. Bebbington PE, Meltzer H , Brugha TS et al. Unequal access and unmet need: neurotic disorders and the use of primary care services. Psychol Med 2000; 30: 1359-1367. 31. Lefebvre J, Lesage A, Cyr M , Toupin J, Fournier L . Factors related to utilization of services for mental health reasons in Montreal, Canada. Soc Psychiatry Psychiatr Epidemiol 1998; 33: 291-298. 32. Dunlop S, Coyte PC, Mclsaac W. Socio-economic status and the utilisation of physicians'services: results from the Canadian National Population Health Survey. Soc Sci Med 2000; 51: 123-133. 33. Martens P, Fransoo R, McKeen N et al. Patterns of Regional Mental Illness Disorder Diagnoses and Service Use in Manitoba: A Population-Based Study. Manitoba Centre for Health Policy, University of Manitoba: September 2004. 34. Shapiro S, Skinner E A , Kessler L G et al. Utilization of health and mental health services: three Epidemiologic Catchment Area sites. Arch Gen Psychiatry 1984; 41: 971-978. 35. Borowsky SJ, Rubenstein L V , Meredith LS, Camp P, Jackson-Triche M , Wells K B . Who is at risk of nondetection of mental health problems in primary care? J Gen Intern Med 2000; 15: 381-388. 93 CHAPTER 4 MENTAL H E A L T H SERVICE UTILIZATION BY CHINESE IMMIGRANTS WITH SEVERE AND PERSISTENT MENTAL ILLNESS IN BRITISH COLUMBIA 1 Studies have demonstrated that immigrants and ethnic minorities, particularly those of Asian or Chinese origin, are less likely to use mental health services (1-11). This raises concerns about the equity of access to mental health services for a population who is particularly at risk as a result of its migration and resettlement experience (12). Moreover, the literature indicates that the disparity in mental health service utilization is not limited to people with mild emotional and adjustment difficulties; under-utilization of mental health services persists among the most vulnerable group of those with serious mental illness. Among Australian patients with schizophrenia or bipolar disorder, all immigrants were found to be under-represented in inpatient and outpatient services compared to the Australian-born, and Asian immigrants had the lowest level of utilization (5). Asian Americans were under-represented in all types of inpatient units and mental health services (3, 8). Among Asian Americans identified as having serious mental illness, the odds of using mental health care was lower than Whites (3). In Canada, Vancouver's community mental health system, the mandate of which is restricted to patients with chronic and serious mental disorders, reported that the proportion of Chinese in the caseload was under two-thirds of that in the population (4, 10). Similarly, an adolescent inpatient psychiatric unit in Calgary reported that the admission rate of Asians in the city was only one-third of that of white Canadians (11). Furthermore, there is evidence that Asians often delay help-seeking and use mental health care only as a "last resort" for very acute symptoms. Asian Americans in New York City's public mental health system, while being under-represented, were more likely than Whites to be diagnosed with schizophrenia and more likely to use emergency services rather than outpatient services (13). Asian-Canadian youths in Calgary were also predominantly admitted to the psychiatric hospital on an emergency basis and had relatively higher rates than whites of diagnoses of psychotic disorders, major mood 1 A version of the chapter will be submitted for publication in a journal. 94 Chapter 4 disorders and substance use disorders whereas none were diagnosed with adjustment disorder or minor mood disorders. Similar observations were also made of patients with English difficulties in Australia, who were more likely diagnosed with psychosis and dementia and less often with situational crises than those who spoke fluent English (14). While Asians have lower probability of receiving any mental health services, some studies suggest that their amount of utilization may not differ once they have contact with the system. A survey of mental health hospitals in the United States showed that while Asian Americans had lower admission rates in all types of institutions, their length of stay, once admitted, was higher or not different from Whites (3). In Calgary, Asian adolescents' length of stay in a psychiatric unit was equivalent to that of white Canadians (11). Among Australian patients with schizophrenia or bipolar disorder, there was no difference by place of origin in number of outpatient contacts, interval between contacts, number of inpatient admissions or length of stay once they began to receive services (5). Access to mental health services by Chinese immigrants is of particular concern in British Columbia (BC), Canada, where 6% of its 3.9 million residents in 2001 were immigrants of Chinese origins. Moreover, almost two-thirds of these Chinese immigrants arrived within the previous decade. Chinese territories remain a major source of immigrants to B C ; in 2001, the People's Republic of China, Taiwan and Hong Kong contributed 35% of the 34,000 new arrivals in Vancouver. A previous study reported that recent Chinese immigrants in B C w,ere far less likely to use mental health services, including physician visits, hospital and outpatient mental health care. In general, Chinese immigrants used under 20% of the services of non-immigrants (15). A large proportion of utilization, as expected, was for less serious mental health problems, such as neurotic and depressive disorders. However, that study was not able to conclude if difference in morbidity level or barrier to care accounted for the disparity in utilization. If the health care system is equitable and the differences in overall utilization can be explained by lower morbidity and need for care among 2 Statistics Canada 2001 Census of Canada. Catalogue number 97F0009XCB2001002 http://wwwl2.statcan.ca/englisr^census01/Products/standara7thernes/DataProducts.cfm?S=l 3 Citizenship and Immigration Canada Facts and Figures 2001 http://www.cic.gc.ca/english/pub/facts2001/4van-02.html 95 Chapter 4 immigrants, the expectation is that immigrants and non-irnmigrants diagnosed with serious conditions will receive similar levels of health services, since they have similar levels of need. Therefore, the next question in assessing equity of care is whether disparity in utilization also exists for Chinese immigrants known to have serious mental disorders. At the same time, there is emerging evidence of the relationship between physicians' demographic characteristics and practice patterns that is particularly relevant to immigrants and minorities. For instance, foreign graduate psychiatrists in the United States were racially more diverse, worked more frequently in the public sector and treated more patients with psychotic disorders (16). Psychiatrists who had high caseloads of substance use disorders were younger and more likely to be foreign graduates and to work in inpatient or publicly-funded settings (17). Primary care physicians who had a preference for providing counselling were more inclined to detect mental health problems in their patients (18). In other areas of health care, Chinese women in Vancouver's Chinatown had the lowest rate of cervical cancer screening; meanwhile, Chinese doctors in the area tended to be older males and performed fewer Pap tests (19). The available evidence, though limited in scope, supports assertions that utilization of health care is related to both patient and provider factors. It also leads to speculations about the role of physician characteristics in the utilization of mental health care by Chinese immigrants. A study in Australia found that patients with poor English proficiency were markedly more likely to consult general practitioners, psychiatrists and psychologists who spoke their first language (14). An earlier study also reported that almost 90% of the mental health services in primary care and almost 70% of psychiatric services received by Chinese immigrants in B C were provided by Chinese-speaking practitioners (20). The question that remains to be answered is whether these providers have distinct practice patterns that may be related to their patients' use of mental health care. The current study aims to study the health service utilization rates and patterns of recent Chinese immigrants diagnosed with severe and persistent mental illness and to explore the relationship of physician characteristics - more specifically, the ability of the physician to speak Chinese - to the utilization. Language skill rather than ethnicity is 96 Chapter 4 chosen because the ability to communicate is considered to be more important in the delivery of care. The research questions for the study are: 1. Do Chinese immigrants identified with severe and persistent mental illnesses use the same level of care as a comparison group with similar diagnoses? What are the differences in utilization if any? 2. Do physicians who speak Chinese have different practice patterns than those who do not? 3. How does the language skill of the physicians providing care affect mental health service utilization among immigrants? METHODS Data Sources This study was a secondary analysis of linked administrative databases. Three data systems provided the data for this study: Landed Immigrant Data System (LIDS) The database from Citizenship and Irnmigration Canada (CIC) consists of individual information on all immigrants who landed in B C between 1985 and 2000. B C Linked Health Data (BCLHD) The B C L H D is a linked database of most administrative health records in the province. Databases used in this study were the Medical Services Plan (MSP) Registry, MSP Payments, Hospital Separations, Vital Statistics-Death, Mental Health Services and Pharmacare files. MSP Registry: The MSP is the publicly-funded provincial health insurance plan which covers all B C residents. In this study, registration in the health plan was used to determine the.period of participation. Vital Statistics Death: The date of death was used to determine the. end point of participation for subjects who died. MSP Payments: The MSP Payments file consists of all claims made by health practitioners for services insured by the health plan. Since most general practitioners in B C are compensated on a fee for service basis, the Payments file contains records of most services received by B C residents from general practitioners (21). In contrast, only 70% 97 Chapter 4 of services provided by psychiatrists are compensated this way; the remainder of psychiatric services are paid through salary or sessional arrangements and their utilization is covered in the Mental Health Services database. Hospital Separations: The Hospital Separations file contains records of all separations from acute care and extended care facilities in B C , including day procedures. Mental Health Services: The Mental Health Services file contains records of utilization of BC ' s community mental health system. Each entry represents an episode of care in one of the system's inpatient or outpatient programs and may include numerous contacts over a lengthy period of time. Information on individual contacts is not currently available in this database. Diagnostic information is also unavailable for many of the entries. Pharmacare: The Pharmacare file encompasses detailed information on all claims for medications dispensed outside of hospitals and which are covered by the publicly-funded Pharmacare program. Only certain population groups in B C receive comprehensive coverage of their medications through one of the Pharmacare plans. During the study period, they included people aged 65 or over (Plan A), residents of long-term care facilities (Plan B), recipients of provincial income assistance (Plan C) and severely handicapped children in a community-based program (Plan F). In addition, patients with financial need also receive psychiatric medications for free when they register at community mental health services (Plan G). B C residents not in one of these plans are in Plan E and they receive coverage for medications only when their expenses exceed a pre-determined threshold. Information about the number of days of medication dispensed is available from 1996 onward. College of Physicians and Surgeons of BC (CPSBC) Register The public register of physicians in B C from 1985-2001 was used to identify the ability of physicians to speak Chinese. Data available from this file include name, specialty, place of graduation and self-reported language skills. f ' As part of a pan-Canadian immigrant health research project, the first two databases— the LIDS and B C L H D - were linked by an independent health services research centre, using probabilistic linkage methods (22). The list of Chinese-speaking 98 Chapter 4 physicians was generated based on the surname, place of graduation and self-reported language skills in the CPSBC register, supplemented by a telephone verification procedure for the purposes of this study. The list was then linked to the M S P Payments file using each physician's unique billing number in order to identify claims submitted by Chinese-speaking physicians. Identifiable personal information was removed before the linked data were released for analyses. Study Design This study adopted an observational retrospective cohort design. Utilization data for each subject was extracted from the B C L H D for the years in the study period when the individual was eligible for the study. Study period for physician visits, hospitalization and outpatient mental health care was from 1992 to 2001; data on psychiatric medications were available only from 1996 to 2001. Subjects were included in the analyses from the first year they received a diagnosis of a severe mental illness. Subjects Subjects were Chinese immigrants and comparison group members diagnosed with a severe and persistent illness. Two steps were involved in selecting subjects for this study. In the first step, immigrants from China, Taiwan, Hong Kong or Macau who were registered in the provincial health plan during the study period of 1992-2001 were identified from the immigration database and the health plan registry. Each linked immigrant was matched by sex, year of birth and area of residence to a comparison subject who was randomly selected from BC's health plan registry, excluding the immigration database. Subjects with severe and persistent mental illness were then identified from these two immigrant and comparison groups according to the two criteria below (23). I. Severe Mental Illness - any one of the following: a) Two or more physician visits for a severe mental health condition in one year b) Any hospital discharge where the most responsible diagnosis was a severe mental health condition or the principal physician was a psychiatrist or the principal service was psychiatry or patient received electroconvulsive therapy 99 Chapter 4 c) Any mental health care activation where one of the diagnoses was a severe mental health condition in one year d) Two or more prescriptions for medications used to treat severe mental health conditions in one year II. Persistent Mental Illness - meeting any one of the above criteria in at least two years during the study period The criteria were chosen to maximize specificity in identifying subjects with severe and persistent mental illness. It was recognized that the criteria would exclude individuals who were resident in B C for under two years, those who did not seek medical care and those who were treated with medications which had multiple indications for use (e.g. anti-convulsants). However, for the purpose of the study, which was to examine the utilization of subjects known to have severe and persistent illness, high specificity was preferred over sensitivity. The diagnostic categories indicating severe mental illness and medications used as treatment are listed in Appendix 4.1. In this report, Chinese immigrants or comparison subjects identified with severe and persistent mental illness refer to those who received one of these diagnoses or prescriptions. Study Variables Rates of Utilization There were six outcome variables for the analyses on rates of utilization. Five of the variables involved all subjects: mental health visits to general practitioners, non-mental health visits to general practitioners, visits to psychiatrists, mental health hospitalizations and length of outpatient mental health care. The sixth outcome, amount of psychiatric medications claimed, applied only to the subset of subjects receiving full coverage of such medications (Pharmacare Plans A,B,C,F,G) for any of the years under study. The variables are described below. 100 Chapter 4 Outcome Variable Description Unit of Measurement Mental health visit to general practitioner Non-mental health visit to general practitioner Mental health hospitalization A l l the inpatient and outpatient services Number of events for which a general practitioner submitted payment claims for a patient in one day where at least one of the diagnoses or service items was related to mental health A l l the inpatient and outpatient services Number of events for which a general practitioner submitted ( payment claims for a patient in one day where neither the diagnoses nor service items were related to mental health Visit to A l l inpatient and outpatient services for Number of events psychiatrist which a psychiatrist submitted claims for a patient in one day (1) An inpatient episode in the Mental Health Services system or (2) A discharge record in the Hospital Separations database which met one of the following criteria: (a) the most responsible diagnosis was psychiatric, or (b) the patient service code was psychiatry, or (c) the most responsible physician specialty was psychiatry, or (d) electroconvulsive therapy was performed during the stay Number of events Outpatient mental health care Activation in the Mental Health Services Number of days of system where the location of the care was activation an outpatient setting 101 Chapter 4 Outcome Variable Description Unit of Measurement Psychiatric Claims for medications dispensed and Number of days of medications frequently used for treatment of medications dispensed psychiatric disorders, including anti-psychotics, anti-depressants, anxiolytics, mood stabilizers and anti-convulsants. With the exception of non-mental health general practitioner visits, two rates were calculated and compared for each type of utilization: rates for severe mental health conditions only and rates including all mental health conditions. Only utilization data and participation time which occurred in the year and after a subject was first identified as suffering from a severe mental disorder were included. Any Diagnosis of Selected Conditions Eight dichotomous variables were created to indicate if each subject was ever diagnosed with each of the eight most common diagnostic categories in physician visits during their participation in the study. These eight conditions were: senile and presenile psychotic conditions, schizophrenic psychoses, affective psychoses, nonorganic psychotic conditions, neurotic disorders, drug dependence, depressive disorder not elsewhere classified and anxiety/depression. The first four conditions were part of the criteria for severe mental illness used in selection of subjects in this study. To be considered as ever being diagnosed, a subject would have to have at least one event (physician visit, hospitalization or mental health care) where they received such diagnosis. Proportion of Mental Health Visits to Chinese-speaking Physicians The proportion of mental health visits made to Chinese-speaking physicians was an individual level variable derived by dividing the number of mental health visits made to general practitioners and psychiatrists identified as Chinese-speaking by the total number of mental health visits made to all general practitioners and psychiatrists. The value of the variable ranged from 0 to 1; a value of 0 meant an individual received none of his/her care from Chinese-speaking physicians while a value of 1 indicated that mental 102 Chapter 4 health care was provided exclusively by Chinese-speaking physicians. Separate proportions were also calculated for the two types of physicians. Data Analysis The rates of all service utilization outcomes in the immigrant and comparison groups were estimated as the total number of outcome units divided by the total time registered in the health plan registry; the rates were directly standardized by sex and age to BC's 1996 census population.4 Separate age-standardized estimations were also done for each sex. Difference between immigrants and the comparison group for each outcome was measured by a rate ratio, obtained by dividing the rate of the immigrant group by the rate of the comparison group. The 95% confidence interval for each rate and rate ratio was determined by bootstrapping (24).5 Frequency distributions were tabulated for the diagnostic categories of all mental health visits to general practitioners and psychiatrists, the most responsible diagnostic categories of all mental health hospitalizations and diagnostic categories of all mental health visits by the language skill of the physician claiming the service. The percentage of all mental health visits claimed by Chinese-speaking physicians for immigrants and comparison group members and the mean proportion of such visits were also calculated. The 95% confidence limits for all the above descriptive statistics were computed by treating multiple visits by an individual as a cluster. Logistic regression modelling was used to estimate the odds ratio between immigrant and comparison subjects of ever receiving a diagnosis of each of the eight selected conditions. Two logistic regression models were run for each diagnosis: one adjusting for sex and age and another adjusting for sex, age and the proportion of visits to Chinese-speaking physicians. Two additional logistic regression models for each selected condition estimated the effect of the proportion of visits to Chinese-speaking 4 Rates for physician visits, hospitalization and mental health care were standardized by ten-year age groups from age 15 to 74 and a group for age 75 and above. The same age groups were used for psychiatric medication utilization but, owing to the small number of subjects, the age groups 45-54 and 55-64 were combined. 5 Bootstrapping computes the variance of an estimated parameter by repeatedly resampling the original data with replacement and re-fitting the model to generate a bootstrap distribution of the estimated parameter. The bootstrap distribution approximates the sampling distribution. 103 Chapter 4 physicians on the odds of ever receiving the diagnosis in the immigrant and comparison groups respectively, adjusting for sex and age at entry to study. RESULTS Subject Selection Only subjects aged 15 or over were included in the analyses because few individuals under 15 (13 immigrants and 117 comparison subjects) met the criteria for the study. Out of over 150,000 Chinese immigrants who landed in B C between 1985 and 2000 and the same number of comparison subjects, 786 immigrants and 3962 comparison subjects aged 15 or over met the criteria for severe and persistent mental illness and were registered in the health plan during 1992-2001. Among this latter cohort, 466 immigrants and 2673 comparison subjects were enrolled in drug plans which fully covered psychiatric medications for some time between 1996 and 2001. Tables 4.1a and 4.1b summarize their demographic characteristics. For both immigrants and the comparison group, there were more women than men among those meeting the criteria for severe and persistent mental illness, although the difference was greater among the immigrants. Immigrant women were also older than comparison women. Immigrants were more likely than the comparison group to live in Vancouver and Richmond but less likely to live in North Shore, Fraser South and Other regions of B C . There were no differences between immigrants and comparison group members in the mean socio-economic quintile of their neighbourhoods or their participation time in the study. The immigrants came predominantly from Hong Kong and most arrived before 1995. The subgroup of subjects who were on Pharmacare plans that fully covered psychiatric medications had similar characteristics as the original cohort except that they were older, because seniors aged 65 or over were automatically eligible for full drug benefits at the time of the study. The more pronounced age effect among immigrants reflected that younger immigrants were less likely to enrol in full coverage plans, especially the plan for income assistance recipients (Plan C), since economic immigrants were selected on the basis of their financial independence and sponsored immigrants were ineligible for social assistance for ten years. 104 Chapter 4 Rates of Mental Health Service Utilization Table 4.2 summarizes the age and sex adjusted rates and rate ratios between Chinese immigrants and comparison subjects for utilization of five types of mental health services, as well as for non-mental health visits to general practitioners. For mental health services, rates are reported for both utilization for severe disorders and utilization for all conditions. When considering only services for severe mental disorders, Chinese immigrants had higher rates of visits to psychiatrists than the comparison group but lower rates in all other types of services. Differences between immigrants and comparison subjects were generally larger among men than women, except in the case of psychiatric visits where immigrant women had 50% more visits than comparison women while there was no significant difference between men. The results for outpatient mental health care for severe mental disorders are shown only for information and must be interpreted with caution since the majority of care episodes in the database lacked a diagnosis. The percentage of unclear diagnosis was higher in the immigrant group than the comparison group (76% vs 52%). When services for all mental health conditions were considered, immigrants had lower rates of utilization than the comparison group in all types of services, including psychiatric visits which barely reached statistical significance at the 5% level. The discrepancy between immigrants and the comparison group was most remarkable in mental health consultations with general practitioners and mental health hospitalizations where immigrants had less than half the rate of utilization of the comparison group. Chinese immigrants identified with severe and persistent mental illness also had a lower rate of visits to general practitioners for non-mental health reasons. Including services for less serious conditions with services for severe mental disorders magnified the difference between immigrant and comparison subjects for visits to general practitioners but reversed the direction of the difference in visits to psychiatrists. The difference in mental health hospitalizations did not change and the difference in psychiatric medications increased slightly. 105 Chapter 4 Reasons for Mental Health Service Utilization The results from the rates of utilization reveal that a large proportion of mental health service utilization is identified with less serious mental disorders, even among a severely mentally i l l cohort. It is confirmed by the frequency distributions of the diagnostic categories in physician visits and in hospitalizations. Table 4.3a lists the distribution of diagnoses reported in mental health visits to general practitioners or psychiatrists made by the study cohort. Of all mental health consultations by Chinese immigrants, 62% were for a severe psychiatric illness; the percentage for the comparison group was lower at 34%. Affective psychoses and schizophrenic psychoses were the main reasons for consultations with physicians by severely i l l immigrants, accounting for almost half of all their mental health visits. However, among members of the comparison group who were also diagnosed with a severe mental illness, the most common reason for receiving mental health care was depressive disorder. Severe mental illnesses were responsible for a higher percentage of hospitalizations, as shown in Table 4.3b. The primary diagnoses for 77% of immigrant hospitalizations and 63% of the comparison group's hospitalizations were severe mental disorders. Schizophrenic psychoses and affective psychoses were the most common reasons for hospital care in both groups. Chinese immigrants were more likely than the comparison group to be hospitalized for psychotic disorders while the comparison group were more likely than immigrants to be hospitalized for depressive and other disorders. Diagnoses of Selected Conditions Results thus far indicate that comparison subjects diagnosed with severe and persistent mental illness were more likely than their immigrant counterparts to use services for less serious conditions. This is corroborated by results of their ever receiving a diagnosis of eight conditions, shown in Table 4.4. There were no differences between the immigrant and comparison groups in the percentages of their respective cohorts being diagnosed with schizophrenic psychoses. Immigrants, however, were less likely than comparison subjects to be diagnosed with affective psychoses (46% vs 53%) and senile/presenile psychoses (8% vs 11%) but more likely to be diagnosed with other nonorganic psychoses (20% vs 16%). For the four less serious conditions which were not 106 Chapter 4 part of the selection criteria of the severely mentally i l l cohorts, no difference was observed between immigrants and comparison subjects in the percentages diagnosed with anxiety/depression. There were relatively fewer immigrants than comparison subjects ever receiving the diagnoses of depressive disorder not elsewhere classified (41% vs 64%), neurotic disorders (39% vs 53%) or drug dependence (1% vs 13%). The first column of odds ratios (OR) in Table 4.4 presents the contrast between immigrants and comparison subjects adjusted for sex and age differences. After adjusting for sex and age, immigrants were more likely than the comparison group to be diagnosed with other nonorganic psychoses (OR=1.35) but were less likely to receive diagnosis of senile and presenile psychoses (OR=0.31) or affective psychoses (OR=0.78). The two cohorts did not differ in their diagnosis of schizopnrenic psychoses. With the exception of anxiety/depression, imrnigrants had consistently lower odds ratios of diagnosis in the other less serious conditions studied (OR=0.07 to 0.55). Physician Factors in Service Utilization and Diagnoses , • The next phase of the analysis examined the contribution of a physician characteristic - the ability of the physician to speak Chinese - to utilization patterns of the immigrant and comparison groups. Results in Table 4.5 confirm that Chinese immigrants diagnosed with severe and persistent mental illness were more likely to receive mental health care from physicians who speak their native language. Overall, 76% of mental health consultations received by Chinese immigrants with severe and persistent mental illness were provided by Chinese-speaking general practitioners and psychiatrists; the corresponding figure for the comparison group was under 10%. On an individual level, the average proportion of mental health services received from Chinese-speaking physicians for Chinese immigrants was 0.81 with 56% of the irrimigrants receiving their care exclusively from physicians who shared their language background. The average proportion for the comparison group was 0.12 and less than 5% received all their care from Chinese-speaking physicians. These results illustrate that mental health care for the immigrant and comparison groups were provided by two almost distinct clusters of physicians differentiated by Chinese language skill. 107 Chapter 4 Table 4.6 presents the diagnostic categories of mental health services claimed in relation to the severely and persistently mentally i l l cohorts by general practitioners and psychiatrists, according to their language skill. Among both immigrants and comparison subjects, Chinese-speaking physicians were more likely than their non-Chinese-speaking colleagues to treat severe mental disorders, although the difference was not statistically significant among immigrants. Among both immigrants and comparison subjects, Chinese-speaking physicians were more likely than non-Chinese-speaking physicians to treat affective psychoses and nonorganic psychoses but less likely to treat depressive disorder not elsewhere classified. Among immigrants, non-Chinese-speaking physicians were treating relatively more schizophrenic psychoses than Chinese-speaking physicians (31% vs 20%, not significant) whereas the reverse was true in the comparison group (11% vs20%). Table 4.7 reports the effect of the proportion of mental health visits to Chinese-speaking physicians on the odds of ever receiving a diagnosis of each of the eight selected conditions, adjusting for sex and age. Among immigrants, receiving more of one's care from Chinese-speaking physicians was associated with higher odds of being diagnosed with anxiety/depression (OR=2.01) but lower odds of being diagnosed with schizophrenic psychoses (OR=0.34), senile and presenile psychoses (OR=0.11), other nonorganic psychoses (OR=0.47), depressive disorder (OR=0.24) and drug dependence (OR=0.10). The odds ratios indicated that those who received mental health care exclusively from Chinese-speaking physicians were twice as likely as those who visited exclusively non-Chinese-speaking physicians to be diagnosed with anxiety/depression but less than half as likely to be diagnosed with most of the other conditions. In the comparison group, receiving care from Chinese-speaking physicians was associated with increased likelihood of diagnosis of schizophrenic psychoses (OR=1.78) but decreased likelihood of diagnoses of affective psychoses (OR=0.62), depressive disorder (OR=0.41), neurotic disorders (OR=0.60) and drug dependence (OR=0.32). The preponderance of visits to Chinese-speaking physicians by immigrants was a plausible confounder of the differences in odds of diagnosis of the selected conditions. To ascertain the difference between immigrants and the comparison group in the odds of diagnosis, analysis was done adjusting for sex, age and proportion of visits to Chinese-108 Chapter4 speaking physicians. The results are shown in the second column of odds ratios in Table 4.4. For most of the conditions, the discrepancy between immigrants and comparison subjects was reduced after adding the physician factor to the model. There was no statistically significant difference between immigrants and comparison subjects in the odds of being diagnosed with schizophrenic psychoses, affective psychoses and anxiety/depression. Immigrants continued to have higher odds of being diagnosed with nonorganic psychoses (OR=1.44) but lower odds of diagnoses of senile/presenile psychoses (OR=0.56) and three of the less serious conditions (OR=0.12 to 0.74). DISCUSSION With respect to the subject of utilization of mental health services by Chinese immigrants with severe and persistent mental illness, findings in this study partially support other reports on rates of utilization. Chinese immigrants are under-represented in the health care system even when serious mental illnesses are involved. Out of the original population containing similar numbers of immigrant and comparison group members, only 20% as many immigrants as comparison subjects met the criteria for severe and persistent mental illness. Severely i l l Chinese immigrants were also more likely than comparison subjects to be treated for the severe disorders, rather than less serious comorbid conditions. In contrast to other studies, however, the current study found that Chinese immigrants used fewer services even after being diagnosed with a serious mental illness; the disparity in utilization was especially pronounced for less serious comorbid conditions. The only exception was that Chinese immigrants had more psychiatric consultations for the severe disorders. The comparison group's lower utilization of psychiatric services might have been compensated by their relatively higher utilization of general practice and of outpatient mental health care, where many psychiatrists practised under alternative payment schemes. This study also offers evidence that the Chinese immigrants diagnosed with severe mental illness are less likely than comparison subjects to be diagnosed with comorbidities of depression, neurotic disorders and drug dependence. Since these individuals are already receiving care from mental health professionals, it is tempting to assume that any comorbid conditions would be promptly identified and to conclude that 109 Chapter 4 Chinese immigrants with severe and persistent mental illness indeed have lower rates of such comorbidities. However, other findings from this study suggest that lower prevalence of less serious conditions is not solely responsible for lower rates of diagnosis and utilization, and that the language skill of the physicians providing care also has an effect. This study found that whether the physicians were able to communicate in Chinese was related to the type of conditions they treated and to the range of diagnoses the patients received in this cohort of severely i l l immigrants and comparison subjects. Physicians who spoke Chinese were more inclined to treat psychoses but less disposed to treat depression and drug dependence. Among both immigrants and comparison subjects, receiving more care from a Chinese-speaking physician was associated with lower probability of being diagnosed with depression and drug dependence. The disparity between the immigrant and comparison groups in odds of diagnosis was partially explained by the proportion of visits to Chinese-speaking physicians. While there are many reasons Chinese immigrants may be receiving fewer services for less serious disorders, this study demonstrates that physician factors also play an important role independent of the ethnicity and immigration status of the patients. Findings from this study underscore some intriguing aspects about mental health services to Chinese immigrants. With respect to treatment for severe mental disorders, Chinese immigrants receive more psychiatric consultations but fewer other services including hospitalization, outpatient mental health care and medications. These findings indicate that access to care, even for the most severe forms of illness, varies according to the type of service. It appears that severely i l l Chinese immigrants compensate for their lower utilization of community mental health and general practice with more visits to fee for service psychiatrists; however, these compensatory efforts do not translate into hospitalizations or prescriptions of medications. Furthermore, when all mental health conditions are considered, even among those receiving treatment for severe illness, Chinese immigrants lag behind the comparison group in utilization of all types of services. In addition, this study draws attention to systematic differences in the practice behaviour of physicians who provide most of the Chinese immigrants' care. Physicians 110 Chapter 4 who speak Chinese tend to treat the more serious disorders rather than the less serious kinds; those who receive their care primarily from these physicians are less likely to be diagnosed with comorbidities of less serious mental disorders. There may be several explanations for this outcome. Chinese-speaking physicians may have developed specialized skills in certain mental disorders and therefore tend to attract referrals for those conditions. There may be differences in coding practice in that Chinese-speaking physicians are predisposed to identify psychoses as the primary problem in submitting a payment claim. They may also be under-diagnosing the less serious conditions, owing to their own cultural background or their training. The potential influence of these two correlates should not be overlooked since almost all of these physicians are ethnically Chinese and 40% of them graduated outside North America. 6 With respect to the delivery of mental health care to Chinese immigrants, the implications of differences in physician practice patterns depend on the reasons for these differences. If Chinese immigrants are drawn to Chinese-speaking physicians because their mental health problems match the specific skills of the practitioners and that they are referred to other non-Chinese-speaking providers when comorbidities are identified, there is no concern about the equity of access. On the other hand, if certain conditions are systematically under-diagnosed and under-treated in practices where most Chinese immigrants congregate for care, the implications can be enormous. The under-diagnosis and under-treatment constitute systemic barriers to adequate services and invariably lead to poor treatment outcome and increased burden of illness. As subjects in this study are selected based on diagnosis of one of the severe conditions, it is not known the extent to which proportion of consultations with Chinese-speaking physicians also affected the rate of diagnosis of these disorders. The consequences are even graver if Chinese immigrants with these most debilitating conditions are missed by the health care system. While the markedly smaller number of severely i l l immigrants identified for the study alleges to this possibility, the current analyses do not allow conclusions to be drawn in this regard. Further research will have to clarify the reasons for differences in practice patterns and the impact. Examining the referral patterns between Chinese-speaking and non-Chinese-6 College of Physicians and Surgeons of BC register 111 Chapter 4 speaking practitioners, for example, will shed light on the treatment path of patients with different diagnoses. . . , In the discourse on disparities in mental health services, the focus is often on financial, language and cultural barriers. In Canada, where insurance coverage is not contingent upon financial status, language interpretation and linguistic match between provider and patient become prominent strategies to reduce disparities. This study has introduced providers' behaviour as another systemic factor that deserves attention. In addition to recruiting bilingual and bicultural professionals, efforts to reduce disparities must consider the systematic differences in the practice pattern of health professionals whom minorities rely on for care. Understanding these differences will enhance the effectiveness of interventions to ensure quality of health care and advance the system towards the goal of equity. There are several limitations to this study. First of all, the use of administrative databases precludes knowledge about individuals who did not receive health care. As discussed earlier, since the definition of severe mental illness is based on the utilization data, the current study is not able to establish if the prevalence of such conditions is indeed five times lower in the Chinese immigrant population or the extent to which seriously i l l immigrants are under-utilizing health care and under-diagnosed. Secondly, the diagnostic information in the databases has not been validated. If practitioners who provide care to immigrants are systematically inclined to favour submitting one diagnosis over another, the results on practice patterns of the providers and diagnostic frequencies of the subjects may be biased. At present, there is no evidence such systematic differences in coding behaviour exist. Validating the diagnostic information by chart reviews or reassessment of patients is a labour intensive task and involves complex ethical and privacy issues, but it is a project worth undertaking to evaluate the usefulness of administrative databases for epidemiological studies. Of central concern to health services research is quality of care and appropriateness of services. The diagnosis submitted with the payment claim does not necessarily reflect the intervention provided; hence, cross-validation between physician claims and, for instance, Pharmacare records will also supply additional information about under-diagnosis and under-treatment. 112 Chapter 4 A third limitation is that only Chinese language skill is investigated as a physician characteristic that may influence the diagnosis and treatment of mental health disorders.., Inclusion of other physician variables would have provided a more comprehensive understanding of the role of physician factors.^  Ethnicity and place of training are two variables that are strongly correlated with Chinese language skill and may likely explain the differences observed in practice pattern. There may also conceivably be other systematic differences in the characteristics of physicians who provide care to Chinese immigrants, such as age, sex or location of practice, and which also affect their practice behaviour. A final caveat of this study is that two other variables can potentially be misclassified and affect the results. Some of the comparison subjects in the study cohort may in fact be Chinese immigrants who are not linked to the immigration database. Conversely, not all the Chinese-speaking physicians are identified. The implication of the second misclassification is that the actual proportion of visits to Chinese-speaking physicians can be higher than reported. For most of the analyses in this study, misclassification of either of these variables attenuates the actual differences and does not affect the findings observed in this study. However, with respect to the effect of the proportion of visits to Chinese-speaking physicians on the odds ratios of receiving specific diagnoses, the results observed in the comparison group may be confounded by the misclassification of Chinese immigrants. Since the "unidentified" Chinese immigrants in the comparison group are more likely to be found among those with higher proportion of visits to Chinese-speaking physicians, the effect observed may reflect the effect of ethnicity and immigration status as well as provider preference. On the other hand, based on census data and a federal immigration report, it is estimated that the percentage of misclassified Chinese immigrants in the overall comparison group from which subjects are drawn is approximately 2.4%. Since Chinese immigrants are only 20% as likely as comparison subjects to be identified as suffering from a severe and persistent mental illness, the actual percentage of misclassified immigrants in this study would be even lower and unlikely to bias the results in any significant way. 113 Chapter 4 CONCLUSION This study offers evidence of disparities in mental health service utilization with respect to Chinese immigrants and of systematic differences in the practice patterns of physicians who provide most of their care. Even among those diagnosed with severe and persistent mental illness, Chinese immigrants use fewer mental health services than the comparison subjects and are also less likely to be diagnosed with comorbidities of less serious mental disorders. At the same time, physicians who speak Chinese and who provide the majority of care to Chinese immigrants tend to treat serious mental disorders rather than the less serious disorders. The proportion of care immigrants and comparison subjects received from Chinese-speaking physicians is associated with diagnoses of depression and drug dependence. These results suggest that, not only are concerns about equity of mental health care to severely i l l immigrants justified, systemic factors such as physician characteristics may contribute to the disparities. Further research is necessary to clarify the reasons behind the disparities and the role that physician factors play. 114 Table 4.1a Characteristics of Immigrant and Comparison Subjects Diagnosed with Severe and Persistent Mental Illness and Aged 15 or Over in 1992-2001 Immigrant (N=786) Sex N % 95 % CI Female 509 64.76% (61.41%, 68.11%) Male 277 35.24% (31.89%, 38.59%) Female Male 95 % CI 95 % CI Mean Age at Entry to Study (Years) 51.27 (49.50,53.05) 45.91 (43.41 ,48.41) Mean Years in Study 5.28 (5.08,5.49) 5.20 (4.89,5.51) Mean Socio-economic Quintile 2.64 (2.53,2.75) 2.72 (2.57,2.87) Health Service Delivery Area Vancouver 53.63% (49.29%, 57.98%) 51.99% (46.09%, 57.88%) Richmond 21.22% (17.66%, 24.78%) 19.49% (14.82%, 24.17%) North Shore/Coast Garibaldi 1.38% (0.36%, 2.39%) 2.17% (0.45%, 3.88%) Fraser North 18.27% (14.91%, 21.64%) 19.49% (14.82%, 24.17%) Fraser South 3.54% (1.93%, 5.14%) 2.89% (0.91%, 4.86%) South Vancouver Island 1.57% (0.49%, 2.65%) 2.89% (0.91%, 4.86%) Other 0.39% (0.00%, 0.94%) 1.08% (0.00%, 2.30%) Year of Landing 1985-1988 15.13% (12.01%, 18.25%) 14.80% (10.61%, 18.99%) 1989-1991 23.38% (19.69%, 27.06%) 25.63% (20.48%, 30.78%) 1992-1994 42.63% (38.33%, 46.94%) 36.10% (30.43%, 41.77%) 1995-1997 17.49% (14.18%,20.79%) 20.94% (16.14%,25.74%) 1998-2000 1.38% (0.36%, 2.39%) 2.53% (0.67%, 4.38%) Place of Origin Hong Kong 70.73% (66.77%, 74.69%) 63.54% (57.86%, 69.22%) China 22.00% (18.40%, 25.61%) 22.74% (17.80%, 27.69%) Taiwan 6.68% (4.51%, 8.85%) 13.36% (9.34%: 17.37%) Macau 0.59% (0.00%, 1.26%) 0.36% (0.00%, 1.07%) Comparison (N=3962) H °A 95 % CI 2231 56.31% (54.76%, 57.86%) 1731 43.69% (42.14%, 45.24%) Female Male 95 % CI 95 % CI 45.98 (45.27,46.70) 45.67 (44.89,46.46) 5.12 (5.02,5.23) 4.96 (4.85,5.08) 2.61 (2.56,2.66) 2.54 (2.48,2.60) 35.50% (33.51%, 37.49%) 41.94% (39.62%, 44.27%) 13.54% (12.12%, 14.96%) 12.59% (11.03% ,14.16%) 4.03% (3.22%, 4.85%) 3.76% (2.86%, 4.65%) 22.28% (20.55%, 24.00%) 21.20% (19.28%, 23.13%) 11.97% (10.62%, 13.32%) 9.24% (7.88%, 10.61%) 3.00% (2.29%, 3.71%) 3.18% (2.35%, 4.00%) 9.68% (8.45%, 10.91%) 8.09% (6.80%, 9.37%) Table 4.1b O Characteristics of Immigrant and Comparison Subjects Diagnosed with Severe and Persistent Mental Illness and on Pharmacare Plans ABCFG in 1996-2001 p Immigrant (N=461) Comparison (N=2617) Sex N % 95 % CI N % 95 % CI Female 306 66.38% (62.05%, 70.71%) 1436 54.87% (52.96%, 56.78%) Male 155 33.62% (29.29%, 37.95%) 1181 45.13% (43.22%, 47.04%) Female Male Female Male 95% CI 95 % CI 95 % CI 95 % CI Mean Age at Entry to Study (Years) 59.46 (57.23,61.69) 53.57 (49.92,57.22) 48.88 (47.90,49.85) 46.77 (45.76,47.78) Mean Years in Study 3.92 (3.72,4.11) 3.38 (3.10,3.66) 3.69 (3.60,3.79) 3.69 (3.59,3.79) Mean Socio-economic Quintile 2.53 (2.39,2.67) 2.52 (2.32,2.72) 2.42 (2.36,2.48) 2.39 (2.32,2.46) Health Service Delivery Area Vancouver 57.52% (51.96%, 63.08%) 56.77% (48.95%, 64.60%) 38.16% (35.65%, 40.68%) 44.96% (42.12%, 47.80%) Richmond 18.95% (14.55%, 23.36%) 21.29% (14.82%, 27.76%) . 12.40% (10.69%, 14.10%) 12.36% (10.48%, 14.24%) North Shore/Coast Garibaldi 1.31% (0.03%, 2.58%) 2.58% (0.08%, 5.09%) 3.06% (2.17%, 3.96%) 3.05% (2.07%, 4.03%) Fraser North 17.32% (13.06%, 21.58%) 15.48% (9.77%, 21.20%) 22.35% (20.20%, 24.51%) 19.05% (16.81%, 21.29%) Fraser South 3.92% (1.74%, 6.10%) 0.65% (0.00%, 1.91%) 11.21% (9.58%, 12.84%) 8.89% (7.27%, 10.52%) South Vancouver Island 0.98% (0.00%, 2.09%) 1.94% (0.00%, 4.11%) 3.06% (2.17%, 3.96%) 3.73% (2.64%, 4.81%) Other 0% 1.29% (0.00%, 3.07%) 9.75% (8.21%, 11.28%) 7.96% (6.41%, 9.50%) Year of Landing 1985-1988 17.97% (13.66%, 22.29%) 19.35% (13.11%, 25.60%) 1989-1991 24.51% (19.67%, 29.35%) 25.16% (18.30%, 32.02%) 1992-1994 41.18% (35.64%, 46.71%) 34.84% (27.31%, 42.37%) 1995-1997 14.71% (10.72%, 18.69%) 18.06% (11.99%, 24.14%) 1998-2000 1.63% (0.21%, 3.06%) 2.58% (0.08%, 5.09%) Place of Origin Hong Kong 69.28% (64.09%, 74.47%) 66.45% (58.99%, 73.91%) China 25.82% (20.90%, 30.74%) 22.58% (15.97% ,29.19%) Taiwan 4.25% (1.98%, 6.52%) 10.97% (6.03%, 15.91%) Macau 0.65% (0.00%, 1.56%) 0% Table 4.2 Rates of Mental Health Service Utilization in 1992-2001 by Immigrant and Comparison Subjects Diagnosed with Severe and Persistent Mental Illness Severe Mental Illnesses All Conditions n Immigrant Rate1 95% CI Comparison Rate1 95% CI Mental Health Physician Visits (General Practitioners + Psychiatrists) All Female Male 3.72 4.06 3.36 (3.41,4.08) (3.63,4.64) (2.88,3.84) 3.13 2.94 3.32 (2.97,3.34) (2.72,3.19) (3.06,3.61) Rate Ratio2 1.19 (1.06,1.32) 1.38 (1.21,1.62) 1.01 (0.86,1.20) Immigrant Rate1 5.98 6.32 5.64 95% CI (5.50,6.46) (5.78,7.01) (4.99 6.66) Comparison Rate1 9.00 9.12 8.88 95% CI (8.64.9.58) (8.61 ,9.70) (8.30,9.75) Rate Ratio2 95% CI 0.66 (0.61,0.73) 0.69 (0.62,0.77) 0.64 (0.55,0.75) Mental Health Visits to General Practitioners All 0.54 (0.45,0.63) 0.78 (0.71,0.86) 0.69 (0.57,0.83) 1.78 (1.56,2.15) 4.34 (4.01,4.78) 0.41 (0.35,0.50) Female 0.61 (0.49,0.76) 0.69 (0.61,0.77) 0.89 (0.70,1.16) 1.77 (1.55.2.00) 4.20 (3.86,4.66) 0.42 (0.36,0.50) Male 0.47 (0.36,0.62) 0.88 (0.78,1.03) 0.53 (0.38,0.71) 1.79 (1.43,2.44) 4.49 (3.90,5.22) 0.40 (0.30,0.54) Visits to Psychiatrists All 3.18 (2.89,3.50) 2.34 (2.17,2.50) 1.36 (1.20,1.53) 4.21 (3.83,4.56) 4.66 (4.43,4.90) 0.90 (0.81,1.00) Female 3.45 (3.02,3.93) 2.26 (2.05,2.49) 1.53 (1.30,1.81) 4.54 (4.04,5.17) 4.92 (4.57,5.32) 0.92 (0.80,1.08) Male 2.90 (2.38,3.30) 2.43 (2.21,2.70) 1.19 (0.98,1.40) 3.85 (3.25,4.37) 4.39 (4.06,4.74) 0.88 (0.74,1.03) Mental Health Hospitalizations All 0.13 (0.10,0.15) Female 0.13 (0.11, Male ,0.16) 0.12 (0.09,0.15) 0.25 0.23 0.27 (0.23,0.27) (0.20,0.25) (0.23,0.31) 0.51 (0.42.0.61) 0.59 (0.46,0.74) 0.43 (0.32,0.58) 0.13 0.14 0.12 (0.11,0.15) (0.11,0.17) (0.09,0.15) 0.27 0.25 0.29 (0.25,0.30) (0.22,0.27) (0.26,0.34) 0.47 (0.38 0.56) 0.54 (0.42.0.69) 0.41 (0.29 0.52) Outpatient Mental Health Care (Days) All 5.36 (5.36,13.22) 24.46 Female 10.04 (5.84,15.95) 25.63 Male 6.02 (2.63,15.01) 23.24 (21.92,27.05) (22.05,29.47) (19.60,27.44) 0.33 (0.21,0.55) 0.39 (0.22,0.64) 0.26 (0.11,0.61) 57.54 60.73 54.22 (48.62 66.84) (48.59,72.91) (41.77,68.96) 81.58 79.10 84.17 (77.41 (73.51 85.86) 84.56) (78.23,91.11) 0.71 0.77 0.64 (0.59,0.83) (0.61.0.93) (0.49.0.82) Psychiatric Medications (Days) 1996-20013 All 251.9 (217.2,289.6) 309.7 (296.5,323.8) 0.81 (0.70,0.94) 384.8 (343.5.427.7) 542.9 (524.8,560.4) Female 268.9 (227.2,339.0) 298.6 (281.4,319.1) 0.90 (0.75,1.16) 408.7 (349.7,478.0) 567.1 (543.7,594.7) Male 234.1 (183.2,283.0) 321.4 (301.3,340.7) 0.73 (0.57,0.90) 359.8 (302.1,420.1) 517.7 (493.2,541.0) 0.71 (0.63,0.79) 0.72 (0.62,0.85) 0.70 (0.58,0.82) Non-Mental Health Visits to General Practitioners All Female • Male 6.35 7.28 5.39 (5.96,6.79) (6.80,7.84) (4.73,6.13) 7.37 8.43 6.27 (7.11 ,7.65) (8.12,8.82) (5.91 ,6.65) 1 Rate is per person-year, male and female rates are age adjusted to the 1996 census population in B.C.; rates for all are adjusted by sex and age 2 Rate ratio in bold type indicates a result significantly different from 1 at the 5% level v 3 Estimates for psychiatric medications are based on a sub-group of subjects who received full coverage for psychiatric medications 0.86 (0.80,0.93) 0.86 (0.79,0.93) 0.86 (0.75,0.99) Table 4.3a Diagnostic Categories of Mental Health Visits to General Practitioners & Psychiatrists of Immigrant & Comparison Subjects with Severe and Persistent Mental Illness Immigrant •• Comparison #of # of #0f #pf Code Diagnostic Category Subiects1 Events % 95% CI Subiects1 Events % • 95% CI 296 Affective Psychoses 350 . 5810 25.47% (22.28% ,28.65%) 1717 30378 16.23% (14.85% 17.60%) 295 Schizophrenic Psychoses 242 5177 22.69% (19.07%, 26.31%) 1154 . .22665 12.11% (10.82% 13.39%) 311 Depressive Disorder NEC 316 2718 11.91% (9.71%, 14.12%) 2535 39283 20.98% (19.52% 22.45%) 300 Neurotic Disorders 290 2552 11.19% . (8.87% ,13.50%) 1952 22539 12.04% (10.81% 13.27%) 50B2 Anxiety/Depression 378 1990 8.72% (6.73%. 10.72%) 1961 15146 8.09% (7.33% 8.85%) 298 Other Nonorganic Psychoses 133 1431 6.27% (4.21%, 8.33%) 486 . 2214 1.18% . (0.98% 1.39%) 290 Senile & Presenile Organic Psychotic Conditions , 58 758 3.32% (1.82%, 4.82%) 354 3089 1.65% (1.31% 1.99%) 304 Drug Dependence 7 399 1.75% (0.00%, 4.41%) 446 25577 13.66% (10.77% 16.56%) 293 Transient Organic Psychotic Conditions 32 335 1.47% (0.64%, 2.30%) 182 1073 0.57% (0.43% 0.71%) 309 Adjustment Reaction 77 321 1.41% (0.81%. 2.01%) 652 4222 2.26% (1.67% 2.84%) 297 Paranoid States 74 313 1.37% (0.82%, 1.92%) 264 1227 0.66% (0.44% 0.87%) 308 Acute Reaction to Stress 75 217 0.95% (0.58%, 1.33%) 873 3211 1.72% (1.40% 2.03%) 307 Special Symptoms or Syndromes NEC 45 156 0.68% . (0.25% ,1.12%) 375 2674 1.43% ". (0.78% 2.08%) 294 Other Organic Psychotic Conditions (Chronic) 23 138 0.60% (0.18%, 1.03%) 189 1092 6.58% (0.40% 0.77%) 310 Nonpsychotic Mental Disorders - Organic Brain Damage 24 113 0.50% (0.18%, 0.81%) 242 1337 0.71% (0.54% 0.89%) Non-Psychiatric Diagnoses3 28 109 0.48% (0.00%. 1.01%) 332 1820 0.97% (0.65% 1.30%) 301 Personality Disorders 32 96 0.42% (0.20%, 0.65%) 516 - 3501 1.87% (1.47% 2.27%) 313 Disturbance of Emotions - Childhood and Adolescence 2 44 0.19% (0.00%, 0.56%) 68 95 0.05% (0.02% 0.08%) 299 Psychoses With Origin Specific to Childhood 6 39 0.17% (0.00%, 0.34%) 89 309 0.17% • (0.11% 0.22%) 292 Drug Psychoses 4 35 . 0.15% (0.00%, 0.43%) 81 175 0.09% (0.06% 0.12%) 312 Disturbance of Conduct NEC 13 19 0.08% (0.03%, 0.13%) 210 537 0.29% (0.20% 0.37%) 306 Physiological Malfunction Arising From Mental Factors 10 14 0.06% (0.02%, 0.11%) 82 237 0.13% (0.05% 0.20%) 314 Hyperkinetic Syndrome of Childhood 3 12 0.05% (0.00%,0.13%) 82 451 0.24% (0.08% 0.40%) 302 Sexual Deviations & Disorders 6 7 0.03% (0.01%, 0.06%) 58 190 0.10% (0.04% 0.16%) 303 Alcohol Dependence Syndrome 6 7 0.03% (0.01%, 0.06%) 328 1996 1.07% (0.80% 1.34%) 305 Nondependent Abuse of Drugs 3 3 0.01% (0.00%, 0.03%) 134 1939 1.04% (0.37% 1.70%) 291 Alcoholic Psychoses 1 . 1 0.00% (0.00%, 0.01%) 60 228 0.12% (0.06% 0.18%) 04A2 General Psychiatric Examination 1 1 0.00% (0.00%, 0.01%) 9 9 0.00% (0.00% .0.01%) Subtotal for Severe Mental Illnesses4 573 14150 62.02% (58.00%, 66.04%) 2778 63787 34.07% (32.04% .36.10%) TOTAL5 725 22815 100% 3785 187214 100% 11ndividual subjects may be treated for more than one diagnosis. The total and sub-total number of subjects do not equal the sum of subjects for the diagnostic categories. 2 MSP diagnostic codes; all the others are ICD-9 codes 3 Includes other conditions of brain and the nervous system (348-349) and general symptoms (780) 4 Includes ICD-9 diagnostic codes 290-299 and 310 5 1 3 missing diagnoses among comparison subjects were not included in the total count Table 4.3b Most Responsible Diagnostic Categories of Mental Health Hospitalizations of Immigrant & Comparison Subjects with Severe and Persistent Mental Illness Immigrant ; Comparison #of #of % 95% CI #0f # 0 i % 95% CI Code Diagnostic Category " Subjects1 Events Subjects' Events 295 Schizophrenic Psychoses 77 144 3470% (27.33% 42.07%) 351 930 20.40% (17.65%.23.16%) 296 Affective Psychoses 53 100 24.10% (17.44% 30.75%) 552 1237 27.14% (24.25% ,30.03%) 298 Other Nonorganic Psychoses 31 36 8.67% (5.56% 11.79%) 159 221 4.85% (3.98% ,5.72%) Non-Psychiatric Diagnoses2 23 25 6.02% (3.37% 8.68%) 235 265 5.81% (4.94%, 6.69%) 297 Paranoid States 12 23 5.54% (1.53% 9.56%) 53 60 1.32% (0.93%, 1.70%) 311 Depressive Disorder NEC 6 19 4.58% (0.00% 11:02%) 133 416 9.13% (4.02%, 14.23%) 300 Neurotic Disorders 15 17 4.10% (2.01% 6.18%) 137 194 4.26% (3.44%, 5.08%) 301 Personality Disorders 12 15 3.61% (1.50% 5.73%) 128 222 4.87% (3.73%, 6.01%) 309 Adjustment Reaction 11 12 2.89% (1.12% 4.66%) 177 233 5.11% (4.20% 6.03%) 290 Senile & Presenile Organic Psychotic Conditions 9 11 2.65% (0.80% 4.51%) 88 101 2.22% (1.69%, 2.74%) 307 Special Symptoms or Syndromes NEC 2 4 0.96% (0.00% 2.47%) 36 81 1.78% (0.83%, 2.72%) 293 Transient Organic Psychotic Conditions 3 3 0.72% (0.00% 1.55%) 52 58 1.27% (0.90%, 1.64%) 294 Other Organic Psychotic Conditions (Chronic) 2 2 0.48% (0.00% 1.16%) 30 36 0.79% (0.48%, 1.10%) 305 Nondependent Abuse of Drugs 2 2 0.48% (0:00% 1.16%) 78 105 2.30% (1.74%',2.86%) 308 Acute Reaction to Stress 2 2 0.48% (0.00% 1.15%) 26 28 0.61% (0.37%, 0.86%) 291 Alcoholic Psychoses 0 0 65 110 2.41% (1.66%, 3.17%) 292 Drug Psychoses 0 0 75 93 2.04% (1.53%, 2.55%) 299 Psychoses With Origin Specific to Childhood 0 0 2 3 0.07% (0.00%, 0.16%) 302 Sexual Deviations & Disorders 0 0 2 2 0.04% (0.00%, 0.10%) 303 Alcohol Dependence Syndrome 0 0 47 , 6 8 1.49% (0.98%. 2.00%) 304 Drug Dependence 0 0 43 53 1.16% (0.78%, 1.55%) 306 Physiological Malfunction Arising From Mental Factors 0 0 5 5 0.11% (0.01%, 0.21%) 310 Nonpsychotic Mental Disorders - Organic Brain Damage 0 0 16 21 0.46% (0.20%. 0.72%) 312 Disturbance of Conduct NEC 0 0 11 12 0.26% (0.10%, 0.43%) 313 Disturbance of Emotions - Childhood and Adolescence 0 0 0 0 314 Hyperkinetic Syndrome of Childhood 0 0 3 4 0.09% (0.00%, 0.19%) Subtotal for Severe Mental Illnesses3 157 319 76.87% (69.70% ; 84.04%) 1141 2870 62.97% (58.70% ,67.23%) TOTAL 192 415 100% 1489 4558 100% 11ndividual subjects may have different diagnoses for each hospitalization. The total and sub-total number of subjects do not equal the sum of subjects for the diagnostic categories. 2 Includes general symptoms (780) and poisoning (963-969,977) 3 Includes ICD-9 diagnostic codes 290-299 and 310 Table 4.4 Percentage and Odds Ratio of Receiving Diagnosis of Selected Mental Health Conditions Immigrant (N=786) Comparison (N=3962) #of subiects % 95% CI #of subiects % 95% CI OR' 95% CI OR 2 95% CI Senile & Presenile Organic Psychotic Conditions 61 7.76% (5.89% ,9.63%) 438 11.06% (10.08%, 12.03%) 0.31 (0.22,0.42) 0.56 (0.36 0.87) Schizophrenic Psychoses 248 31.55% (28.30% ,34.80%) 1216 30.69% (29.25%, 32.13%) 1.12 (0.94,1.34) 1.02 (0.81 1.30) Affective Psychoses 364 46.31% (42.82%, 49.80%) 2103 53.08% (51.52%, 54.63%) 0.78 (0.66,0.92) 1.07 (0.85 1.36) Other Nonorganic Psychoses 157 19.97% (17.18%,22.77%) 625 15.77% (14.64%, 16.91%) 1.35 (1.10,1.65) 1.44 (1.09 1.91) Neurotic Disorders 305 38.80% (35.40% ,42.21%) 2083 52.57% (51.02%,54.13%) 0.55 (0.46,0.65) 0.74 (0.59 0.93) Drug Dependence 9 1.15% (0.40%, 1.89%) 520 13.12% (12.07%, 14.18%) 0.07 (0.04,0.13) 0.12 (0.06 0.26) Depressive Disorder NEC 325 41.35% (37.90%, 44.79%) 2524 63.71% (62.21%, 65.20%) 0.36 (0.30,0.42) 0.69 (0.55 0.88) Anxiety/Depression 383 48.73% (45.23%, 52.22%) 1949 49.19% (47.64%, 50.75%) 0.95 (0.81,1.12) 0.90 (0.72 1.13) 1 OR is immigrant vs comparison subjects, adjusting for sex and age at entry to study 2 OR is immigrant vs comparison subjects, adjusting for sex, age at entry to study and proportion of mental health visits to Chinese-speaking physicians OR in bold type indicates a result significantly different from 1 at 5% level n cr a to o Table 4.5 Rate of Mental Health Services Received from Chinese-speaking Physicians Immigrant Total N Rate 95% CI Percentage of All Mental Health Visits Claimed by Chinese-speaking Physicians General Practitioners + Psychiatrists 22374 75.65% (71.36%. 79.94%) General Practitioners 7167 83.09% (75.86%, 90.32%) Psychiatrists 15207 72.14% (67.44%, 76.85%) Mean Individual Proportion of Mental Health Visits to Chinese-speaking Physicians General Practitioners + Psychiatrists 723 0.81 (0.79.0.83) General Practitioners 666 0.89 (0.87.0.91) Psychiatrists 541 0.72 (0.69,0.76) Total N 181230 86507 94723 3773 3528 2902 Comparison Rate 9.79% 9.68% 9.88% 0.12 0.15 0.11 Percentage of Subjects Who Received All Mental Health Services from Chinese-speaking Physicians (Proportions!) General Practitioners + Psychiatrists 723 57.95% (54.35%,61.56%) 3773 4.66% General Practitioners 666 74.02% (70.69%, 77.36%) 3528 8.56% Psychiatrists 541 56.93% (52.75%, 61.12%) 2902 6.27% 95% CI (8.77%, 10.80%) (8.30%, 11.07%) (8.53%, 11.24%) (0.12,0.13) (0.14,0.16) (0.10,0.12) (3.99%, 5.34%) (7.64%, 9.48%) (5.39%, 7.15%) n 3" £> CD >-t Table 4.6 Diagnostic Categories of Mental Health Visits of Immigrant and Comparison Subjects to General Practitioners & Psychiatrists by Language Skill of Physician IMMIGRANT COMPARISON Speaks Chinese No Chinese Speaks Chinese No Chinese Code Diagnostic Category 296 Affective Psychoses 295 Schizophrenic Psychoses 300 Neurotic Disorders 50B1 Anxiety/Depression 311 Depressive Disorder NEC 298 Other Nonorganic Psychoses 290 Senile & Presenile Organic Psychotic Conditions 293 Transient Organic Psychoses 309 Adjustment Reaction 297 Paranoid States 308 Acute Reaction to Stress 294 Other Organic Psychotic Conditions (Chronic) 307 Spec. Symp7Svnd. NEC 310 Nonpsychotic Disorders-Organic Brain Damage 301 Personality Disorders 313 Disturbance of Emotions - . Childhood and Adolescence 292 Drug Psychoses 304 Drug Dependence 299 Psychoses (Child) 312 Disturbance of Conduct NEC 306 Physiological Malfunction Arising From Mental Factors 302 Sexual Deviations/Disorders 303 Alcohol Dependence 291 Alcoholic Psychoses 305 Nondependent Drug Abuse 314 Hyperkinetic Synd. (Child) 04A1 General Psych. Examination Other Diagnoses Subtotal for Severe Illnesses2 TOTAL #of % #of 95% CI #of 95% CI #of Visits 95% CI Visits % Visits % Visits % 95% CI 4918 28.6% (24.7% 32.4%) 892 15.9% (11.9% ,20.0%) 3927 21.4% (17.7%, 25.0%) 26451 15.7% (14.2% 17.1%) 3459 20.1% (16.2% 23.9%)- 1718 30.7% (23.3%, 38.1%) 3675 20.0% (16.2%, 23.8%) 18990 11.2% (10.0% 12.5%) 2253 13.1% (10.2% 16.0%) 299 5.3% (2.4%, 8.2%) 2415 13.1% (10.2%, 16.1%) 20124 11.9% (10.6% 13.2%) 1729 10.0% (8.0% 12.1%) 261 4.7% (0.0%, 9.4%) 1643 8.9% (7.4%, 10.5%) 13503 8.0% (7.2% 8.8%) 1490 8.7% (6.4% 11.0%) 1228 21.9% (16.4%, 27.5%) 2548 13.9% (10.8%, 16.9%) 36735 21.8% (20.2% 23.3%) 1332 7.7% (5.1% 10.4%) 99 1.8% (0.7%, 2.9%) 500 2.7% (1.6%, 3.9%) 1714 1.0% (0.8% 1.2%) 545 3.2% (1.5% 4.8%) 213 3.8% (1.5%, 6.1%) 477 2.6% (1.4%, 3.8%) 2612 1.5% (1.2% 1.9%) 286 1.7% (0.6% 2.7%) 49 0.9% (0.0%, 1.8%) 295 1.6% (0.7%, 2.5%) 778 0.5% (0.3% 0.6%) 272 1.6% (0.8% 2.4%) 49 0.9% (0.4%, 1.4%) 273 1.5% (1.0%, 2.0%) 3949 2.3% (1.7% 3.0%) 265 1.5% (0.8% 2.2%) 48 0.9% (0.4%, 1.3%) 195 1.1% (0.5%, 1.6%) 1032 0.6% (0.4% 0.8%) 181 1.1% (0.6% 1.5%) 36 0.6% (0.3%, 1.0%) 316 1.7% (1.2%, 2.2%) 2895 1.7% (1.4% 2.1%) 112 0.7% (0.1% 1.2%) 26 0.5% (0.0%, 1.1%) 175 1.0% (0.3%, 1.6%) 917 0.5% (0.4% 0.7%) 68 0.4% (0.2% 0.6%) 88 1.6% (0.0%, 3.3%) 105 0.6% (0.2%, 0.9%) 2569 1.5% (0.8% 2.2%) 48 0.3% (0.1% 0.5%) 65 1.2% (0.0%, 2.3%) 72 0.4% (0.1%, 0.7%) 1265 0.7% (0.6% 0.9%) 47 0.3% (0.1% 0.5%) 49 0.9% (0.3%, 1.4%) 256 1.4% (0.7%, 2.0%) 3245 1.9% (1.5% 2.4%) 43 0.2% (0.0% 0.7%) 1 0.0% (0.0%,o!l%) 9 0.0% (0.0%, 0.1%) 86 0.1% (0.0% 0.1%) 34 0.2% (0.0% 0.6%) 1 0.0% (0.0%, 0.1%) 16 0.1% (0.0%, 0.2%) 159 0.1% (0.1% 0.1%) 34 0.2% (0.0% 0.5%) 365 6.5% (0.0%, 16.8%) 1046 5.7% (1.2%, 10.1%) 24531 14.5% (11.5% 17.6%) 25 0.1% (0.0% 0.3%) 14 0.3% (0.0%, 0.6%) 8 0.0% (0.0%, 0.1%) 301 0.2% (0.1% 0.2%) 19 0.1% (0.0% 0.2%) 0 74 0.4% (0.0%, 1.0%) 463 0.3% (0.2% 0.3%) 11 0.1% (0.0% 0.1%) 3 0.1% (0.0%, 0.1%) 13 0.1% (0.0%, 0.1%) 224 0.1% (0.0% 0.2%) 6 0.0% (0.0% 0.1%) . 1 0.0% (0.0%, 0.1%) 21 0.1% (0.0%, 0.2%) 169 0.1% (0.0% 0.2%) 2 0.0% (0.0% 0.0%) 5 0.1% (0.0%, 0.2%) 197 1.1% (0.7%, 1.5%) 1799 1.1% (0.8% 1.4%) 1 0.0% (0.0% 0.0%) 0 11 0.1% (0.0%, 0.1%) 217 0.1% (0.1% 0.2%) 1 0.0% (0.0% 0.0%) 2 0.0% (0.0%, 0.1%) 13 0.1% (0.0%, 0.1%) 1926 1.1% (0.4% 1.9%) 1 0.0% (0.0% 0.0%) 11 0.2% (0.0%, 0.5%) 451 0.3% (0.1% 0.4%) 1 0.0% (0.0% 0.0%) 0 9 0.0% (0.0% 0.0%) 33 0.2% (0.0% 0.3%) 76 1.4% (0.0%, 3.4%) 106 0.6% (0.0%, 1.4%) 1714 1.0% (0.7% 1.4%) 11025 64.0% (59.9% 68.1%) 3125 55.8% (47.0%, 64.6%) 9351 50.9% (46.2%, 55.5%) 54436 32.2% (30.1% 34.3%) 17216 100% 5599 100% 18386 100% 168828 100% 1 MSP diagnostic codes; all the others are ICD-9 codes 2 Includes ICD-9 diagnostic codes 290-299 and 310 Table 4.7 Odds Ratio of Receiving Diagnosis of Selected Mental Health Conditions by Proportion of Mental Health Visits to Chinese-speaking Physicians Immigrant Comparison OR 95% CI OR 95% CI Senile & Presenile Organic Psychotic Conditions 0.11 (0.04,0.28) 0.67 (0.44,1.02) Schizophrenic Psychoses 0.34 (0.20,0.58) 1.78 (1.38,2.31) Affective Psychoses 1.10 (0.64,1.88) 0.62 (0.48,0.80) Other Nonorganic Psychoses 0.47 (0.27,0.82) 1.26 (0.92,1.72) Neurotic Disorders 0.96 (0.58,1.60) 0.60 (0.46,0.77) Depressive Disorder NEC 0.24 (0.14,0.40) 0.41 (0.32,0.53) Drug Dependence 0.10 (0.02,0.57) 0.32 (0.20,0.53) Anxiety/Depression 2.01 (1.21,3.34) 0.97 (0.75,1.24) 1 OR is the odds of ever being diagnosed with a condition of an individual who received all mental health care from Chinese-speaking physicians (proportions) versus the odds of ever being diagnosed with a condition of an individual who received ail mental health care from non-Chinese-speaking physicians (proportions), adjusting for sex and age; bold type indicates an OR that is significantly different from 1 at 5% level Chapter 4 Appendix 4.1 ICD-9 Diagnostic Categories for Severe Mental Illness 290 Senile and presenile organic psychotic conditions 291 Alcoholic psychoses 292 Drug psychoses 293 Transient organic psychotic conditions 294 Other organic psychotic conditions 295 Schizophrenic psychoses 296 Affective psychoses 297 Paranoid states 298 Other nonorganic psychoses 299 Psychoses with origin specific to childhood 310 Nonpsychotic disorders of organic brain damage CDIC Therapeutic Classes of Psychiatric Medications for Treatment of Severe Mental Illness 28160803 Chlorpromazine 28160804 Fluphenazine 28160805 Mesoridazine 28160806 Perphenazine 28160808 Prochlorperazine 28160809 Promazine 28160810 Thioridazine 28160811 Trifluoperazine 28160814 Droperidol 28160815 Haloperidol 28160816 Loxapine 28160818 Thiothixene 28160819 Flupenthixol 28160820 Pericyazine 28160821 Thioproperazine 28160822 Methotrimeprazine 28160823 Pepotiazine 28160824 Pimozide 28160826 Fluspirilene 28160827 Clozapine 28160828 Risperidone 28160829 Remoxipride 28160830 Zuclopenthixol 28240818 Olanzapine 28280001 Lithium 124 Chapter 4 BIBLIOGRAPHY 1. Cooper L A , Roter D L , Johnson R L , Ford DE, Steinwachs D M , Powe NR. Patient-centered communication, ratings of care, and concordance of patient and physician race. Ann Intern Med 2003; 139: 907-915. 2. Flaskerud JH. The effects of culture-compatible intervention on the utilization of mental health services by minority clients. Community Ment Health J 1986; 22: 127-141. 3. Harris K M , Edlund M J , Larson S. Racial and ethnic differences in the mental health problems and use of mental health care. Med Care 2005; 43: 775-784. 4. Intercultural Committee. Multiculturalism and Mental Health: Developing Culturally Competent Systems of Care. Vancouver: Greater Vancouver Mental Health Service: 1999. 5. Maramba G G , Hall G C N . Meta-analyses of ethnic match as a predictor of dropout, utilization, and level of functioning. Cultural Diversity and Ethnic Minority Psychology 2002; 8: 290-297. 6. Newbold B. Health status and health care of immigrants in Canada: a longitudinal analysis. J Health Serv Res Policy 2005; 10: 77-83a. 7. Review of the Literature on Migrant Mental Health. Canadian Task Force on Mental Health Issues Affecting Immigrants and Refugees: 1988. 8. Matsuoka JK, Breaux C, Ryujin D H . National utilization of mental health services by Asian Americans/Pacific Islanders. J Community Psychol 1997; 25: 141-145. 9. Takeuchi DT, MokuauN, Chun C-A. Mental health services for Asian Americans and Pacific Islanders. J Ment Health Adm 1992; 19: 237-245. 10. Peters R. Increasing the responsiveness of mainstream mental health services to ethnocultural minorities. In: Health and Cultures Vol . II Programs, Services and Care. Edited by Masi R, Mensah L, McLeod K A . Oakville: Mosaic Press: 1993: 187-203. 11. Sue S, Fujino DC, Hu L, Takeuchi DT, Zane NWS. Community mental health services for ethnic minority groups: A test of the cultural responsiveness hypothesis. J Consult Clin Psychol 1991; 59: 533-540. 12. Yeh M , takeuchi DT, Sue S. Asian-American children treated in the mental health system: A comparison of parallel and mainstream outpatient service centers. J Clin Child Psychol 1994; 23: 5-12. 13. Bebbington PE, Meltzer H , Brugha TS et al. Unequal access and unmet need: neurotic disorders and the use of primary care services. Psychol Med 2000; 30: 1359-1367. 125 Chapter 4 14. Chen C-N, Wong J, Lee N , Chan-Ho M - W , Lau JT-F, Fung M . The Shatin Community Mental Health Survey in Hong Kong. Arch Gen Psychiatry 1993; 50: 125-133. 15. Takeuchi DT, Sue S, Yeh M . Return rates and outcomes from ethnicity-specific mental health programs in Los Angeles. A m J Public Health 1995; 85: 638-643. 16. Snowden LR. Bias in mental health assessment and intervention: theory and evidence. A m J Public Health 2003; 93: 239-243. 17. Jerrell J M . Effect of ethnic matching of young clients and mental health staff. Cult Diversity Ment Health 1998; 4: 297-302. 18. Borowsky SJ, Rubenstein L V , Meredith LS , Camp P, Jackson-Triche M , Wells K B . Who is at risk of nondetection of mental health problems in primary care? J Gen Intern Med 2000; 15: 381-388. 19. Snowden LR, Cheung F K . Use of inpatient mental health services by members of ethnic minority groups. A m Psychol 1990; 45: 347-355. 20. Chen A W . (Unpublished manuscript) Determinants of Mental Health Consultations by Recent Chinese Immigrants in British Columbia. University of British Columbia: 2006 21. Kazanjian A , Reid RJ, Pagliccia N , Apland L , Wood L. Issues in Physician Resources Planning in B.C.: Key Determinants of Supply and Distribution, 1991-96. H H R U 00:02. Vancouver: Centre for Health Services and Policy Research, University of British Columbia: 2000. i 22. Snowden L R , Hu T-W. Outpatient service use in minority-serving mental health programs. Adm Policy Ment Health 1996; 24: 149-159. 23. Sue S. Psychotherapeutic services for ethnic minorities: Two decades of research findings. A m Psychol 1988; 43: 301-308. 24. Saha S, Komaromy M , Koepsell TD, Bindman A B . Patient-physician racial concordance and the perceived quality and use of health care. Arch Intern Med 1999; 159: 997-1004. 126 CHAPTER 5 MENTAL H E A L T H CONSULTATION BY IMMIGRANT AND CANADIAN-BORN CHINESE IN BRITISH COLUMBIA 1 Chinese immigrants began arriving in British Columbia (BC), Canada in 1858, drawn by the discovery of gold in the interior of the province ' . The first group of approximately 300 came via California; by the following year, a few thousands were sailing to Canada's west coast directly from Asia. These first irrimigrants intended to be migrant workers rather than settlers. Nevertheless, in 1861, the first Chinese baby was born on Canadian soil in Victoria. The first census in 1871 estimated that there were 3000 Chinese residing in the province. The proliferation of ethnic Chinese created a backlash in the country and, in subsequent years, the Canadian government introduced various legislative measures to restrict immigration from China. Since changes in Canada's immigration policies in 1967, when national origin was no longer a criterion in the selection of immigrants, Chinese immigrants began to arrive again in great numbers and proportions. In 2001, nationals from People's Republic of China, Taiwan and Hong Kong comprised 35% of the 34,000 newcomers to the metropolitan area of Vancouver.4 The ethnic makeup of the population in B C was rapidly transformed. In 2001, almost 10% of the 3.9 million residents in B C reported to be of Chinese ethnicity and 6% were born in predominantly Chinese territories outside of Canada.5 Concomitant with the growing proportion of Chinese in the province, there are concerns raised that ethnic minorities, especially Chinese and Chinese immigrants are under-served by the mental health care system. The sparse evidence in published literature seems to justify these concerns. Asian Americans consistently showed a lower probability of using mental health services (1-6). Among Australian patients with 1 A version of this journal will be submitted for publication in a journal. 2 Marjorie P. Kohli, University of Waterloo, http://ist.uwaterloo.ca/~marj/genealogy/chinese.html 3 Chinese Canadian National Council, Toronto Chapter, http://www.ccnc.ca/toronto/history/index.htmI 4 Citizenship and Immigration Canada Facts and Figures 2001 http://www.cic.gc.ca/english/pub/facts2001/limm-05.html 5 Statistics Canada 2001 Census of Canada. Catalogue numbers 97F0010XCB2001001 & 97F0009XCB2001002 http://wwwl2.statcan.ca/english/census01/Products/standard/themes/DataProducts.cfm?S=l 127 Chapter 5 serious psychiatric disorders, all immigrants were found to be under-represented in inpatient and outpatient services compared to the Australian-born and Asian immigrants had the lowest level (7). In Canada, Chinese were under-represented in Vancouver's community mental health system and older Chinese in a national survey reported lower rates of visiting psychologists and psychiatrists (8-10). At the same time, there is also evidence that the prevalence of mental disorders is lower among the Chinese population. Chinese Americans in Los Angeles were found to have lower rates of depression and dysthymia (11). Results of epidemiological studies also demonstrated that prevalence of mental disorders was lower in Taiwan, Hong Kong and China than in most other countries (12-14). Therefore, possible explanations for the apparent under-utilization of mental health services by Chinese in Canada can be either better mental health status or inequitable access. Chinese as an ethnic group may be at lower risk of psychiatric problems, or Chinese immigrants may enjoy the advantage of the "healthy immigrant effect", thus obviating the need for mental health intervention. On the other hand, language or cultural barriers may impede Chinese immigrants' access to mental health services. Few studies on mental health service utilization by Chinese minorities distinguish between Chinese immigrants and Chinese who are native-born. Previous studies using BC's linked immigration and provincial health administrative databases concluded that recent Chinese immigrants to B C , including those who were diagnosed with severe and persistent mental illness, had lower rates of utilization of all types of inpatient and ambulatory mental health services (15,16). There is some evidence that American-born Chinese are more likely than foreign-born Chinese to have professional consultations for psychological problems (17). Little is known about the utilization of mental health services by Canadian-born Chinese. Comparing utilization patterns of immigrants with the Canadian-born will offer some clues as to the reasons for the lower levels of utilization reported in the literature. If language barrier or the "healthy immigrant effect" are the primary reasons for less utilization, the expectation is that the next generation will exhibit the same pattern of utilization as other Canadians. If better mental health status related to Chinese ethnicity accounts for the differences in utilization rates, then not only 128 Chapter 5 should the prevalence of psychiatric disorders be commensurate with rates of utilization, utilization will also be similar between the immigrants and the Canadian-born. The aim of the current study is to determine if there are differences in mental health service utilization between Canadian-born Chinese, immigrant Chinese and the non-Chinese population in the province of BC. The questions to be addressed are: 1. Are Canadian-bom Chinese and immigrant Chinese as likely to consult professionals for mental health reasons as other British Columbians? 2. Can the difference in utilization be explained by differences in morbidity, such as prevalence of depression? 3. What are the factors associated with mental health consultation among the Chinese? METHODS Study Design This study is a secondary analysis of the Canadian Community Health Survey (CCHS) Cycle 1.1.6 The CCHS is a cross-sectional survey conducted nationally every year by Statistics Canada. The master microdata file is available for research purposes by application to the Social Science and Humanities Research Council and Statistics Canada. Cycle 1.1 was a large general health survey designed to collect data at the health region level. It surveyed persons aged 12 or over living in private dwellings in all the provinces and territories of Canada (18). Households were sampled at the health region and provincial level using three sampling frames: an area frame, a random digit dialling frame and a telephone list frame. Respondents were randomly selected from households with over-sampling of youths and seniors. Interviews were conducted in person or by telephone using computer assisted methods. Besides English and French, the questionnaire was translated into Chinese, Punjabi and Inuktitut. Interviewers with a variety of language skills were recruited. Data were collected from September 2000 to 6 The general health survey Cycle 1.1 was chosen for this study, rather than Cycle 1.2, the mental health and wellbeing survey, because it had a larger sample size,, thus allowing analyses to be done on the sub-sample of Chinese respondents. Cycle 2.1 was not used because BC did not opt into the depression module in this survey conducted in 2002/3 and examination of the relationship between mental health consultation and morbidity would not be possible. 129 Chapter 5 November 2001. The response rate in B C was 84.7%, yielding a sample of 18302 individuals. Subjects Subjects were the 18,302 respondents in BC. The ethnicity of each respondent was identified by the language of interview, the self-reported ethnicity of ancestors and respondent's cultural/racial background. Respondents were considered ethnic Chinese if response to one of the above questions was Chinese. A l l respondents who reported being an immigrant or whose place of birth was outside of Canada were considered immigrants. Study Variables Contents of the Cycle 1.1 covered a wide range of health status and health behaviour topics, in addition to demographic information. The primary outcome of this study was mental health consultation, which was a dichotomous variable in the CCHS. Respondents were asked if they had any contact with mental health professionals in the past 12 months. If the answer was affirmative, they were then asked about the number of contacts and the type of professionals, which included general practitioners, psychiatrists, psychologists, nurses, social workers and counsellors. Other variables of interest to this study were self-reported depressive symptoms, number of general medical consultations and self-reported unmet health care needs in the previous 12 months. The experience of depressive symptoms in the past 12 months was measured by the Composite International Diagnostic Interview (CIDI) Short Form version, which yielded a score of 0 to 8. The scoring scheme of the instrument assigned to each score a probability that the respondent suffered a clinical depressive episode and higher score represented a higher probability. For general medical consultations and unmet needs, respondents were asked respectively to report the number of contacts they had with their family doctor or paediatrician in the past 12 months and if they failed to receive needed health care. In addition to depression and general medical contacts, factors associated with Chinese respondents' mental health consultations examined in this study were sex, age, place of birth, household income, education, marital status, job status in the last year, Chinese language skill, general health status and impairment; for immigrant Chinese, 130 Chapter 5 additional factors were age at immigration, years in Canada and English language skill. Impairment was denoted by how often a health problem restricted the activities they could perform and if they required assistance in six activities of daily living. A l l the data were self-reported. Data Analysis Respondents were assigned to one of four groups: immigrant Chinese, Canadian-born Chinese, immigrant non-Chinese, Canadian-born non-Chinese. Weighted distributions and means of results of mental health consultation, unmet health care need, depression and number of medical consultations were stratified by ethnicity and immigration status. The weights were supplied by Statistics Canada and took into account the sampling frames, non-response rates and other sampling issues (18). The resulting statistics were estimates for the entire non-institutionalized population in B C aged 12 or over. When the number of respondents in a category was too small, analysis was stratified only by ethnicity (Chinese versus non-Chinese) and not by immigration status. Weighted logistic regression models of the odds ratios (OR) of having any contacts with mental health professionals by ethnicity and immigration status were also run, with and without adjustment for depression score and number of general medical consultations. The factors associated with mental health consultation of Chinese respondents were analyzed first by weighted univariate logistic regression models of each factor and then by models of the same factors adjusting for depression score. The 95% confidence limits of all analyses were calculated with the B O O T V A R macro developed by Statistics Canada, which applied 500 bootstrap weights the agency generated. RESULTS Description of Subjects Of the 18302 B C respondents to the survey, 879 were immigrant Chinese and 205 were Canadian-born Chinese, representing the estimated 7.9% immigrant Chinese and 1.6% Canadian-born Chinese population in BC. Table 5.1 shows the weighted means and frequencies of the demographic characteristics of the survey population. Canadian-131 Chapter 5 born Chinese were significantly younger and immigrant non-Chinese were slightly older than the overall population, reflecting historical irnmigration trends in the province. Only health regions with a substantial number of Chinese residents are named in the table; regions with few Chinese residents are aggregated in the "Other" category. With the exception of the Capital health region which serves the provincial capital of Victoria, all the health regions shown are located in the metropolitan Vancouver area. Approximately 95% of immigrant Chinese lived in greater Vancouver, concentrating in the cities of Vancouver, Richmond and Burnaby; Canadian-born Chinese mostly resided in Vancouver, South Fraser Valley and Burnaby, with more than 16% scattered in the Capital and "Other" regions. In contrast, close to half of the Canadian-born non-Chinese population lived outside of greater Vancouver and Victoria. The Chinese immigrants had been in this country for an average of almost 13 years and the majority were born in China or Hong Kong. Descriptive Results Mental Health Consultation Tables 5.2a and 5.2b present the weighted descriptive results of the percentage of respondents who contacted mental health professionals and the number of contacts among those who did. Approximately 9.5% of the population of B C reported having contacted a mental health professional in the previous 12 months. The percentage was much higher among the Canadian-born non-Chinese (11%) than among the immigrant and Canadian-born Chinese (2.9% and 3.4%). Although immigrant Chinese who did consult appeared to have the most consultations, the differences in the number of contacts were not statistically different. Depression Tables 5.3a and 5.3b show the mean depression score and percentage of B C respondents who had a moderate to high probability of clinical depressive episode in the past year. For this study, moderate probability was a score of 2 to 3 on the CIDI, i indicating a probability of 0.25-0.50; score of 4 or more translated into a high probability of 0.80-0.90. Out of a maximum score of 8, the mean depression score for Canadian-born non-Chinese was 0.73, with 13% scoring 2 or above. For immigrant and Canadian-132 Chapter 5 born Chinese, the mean depression scores were 0.26 and 0.43 respectively, with 5% and 8% scoring in the moderate to high probability range. Even among respondents who were at moderate to high risk of depression, relatively fewer Chinese than non-Chinese consulted mental health professionals in the previous year (22.6% and 40.3%), as shown in Table 5.3c. Medical Consultations The contrast between Chinese and non-Chinese was less drastic in the number of medical consultations reported in the last 12 months (Table 5.4). Canadian-bom non-Chinese reported 4.75 contacts with their general practitioners or paediatricians while immigrant Chinese and Canadian-born Chinese reported 3.67 and 3.40 respectively. Unmet Health Care Needs In terms of unmet health care needs, Table 5.5a shows that 14% of Canadian-born non-Chinese reported not receiving needed health care in the past 12 months while only half as many immigrant and Canadian-born Chinese did the same (6% and 7%). Of those who had unmet health care needs, 9% of the non-Chinese respondents compared to only 5% of the Chinese respondents acknowledged that the need was related to an emotional or mental health problem (Table 5.5b). Regression Results Results of weighted logistic regression modelling corroborated the descriptive findings on mental health consultation, even though the sampling variability was very high in many instances. As shown in Table 5.6a, immigrant and Canadian-born Chinese were much less likely than Canadian-born non-Chinese to contact mental health professionals (OR=0.24 and 0.28). Even after adjusting for their lower level of depression and general medical consultations, the difference was barely affected (OR=0.33 and 0.31). Among those who had moderate to high probability of depression, similar observations were made (Table 5.6b). Chinese were less than half as likely as non-Chinese to make contact with mental health professionals (OR=0.43). 133 Chapter 5 Factors Associated with Mental Health Consultation among Chinese Respondents Owing to the small number of Chinese respondents and the absence of significant difference between immigrant Chinese and Canadian-born Chinese in the results thus far, the two groups were combined in the analyses of factors associated with mental health consultation, except for the three immigrant factors of age at immigration, years in Canada and English language skill. The results of the univariate models and the models adjusting for depression score are summarized in Table 5.7. Of all the factors analyzed, the only ones found to have statistically significant impact on the odds of mental health consultation in the univariate models were the ability to converse in Chinese, restriction of activities, number of general medical consultations and depression score. Not surprisingly, higher depression score was associated with higher odds of having contacts with mental health professionals. In addition, those who were able to converse in Chinese, those who were sometimes or often impaired by health conditions and those with higher number of general medical consultations were also more likely to have mental health consultations. Many of the other factors also showed strong association with mental health consultation although the effects did not reach statistical significance. Women were more likely to consult than men. Those who spoke English were more likely to consult than those who did not. Chinese born in China and those with secondary education or less were least likely to consult. Those who were separated, divorced or widowed were more likely to consult than single or married individuals. Those who suffered from poor health and those who required assistance with daily activities were more likely to have mental health consultations. Unexpectedly, those who had a job in the past year were more likely to consult than those who did not. After adjusting for depression score, the only factor that retained statistical significance in odds of mental health consultation was number of medical consultations. DISCUSSION This study demonstrates that Chinese in BC are less likely to consult mental health professionals, notwithstanding their lower level of risk for depression. Furthermore, lower levels of depression and mental health consultation are found in both the Canadian-born and immigrant Chinese. Even though Canadian-born Chinese seemed 134 Chapter 5 somewhat more likely than immigrant Chinese to suffer from depression, a statistically significant difference was not established in this study and the Canadian-born showed no greater inclination to seek professional help for mental health reasons. Although the younger age of the Canadian-born Chinese might also have affected the propensity to seek help, ethnicity, rather than immigration status, emerged as the more powerful influence in this study where the Chinese respondents, regardless of place of birth, reported levels of mental health consultation lower than the non-Chinese, including the non-Chinese immigrants. On the other hand, immigration status played a significant role among the non-Chinese. Ethnic variation within the heterogeneous non-Chinese population was not explored in this study; nonetheless, non-Chinese immigrants as a group were significantly less likely than non-immigrants to report depressive episodes or to contact mental health professionals. Findings from this study do not support assertions that language barrier or the "healthy immigrant effect" account for the lower utilization of mental health services observed in immigrants. While these factors may explain the utilization pattern among non-Chinese immigrants and minorities, there are apparently other elements related to the ethnicity of the Chinese respondents that keep their use of mental health resources and, to a certain degree, their prevalence of depression, lower than those of the rest of the population. Although language barriers and better mental health status may indeed exist in the Chinese community, the results of this study support cultural explanations of differences in rate of mental health service utilization. Cultural issues underlie the mental health experience of ethnic Chinese, both in their expression of emotional difficulties and their response to them. Depressive disorder, as defined by diagnostic criteria developed primarily in the North American context, represents only one formulation of emotional i difficulties and may not fully describe the experience by individuals from other cultures. Variations across cultures in the presentation of affective states have long been recognized, with somatic symptoms being core complaints in many cultures and these complaints are often not attributed to psychological origins (19, 20). Furthermore, although the instrument used to measure depression - the CJDI - was selected by the World Health Organization for international epidemiological surveys, its cross-cultural 135 Chapter 5 validity is questionable. The initial field testing of the CJDI was carried out in 18 countries involving only 575 subjects; only feasibility and inter-rater reliability were assessed. Subsequent validation studies were conducted primarily in western countries (21-23). Thus, this depression scale may not capture the entire spectrum of psychological distress that Chinese respondents experience. Even when they encounter mental health problems, Chinese may not interpret them as emotional in origin or as issues for health care (19). This is corroborated by the finding in this and other studies that they are less likely to perceive any unmet health or mental health needs, even when they are accessing less of the available resource (2). At least three cultural factors have been postulated to come into play in the help-seeking process: acculturation, self-construal or self-concept, and ethnic identity. Owing to their communication styles, interdependent self-concept and aversion to shame, Chinese may prefer informal resources, such as community and family support, over formal services. The western approaches to mental health treatment, which emphasize individual cognition and well-being, are incongruent with the Chinese view of a collective identity. Strong stigma against mental health problems and treatment in the Asian community is another deterrent to seeking psychological and psychiatric help. Moreover, these cultural factors are passed on to the second generation through parenting and community dynamics and it may take several generations for Asians in North America to be completely acculturated with respect to mental health service utilization (24, 25). Acculturation is multidimensional and each dimension may occur at a different pace (26). ; While the use of general medical care by the Chinese in this study lagged only slightly behind the non-Chinese, cultural issues specific to mental health created a much wider chasm between the two populations in mental health consultation. In spite of the inadequacies of the diagnostic category of depressive disorder and the CIDI and the disinclination of the Chinese to seek mental health care, the distress that was captured by the survey instrument was found to be a strong predictor of mental health consultation. In terms of appropriateness of services, this finding is positive evidence that mental health services are accessed by those with demonstrated need. While depression score account for much of the influence of overall health and functioning on mental health service utilization, the number of general medical 136 Chapter 5 consultations had an impact independent of depression. The association between mental health consultation and general medical consultation is explained by more than comorbid physical and mental health conditions; it may also be the result of individual help-seeking behaviours or the increased opportunity for users of health care to be referred to mental health resources. Statistically significant associations could not be established between most of the other respondent attributes and mental health consultation owing to the small sample size and large sampling variability, although the strength of the effects suggests that some relationships probably exist. Many of these non-significant associations are in the expected direction. Women and those with poorer health and functional status are more likely to contact mental health professionals. The effect of English language skill indicates that language barrier does play a role. The contribution of other factors, such as education, place of birth and job status, is likely more complex and will need to be explored further in future studies. Worthy of note is the negative effect of the ability to converse in Chinese, which is much reduced by depression score. As an indication of acculturation, this effect of Chinese language ability is consistent with the hypothesis that affinity with the Chinese culture predisposes an individual against the use of mental health services, and this association is partially explained by their lower levels of depression as measured by the CIDI. A limitation of this study, besides the small number of Chinese respondents, is that data are self-reported and are subject to recall errors. This is a limitation inherent to survey methodology. A study in Ontario comparing mental health care use reported in a survey and recorded in the administrative databases found only moderate agreement between the data sources in reporting any use after adjusting for chance and low agreement in the volume of use (27). However, since the survey question defined mental health care more broadly than the medical resources captured in the administrative databases, not all differences could be attributed to respondent error. Studies on self-reported use of general health care result in similar findings of discrepancies between self-reported data and medical or administrative records, but the implications of these findings on the current study are mixed (28-37). On one hand, respondents are found to be more accurate in recalling whether they had any contact than the number of contacts. On the other hand, the magnitude of discrepancy is related to the age, sex, immigration 137 Chapter 5 status, type of health service, volume of use and the health and emotional status of the respondent; discrepancies also increase with the recall period. Recall of psychiatric history and use of mental health services may also be affected by respondent's current emotional state and intensity of past episodes (27, 38). Although medical and administrative data are also subject to errors, they have been considered the "gold standard" in measuring health service utilization where comprehensive data can be collected. Interpretations of results from this and other studies based on survey data will have to take into account the variances reported between these data sources. CONCLUSION Data from the CCHS confirm that ethnic Chinese are much less likely to use professional mental health resources than non-Chinese, even after taking into account their lower probability of suffering from clinical depression. Unlike the non-Chinese population, place of birth has little effect among the Chinese, with the Canadian-born Chinese being as unlikely to contact mental health professionals as the immigrants. Cultural factors, rather than language barriers or immigration status, provide more plausible explanations for the difference observed. The results illustrate the resilience of a minority culture from generation to generation, at least where mental health issues are concerned. Efforts to promote mental health in the Chinese and other ethnic minority communities must consider the cultural differences in perception of mental health problems and help-seeking behaviours, as well as the persistence of these differences beyond the first generation of immigrants. Disclaimer The research and analysis are based on data from Statistics Canada and the opinions expressed do not represent the views of Statistics Canada. 138 Table 5.1 Characteristics of BC Respondents to CCHS Cycle 1.1 Immigrant Chinese Canadian-born Chinese (N=879) (N=205) 95% CI 95% CI Percentage in Population 7.87% (7.19% 8.56%) 1.60% (1.31% 1.89%) Sex (%) Female 49.60% (45.87% 53.33%) 50.45% (42.21% 58.69%) Male 50.40% (46.67% 54.13%) 49.55% (41.31% 57.79%) Mean Age (Years) 41.57 (40.41 42.73) 27.83 (25.86 29.80) Health Region (%) Vancouver 41.94% (37.52% 46.36%) 36.44% (28.08% 44.80%) Richmond 20.36% (17.56% 23.15%) 9.66% (5.57% 13.74%) North Shore 2.63% (1.60% 3.66%) 2.16% (0.38% 3.94%) Burnaby 15.15% (12.57% 17.73%) 13.92% (7.76% 20.08%) Simon Fraser 6.19% (3.90% 8.47%) 4.59% (1.77% 7.41%) South Fraser Valley 8.50% (5.49% 11.51%) 16.47% (8.72% 24.23%) Capital 3.14% (1.71% 4.57%) 5.35% (1.64% 9.06%) Other 2.10% (1.22% 2.98%) 11.41% (7.06% 15.76%) Mean Number of Years In Canada* 12.56 (11.64 13.48) Place of Birth* (%) China 36.40% (32.13% 40.67%) Hong Kong 35.69% (31.47% 39.91%) Taiwan 14.71% (11.76% 17.67%) Other 13.19% (10.41% 15.98%) 'Immigrant Chinese only Immigrant Non-Chinese Canadian-born TOTAL (N=3439) Non-Chinese (N=13779) (N=18302) 95% CI 95% CI 21.20% (20.31% 22.10%) 69.33% (68.27% 70.39%) 100.00% 49.74% (47.75% 51.72%) 51.15% (50.46% 51.84%) 50.71% 50.26% (48.28% 52.25%) 48.85% (48.16% 49.54%) 49.29% 48.62 (47.83 49.42) 41.19 (40:97 41.41) 42.58 17.73% (15.86% 19.61%) 10.10% (9.45% 10.75%) 14.65% 4.38% (3.67% 5.08%) 2.20% (1.92% 2.47%) 4.21% 7.69% (6.72% 8.66%) 4.08% (3.79% 4.37%) 4.70% 6.80% (6.07% 7.53%) 3.15% (2.84% 3.46%) 5.04% 9.16% (7.92% 10.40%) 8.16% (7.73% 8.58%) 8.16% 19.58% (17.28% 21.88%) 13.86% (13.08% 14.63%) 14.69% 7.52% (6.38% 8.65%) 9.13% (8.74% 9.52%) 8.26% 27.15% (25.60% 28.69%) 49.33% (48.58% 50.08%) 40.30% / Chapter 5 Table 5.2a Percentage of BC Residents Who Contacted Mental Health Professionals in Past 12 Months N ! % 9 5 % CI Immigrant Chinese 867 2 .89% (1 .65%, 4.13%) Canadian-born Chinese 205 3 .39% (1 .04%, 5.74%) Immigrant Non-Chinese 3 4 1 7 7 .67% (6 .58%, 8.76%) Canadian-born Non-Chinese 13545 10.98% (10.31%, 11.65%) TOTAL 18034 9 . 5 1 % (8 .97%, 10.05%) #N = number of respondents in BC who contributed to the analysis Table 5.2b Mean Number of Mental Health Consultations of BC Residents Of Those Who Contacted Mental Health Professionals N? No. 9 5 % CI Immigrant Chinese 28 8.66 (2 .83 ,14 .49 ) Canadian-born Chinese 9 4 . 1 1 (0 .05 ,8 .17 ) Immigrant Non-Chinese 263 5.45 (4 .18 ,6 .73 ) Canadian-born Non-Chinese 1515 7 . 1 5 (6 .34 ,7 .95 ) TOTAL 1815 6.87 (6 .18 ,7 .57 ) #N = number of respondents in BC who contributed to the analysis 140 Chapter 5 Table 5.3a Mean Depression Score of BC Residents (Range=0 to 8) N! Score 95% CI Immigrant Chinese 861 0.26 (0.16,0.35) Canadian-born Chinese 204 0.43 (0.19,0.67) Immigrant Non-Chinese 3380 0.47 (0.41 ,0.54) Canadian-born Non-Chinese 13417 0.73 (0.69,0.77) TOTAL 17862 0.63 (0.60,0.66) # N = number of respondents in BC who contributed to the analysis Table 5.3b Percentage of BC Residents Who Had Moderate to High Probability of Depression (Score>=2) N ! % 95% CI Immigrant Chinese 861 5.11% (3.22%, 6.99%) Canadian-born Chinese 204 7.86% (3.26%, 12.45%) Immigrant Non-Chinese 3380 8.86% (7.67%, 10.06%) Canadian-born Non-Chinese 13417 13.44% (12.74%, 14.14%) TOTAL 17862 11.72% (11.16%, 12.27%) #N = number of respondents in BC who contributed to the analysis Table 5.3c Percentage of BC Residents Who Contacted Mental Health Professionals in Past 12 Months Of Those Who Had Moderate to High Probability of Depression (Score>=2)' N! % 95% CI Chinese 16 22.63% (9.89%, 35.36%) Non-Chinese 892 40.26% (37.25%, 43.26%) TOTAL 908 39.46% (36.54%, 42.38%) Results not reported by immigration status due to Statistics Canada's disclosure policies #N = number of respondents in BC who contributed to the analysis 141 Chapter 5 Table 5.4 Mean Number of Medical Consultations of BC Residents in Past 12 Months N ! No. 95% CI Immigrant Chinese 875 3.67 (3.30,4.05) Canadian-born Chinese 204 3.40 (2.67,4.13) Immigrant Non-Chinese 3421 4.82 (4.52,5.13) Canadian-born Non-Chinese 13695 4.88 (4.73,5.04) TOTAL 18195 4.75 (4.62,4.89) # N = number of respondents in BC who contributed to the analysis Chapter 5 Table 5.5a Percentage of BC Residents Who Had Unmet Health Care Needs in Past 12 Months N ! °A 9 5 % CI Immigrant Chinese 878 5 . 5 1 % (3 .77% ,7 .26%) Canadian-born Chinese 205 6 .82% ( 2 . 9 4 % , 10.69%) Immigrant Non-Chinese 3432 10.45% ( 9 . 2 3 % , 11.66%) Canadian-born Non-Chinese 13756 13.53% (12 .79%, 14.26%) TOTAL 18271 12.13% (11 .50%, 12.76%) *N = number of respondents in BC who contributed to the analysis Table 5.5b Percentage of BC Residents Who Had Unmet Mental Health Care Needs in Past 12 Months Of Those with Unmet Health Care Needs* N! % 9 5 % CI Chinese 77 4 .58% (0 .00%, 9.75%) Non-Chinese 2336 8 .95% (7 .50%, 10.40%) TOTAL 2 4 1 3 8 .75% ( 7 . 3 1 % , 10.20%) * Results not reported by immigration status due to Statistics Canada's disclosure policies *N = number of respondents in BC who contributed to the analysis 143 Chapter 5 Table 5.6a Odds Ratio of Contacting Mental Health Professionals in Past 12 Months Controlling for Depression & Univariate Medical Consultations N! OR 95% CI N ! OJR 95% CI Immigrant Chinese 867 0.24 (0.15,0.39) 856 0.33 (0.20,0.53) Canadian-born Chinese 205 . 0.28 (0.13,0.64) 203 0.31 (0.13,0.73) Immigrant Non-Chinese 3417 0.67 (0.57,0.80) 3361 0.78 (0.64,0.94) Canadian-born Non-Chinese 13545 1.00 13341 1.00 *N = number of respondents in BC who contributed to the analysis Table 5.6b Odds Ratio of Contacting Mental Health Professionals in Past 12 Months* Of Those With Moderate to High Probability of Depression (Score>=2) Controlling for Medical Univariate Consultations N! OR 95% CI N# OR Chinese 63 0.43 (0.20,0.94) 63 0.43 Non-Chinese 2149 1.00 2139 1.00 Not stratified by immigration status owing to small number of respondents *N = number of respondents in BC who contributed to the analysis 95% CI (0.20,0.94) 144 Chapter 5 Table 5.7 Odds Ratio of Factors Associated with Mental Health Consultat ion Among Chinese Respondents Univariate Sex Female Male Age Age at Immigrat ion* Years In Canada* English Abil i ty* English No English Chinese Abil i ty Chinese No Chinese Place of Birth China Hong Kong Taiwan Other Canada Respondent Educat ion More than bachelor's degree Bachelor's degree Some post-secondary Secondary or less Marital Status Married Separated/Divorced/Widowed Single Income Low Low-middle High-middle High Job Status Last Year Had job No job General Health Status Good/excellent Poor/fair Restriction of Activit ies Sometimes/often Never Needing Help Need Help No help Number of Medical Consultat ions Depression Score •Immigrants only OR 1.82 0.99 0.98 1.00 1.59 0.38 0.25 0.86 1.67 1.21 3.43 2.87 1.78 0.97 1.71 3.64 2.17 1.11 1.68 95% CI (0.72,4.59) (0.97,1.01) (0.96,1.01) (0.95,1.04) (0.17,15.16) (0.17,0.89) (0.00,16.26) (0.27,2.78) (0.25,11.35) (0.18,8.00) (0.01,1153.69) (0.88,9.40) (0.61 ,5.20) (0.37,2.54) (0.01,224.45) Control l ing for Depression 95% CI QB 1.74 0.99 0.98 1.00 1.52 0.57 0.37 1.20 2.32 1.52 2.53 3.78 1.86 1.13 1.52 2.85 (0.50,16.35) 0.42 (0.13,1.34) (1.37,9.66) (0.46,10.18) (1.06,1.17) (1.44,1.94) (0.61,4.92) (0.97,1.01) (0.94,1.02) (0.95,1.05) (0.14,16.97) (0.22,1.49) (0.01,26.34) (0.32,4.53) (0.30,17.74) (0.21,10.88) (0.01,1015.95) (0.92,15.53) (0.60,5.82) (0.40,3.16) (0.01,180.80) 2.55 (0.02,367.65) 2.15 (0.01,336.04) 1.64 (0.01,215.33) 1.70 (0.01,235.01) 3.24 (0.03,406.29) 4.45 (0.03,592.18) 2.87 (0.49,16.78) 0.62 (0.19,2.02) 2.10 (0.88,5.00) 1.29 (0.35,4.66) 1.08 (1.04,1.13) 145 Chapter 5 BIBLIOGRAPHY 1. Cheung F K , Snowden LR. Community mental health and ethnic minority populations. Community Ment Health J 1990; 26: 277-291. 2. Harris K M , Edlund M J , Larson S. Racial and ethnic differences in the mental health problems and use of mental health care. Med Care 2005; 43: 775-784. 3. Leong FTL. Asian Americans' differential patterns of utilization of inpatient and outpatient public mental health services in Hawaii. J Community Psychol 1994; 22: 82-96. 4. Matsuoka JK, Breaux C, Ryujin D H . National utilization of mental health services by Asian Americans/Pacific Islanders. J Community Psychol 1997; 25: 141-145. 5. Meinhardt K , Vega W. A method for estimating underutilization of mental health services by ethnic groups. Hosp Community Psychiatry 1987; 38: 1186-1190. 6. Snowden LR, Cheung F K . Use of inpatient mental health services by members of ethnic minority groups. A m Psychol 1990; 45: 347-355. 7. Klimidis S, McKenzie DP, Lewis J, Minas IH. Continuity of contact with psychiatric services: irrimigrant and Australian-born patients. Soc Psychiatry Psychiatr Epidemiol 2000; 35: 554-563. 8. Peters R. Increasing the responsiveness of mainstream mental health services to ethnocultural minorities. In: Health and Cultures Vol . II Programs, Services and Care. Edited by Masi R, Mensah L , McLeod K A . Oakville: Mosaic Press: 1993: 187-203. 9. Intercultural Committee. Multiculturalism and Mental Health: Developing Culturally Competent Systems of Care. Vancouver: Greater Vancouver Mental Health Service: 1999. 10. Lai D W L , Tsang K T , Chappell N L , Lai D C Y , Chau S B Y . Health and Well Being of Older Chinese in Canada. Calgary: Faculty of Social Work, University of Calgary: 2003. 11. Takeuchi DT, Chung R C - Y , Lin K - M et al. Lifetime and twelve-month prevalence rates of major depressive episodes and dysthymia among Chinese Americans in Los Angeles. A m J Psychiatry 1998; 155: 1407-1414. 12. Chen C-N, Wong J, Lee N , Chan-Ho M-W, Lau JT-F, Fung M . The Shatin Community Mental Health Survey in Hong Kong. Arch Gen Psychiatry 1993; 50: 125-133. 13. Hwu H-G, Yeh E-K, Chang L - Y . Prevalence of psychiatric disorders in Taiwan defined by the Chinese Diagnostic Interview Schedule. Acta Psychiatr Scand 1989; 79: 136-147. 146 Chapter 5 14. WHO World Mental Health Survey Consortium. Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. J A M A 2004; 291: 2581-2590. 15. Chen A W . (Unpublished manuscript) Disparities in Utilization of Mental Health Services by Chinese Immigrants in British Columbia. University of British Columbia: 2006 16. Chen A W . (Unpublished manuscript) Determinants of Mental Health Consultations by Recent Chinese Immigrants in British Columbia. University of British Columbia: 2006 17. Ying Y - W , Miller LS . Help-seeking behavior and attitude of Chinese Americans regarding psychological problems. A m J Community Psychol 1992; 20: 549-556. 18. Beland Y . Canadian community health survey—methodological overview. Health Rep 2002; 13: 9-14. 19. Kirmayer LJ . Cultural variations in the clinical presentation of depression and anxiety: Implications for diagnosis and treatment. J Clin Psychiatry 2001; 62: 22-28. 20. Kirmayer L J , Groleau D. Affective disorders in cultural context. The Psychiatric Clinics of North America 2001; 24: 465-478. 21. Wittchen H U , Robins L N , Cottier L B , Sartorius N , Burke JD, Regier D A . Cross-cultural feasibility, reliability and sources of variance of the Composite International Diagnostic Interview (CIDI). Br J Psychiatry 1991; 159: 645-653. 22. Wittchen H U . Reliability and validity studies of the W H O - Composite International Diagnostic Interview (CLDI): A critical review. J Psychiatr Res 1994; 28: 57-84. 23. Andrews G, Peters L . The psychometric properties of the Composite International Diagnostic Interview. Soc Psychiatry Psychiatr Epidemiol 1998; 33: 80-88. 24. Yamashiro G, Matsuoka JK. Help-seeking among Asian and Pacific Americans: a multiperspective analysis. Soc Work 1997; 42: 176-186. 25. Barry DT, Grilo C M . Cultural, psychological, and demographic correlates of willingness to use psychological services among East Asian immigrants. J Nerv Ment Dis 2002; 190: 32-39. 26. Clark L , Hofsess L . Acculturation. In: Handbook of Immigrant Health. Edited by Loue S. New York: Plenum Press: 1998: 37-59. 27. Rhodes A E , Lin E, Mustard C A . Self-reported use of mental health services vs. administrative: Should we care? Int J Methods Psychiatr Res 2002; 11: 125-133. 147 Chapter 5 28. Cleary PD, Jette A M . The validity of self-reported physician utilization measures. Med Care 1984; 22: 796-803. 29. Glandon G L , Counte M A , Tancredi D. An analysis of physician utilization by elderly persons: Systematic differences between self-report and archival information. J Gerontol 1992; 47: S245-S252. 30. Hennessy K D , Reed SK. Validating self-reports of mental health service use in a chronic population. J Nerv Ment Dis 1992; 180: 399-400. 31. Manderscheid RW, Rae DS, Narrow WE, Locke B Z , Regier D A . Congruence of service utilization estimates from the Epidemiologic Catchment Area Project and other sources. Arch Gen Psychiatry 1993; 50: 108-114. ,32. Raina P, Torrance-Rynard V , Wong M , Woodward C. Agreement between self-reported and routinely collected health-care utilization data among seniors. Health ServRes 2002; 37: 751-774. 33. Reijneveld SA. The cross-cultural validity of self-reported use of health care: A comparison of survey and registration data. J Clin Epidemiol 2000; 53: 267-272. 34. Ritter PL, Stewart A L , Kaymaz H , Sobel DS, Block D A , Lorig K R . Self-reports of health care utilization compared to provider records. J Clin Epidemiol 2001; 54: 136-141. 35. Roberts RO, Bergstralh EJ, Schmidt L , Jacobsen SJ. Comparison of self-reported and medical record health care utilization measures. J Clin Epidemiol 1996; 49: 989-995. 36. Rozario PA, Morrow-Howell N , Proctor E. Comparing the congruency of self-report and provider records of depressed elders' service use by provider type. Med Care 2004; 42: 952-959. 37. Tisnado D M , Adams JL, Liu H et al. What is the concordance between the medical record and patient self-report as data sources for ambulatory care? Med Care 2006; 44: 132-140. 38. vSimon GE, VonKorff M . Recall of psychiatric history in cross-sectional surveys: Implications for epidemiologic research. Epidemiol Rev 1995; 17: 221-227. 148 CHAPTER 6 CONCLUSION This chapter summarizes the findings from Chapters 2 to 5 and discusses the general implications of these findings on mental health service policy and delivery. It concludes with some comments on the strengths and weaknesses of the research design and the databases used and suggestions for future research in this field. SUMMARY OF RESEARCH FINDINGS This thesis is a population-based research study of mental health service utilization by Canadian Chinese. Using existing administrative and survey databases, it examines rates, patterns and related issues of mental health service utilization by Chinese in British Columbia (BC). The body of the thesis consists of four manuscripts. Three of the four manuscripts concern recent Chinese immigrants to the province and data are derived from several linked administrative databases, including a federal immigration database, provincial health databases and the provincial physicians' register. The fourth manuscript is based on a Canadian general health survey and includes both immigrant and Canadian-born Chinese. The chapter on overall utilization of mental health services by Chinese immigrants (Chapter 2) reports that recent Chinese immigrants to the province are much less likely than the comparison group to use most kinds of mental health services, including primary care, specialist care, hospital care and community mental health. The disparity in utilization is quite startling, with immigrants having only 21% as many mental health visits to general practitioners as comparison group members and about 14% for other forms of care. Relative to comparison subjects, Chinese immigrants are also more likely to visit psychiatrists for psychotic disorders than for less serious mental disorders. These results suggest that Chinese immigrants rely more on primary care for their mental health needs and access specialist services only for serious problems. The relative disparity in utilization also varies by sex and age; it is smaller among women, the young and the elderly for some forms of care, consistent with other literature that these may be the most vulnerable groups within the immigrant population. This chapter raises 149 Chapter 6 several possible explanations for the under- utilization relative to the comparison group, such as lower psychiatric morbidity and barriers to access, but is not able to conclude the reason for the disparity. The examination of the determinants that contribute to mental health visits to general practitioners and psychiatrists within this population of recent Chinese immigrants (Chapter 3) finds significant variation among sub-groups. The determinants explored include individual characteristics as well as contextual factors such as the supply of physicians and, more importantly, the supply of Chinese-speaking physicians. Of the individual characteristics, years since landing and the rate of non-mental health visits to general practitioners are found to be the most consistent predictors of mental health consultations in both primary care and specialist care. Other factors, such as age, place of origin, educational level, marital status and English language skill, also show varying effects on mental health consultations depending on the sex, age group of the irrimigrants and the specialty of the physicians. Although contextual factors appear to play a role in that utilization varies by the health service delivery area, no clear relationship with the supply of physicians is observed and the mechanism by which health jurisdiction affects utilization remains to be clarified. This chapter, however, highlights the relevance of provider factors by demonstrating unequivocally that Chinese immigrants receive their mental health care predominantly from physicians who speak their native language. The most vulnerable group among both Chinese immigrants and comparison subjects - those diagnosed with severe and persistent mental illness - is the focus in Chapter 4. Results show that, with the exception of psychiatric consultations for severe disorders, Chinese immigrants have lower rates of utilization of all types of mental health services, including primary care, specialist care, hospital care, outpatient mental health care and psychiatric medications. However, the gap between Chinese immigrants and comparison subjects in this subgroup, ranging from 40% to 90%, is relatively smaller compared to the disparity found in the population at large in Chapter 2. Differences between immigrants and the comparison group are generally greater in services for less serious mental disorders. This chapter demonstrates that, even among a cohort of severely i l l individuals receiving mental health care, Chinese immigrants use fewer 150 Chapter 6 services and are less likely to receive comorbid diagnoses of the less serious mental disorders. This finding supports the contention that lower prevalence of mental disorders, at least the less serious disorders, may account for some of the under-utilization of mental health services commonly observed and reported. At the same time, there is preliminary evidence that, in addition to lower prevalence, the practice pattern of the physicians providing care also explains part of the disparity in utilization. This chapter builds on the previous one by examining physicians' language skill as a contributor to Chinese immigrants' service utilization. Compared to physicians without Chinese language skills, physicians who are able to speak Chinese tend to treat more serious rather than less serious mental disorders. As a result, high proportion of mental health care received from Chinese-speaking physicians, which describes most of the Chinese immigrants, is associated with lower odds of receiving diagnoses of depressive disorder and drug dependence. Chapter 5 reports the only study in this thesis using survey data rather than administrative databases and is the only one that provides information about Canadian-born Chinese. The results from self-reported mental health service utilization of Chinese respondents corroborate findings from administrative databases that Chinese are less likely to consult professionals for mental health reasons. This chapter adds two significant findings to the knowledge about Chinese' utilization of mental health services. First, it confirms that Chinese in B C have a lower probability of suffering from major depression but that, even after taking into account their lower level of psychiatric morbidity, they are still less likely to have contacts with mental health professionals. Secondly, no difference is found between immigrant and Canadian-born Chinese in terms of likelihood of mental health consultation. Hence, this chapter reinforces the general observation that Chinese are less likely to use mental health services, irrespective of the morbidity level. Furthermore, cultural factors which are passed on to the next generation, rather than language barriers or immigration experience, seem to be stronger influences on the use of mental health care. Other findings in this chapter confirm that probability of depression and rate of general medical consultations are predictive of likelihood of mental health consultation, supporting the important role of need as a determinant of utilization, in addition to other individual and contextual factors. 151 Chapter 6 Overall, the population-based data in B C confirm other studies and anecdotal reports that Chinese, like many minority groups, use fewer mental health resources than the rest of the population (1-11). The relative under-utilization is observed in all arenas of formal health services and is universal among sub-groups of the Chinese population, regardless of sex, age or place of birth. On the other hand, there is variation among the sub-groups in terms of the likelihood of utilization and the magnitude of the disparity. As hypothesized, the gap is smaller in the group with the highest need - those diagnosed with severe and persistent mental illness and those with higher probability of major depression. Groups reported in the literature to be at higher risk of emotional distress -immigrant women, youth and elderly - also seem to be accessing services relatively more than other immigrants (12). IMPLICATIONS FOR H E A L T H SERVICES RESEARCH AND POLICIES In health services research, the key concern goes beyond the rate of utilization and centres around equity in access to services and outcomes. Therefore, the reasons behind the observed differences in utilization are crucial to assessing if access is indeed equitable for immigrants and ethnic minorities. Explanations for under-utilization of mental health services generally fall into two categories: (a) better mental health status or lower need (b) barriers to access, including financial, language, cultural and practice barriers (13). This thesis offers some preliminary evidence with respect to both of these explanations as applied to Chinese in B C . Findings suggest that lower psychiatric morbidity and, therefore, lower need for mental health services, account for some of the observed differences in utilization. Moreover, lower morbidity is not attributable to the "healthy immigrant effect" alone, since Canadian-born Chinese are as unlikely to suffer from major depression as their immigrant counterparts. These findings, coupled with results of epidemiological surveys among Chinese populations in other countries, lend credence to the view that Chinese are less likely to suffer from mental illness, at least as defined by existing diagnostic systems (14-17). Whether these diagnostic systems are adequate to capture psychological distress in the Chinese population is a separate discourse that will not be discussed here. The "protective" factors associated with Chinese ethnicity are also resilient to changes in the 152 Chapter 6 cultural environment and appear to operate in the generation bom and raised in North America. On the other hand, this thesis also offers strong evidence that, even after accounting for effects of mental health status, Chinese are still less likely to use mental health services. These findings suggest that barriers to access cannot be ruled out. Under Canada's universal health care policy, there are few financial barriers to health care. Socio-economic context of the individual is not found to be a predictor of utilization in Chapter 3 and in fact, those with lower income tend to be more likely to contact mental health professionals in Chapter 5. Evidence on language barrier is rather unclear. English language skill is found to affect mental health service utilization in different ways among Chinese immigrants in Chapter 3 and to have a positive, albeit statistically non-significant, effect on mental health contacts in Chapter 5. On the other hand, Chinese language skill shows a more unequivocal association with such mental health contacts. Cultural barriers cover a variety of issues, from health beliefs and attitudes to values in health care, and are not assessed directly in this thesis. The fact that Canadian-born Chinese are equally unlikely as immigrants to consult mental health professionals points to the dominant role of cultural factors over immigration factors. Acculturation has been conceptualized as a matrix of two cultural affiliations - with one's native culture and with one's new culture (18). Individuals of high affiliation with the new culture but low affiliation with the native culture are considered "acculturated" while those exhibiting the opposite tendencies are "unacculturated". Those highly affiliated with both cultures are described as "bicultural" and those who identify with neither are the "marginalized". The stronger effect of Chinese language skill relative to English language skill suggests that retaining or inheriting one's native/ancestral culture is more powerful than adoption of the new culture in determining the behaviour of mental health service utilization. Barriers related to practice and service delivery are embedded in the health care environment and receive less attention than individual attributes as factors influencing health care utilization. This thesis introduces physician practice pattern as an example of such barrier. For Chinese immigrants, a form of systemic discrimination and racism inadvertently occurs when physicians who provide most of their care are systematically 153 Chapter 6 less inclined to make a diagnosis of the more common forms of mental disorders (19). Since mental health service utilization is measured by these diagnoses and access to specialist care is contingent upon such diagnoses in primary care, it is not surprising that Chinese immigrants consistently turn out to be under-diagnosed, under-treated and under-utilizing services. According to Andersen's behavioural model, services are equitable if need factors determine the use of services (20-23). While this thesis shows that there is some equity in the mental health service system in B C in that mental health needs, measured by diagnosis of serious mental disorders in Chapter 4 and risk of depression in Chapter 5, are strong predictors of service utilization, it also demonstrates that a segment of the Chinese population may not be receiving services even when needs exist. With respect to health policy and service delivery, the explanations for lower rates of service utilization each carries different implications. At first glance, under-utilization of mental health services due to less need is a positive sign that services are being accessed appropriately. On the other hand, the notion that immigrants experience lower levels of psychological distress seems to contradict what has been reported in the literature (12). It is likely that immigrants express their distress in ways that are not captured by formal psychiatric diagnostic systems; the relatively high proportion of visits in primary care seen in Chapters 2 and 4 for the vague diagnostic category of anxiety/depression attests to that possibility. The implications are twofold. On one hand, clinicians will have to be better prepared to identify these atypical symptoms of distress and provide appropriate intervention. On the other, assessment of mental health needs in a community cannot rely solely on existing psychiatric diagnostic criteria and must take into account the subjective experience of distress and burden of illness. Lower utilization due to cultural and other barriers is frequently discussed in the literature and a rallying point for immigrant and minority advocates. Solutions have to be multi-faceted. The bulk of mental health service providers are currently located within the medical sector, as a result of the health system's conceptualization of emotional problems and the health insurance scheme. While these providers are encouraged to develop culturally responsive programs and services, they are still dependent to a large extent on patients' initiative to present their complaints and to seek care. Discordant 154 Chapter 6 interpretations of the nature of the problem prevent individuals needing help from taking advantage of resources that are available. The health financing structure prevents the system from offering alternatives that are deemed appropriate. The barriers are deeply ingrained when even those born and raised in this country are reluctant to use available professional mental health interventions. BC's mental health plan states that care must be sensitive to the ethnocultural background of clients and service strategies are to be responsive to the values and beliefs of community members.1 To realize these goals, efforts to improve access will have to extend beyond the formal health sector and involve education, social and community services, and perhaps the complementary health practice sectors so that interventions for psychological difficulties can be provided early and in a milieu and manner acceptable to the individuals involved. The importance of language barrier cannot be overlooked. To overcome this barrier, many have advocated for recruiting bilingual clinicians and training members from minority groups to deliver services. However, this thesis reveals a paradoxical phenomenon that, for Chinese immigrants at least, receiving care from Chinese-speaking physicians is actually associated with lower rates of utilization and less likelihood of being diagnosed with some types of mental disorders. This thesis does not dispute the effectiveness of bilingual practitioners in facilitating access to care. Ethnically or linguistically matched providers are associated with lower dropout rate and higher utilization in community mental health settings and with patient satisfaction in general medical care (24-31). Without physicians who are able to communicate in their native language, Chinese immigrants may be even less likely to receive any mental health care or diagnosis. Instead, this thesis points out that the potential of these bilingual professionals in improving access to care is not fully utilized without paying attention to their practice behaviours. Cultural bias on the practitioners' part cannot be ruled out. Just as physicians of a different cultural background are susceptible to misinterpreting clinical presentations of Chinese and other minority patients, Chinese physicians may also overlook minor mental health symptoms in their patients as a result of their own perception of mental health or their disinclination to provide the necessary intervention 1 Revitalizing and Rebalancing British Columbia's Mental Health System. The 1998 Mental Health Plan. http://www.healthservices.gov.bc.ca/mhd/pdf7rnhpd.pdf 1 155 Chapter 6 (32, 33). As many researchers and practitioners in the mental health field have begun to propose, cultural competence of the providers and the service environment may be more important than the language medium or common heritage (27, 34-37). In addition to ensuring unimpeded communication between clinicians and patients, mental health services and those who provide them must also be sensitive to the cultural orientation of minorities and be equipped to respond to mental health issues. The competencies include an understanding of minorities' experience of mental health difficulties and the skills to deliver appropriate intervention. In terms of human resources, attention must'be paid to the training of physicians and mental health professionals, besides the recruitment of bilingual personnel. STRENGTHS AND WEAKNESSES OF RESEARCH DESIGN The research design in this thesis avoids some of the shortcomings in existing literature with respect to utilization of mental health services by minority groups. By focussing on one ethnic group, it minimizes confounding^effects of ethnic variation inherent in many studies that treat immigrants or ethnic minorities as a homogeneous entity. It is also able to isolate the effects of immigration by studying immigrants and Canadian-born Chinese separately, although the sample of the latter group is small. Secondary analysis of both administrative and survey databases is the primary methodology. The advantage of using available databases is the low cost, both in terms of time and money required to acquire data and produce results. The disadvantage is that data are not collected to address the specific questions of the research study and the answers that can be generated are limited. In addition, the administrative and survey databases used in this thesis have each their strengths and weaknesses. . The administrative databases provide comprehensive data that are not subject to recall biases. On the other hand, the quality and appropriateness of the data to answer the research questions vary. While the health databases are fairly accurate records of utilization of insured services, they do not indicate the prevalence of need or the outcome of utilization; nor do they contain information about use of non-medical or informal resources. The diagnostic information in the databases provides at best an estimate of the treated prevalence rate of disorders. Even then, diagnostic information, especially in 156 Chapter 6 physician payment files, is not validated and must not be interpreted as precise representation of incidence or prevalence of conditions. Diagnostic categories which share similar symptom profiles are probably used interchangeably by some physicians and distinctions between these categories are often blurred (38). The usefulness of the diagnostic information in this thesis is predicated on the assumption that inaccuracies in diagnosis are similarly distributed among immigrant and comparison subjects. By linking different databases, some of the deficiencies in the administrative databases can be overcome. For instance, although health databases do not contain information on patients' ethnicity or immigration status, linking them to the immigration database allows Chinese immigrants' utilization records to be extracted. Similarly, linking the health databases with the physician database allows the language skill of the physicians providing service to be identified. Data linkage, however, also has its caveats. Overall linkage rate for the Chinese immigrants in this thesis is 73%, with variation within the population. Appendix B presents the linkage rates by various immigrant characteristics. While many of the immigrants fail to be linked because they have left the province and are not eligible subjects for this thesis, it should be acknowledged that some immigrants who are still resident in BC are probably not linked for other reasons. These unlinked residents threaten to bias the results and these issues are discussed in Appendix C. In any case, the potential bias is estimated to be small and the effect generally is to attenuate the results. Therefore, it is not expected to alter the findings substantially. The survey database presents different strengths and weaknesses. In this thesis, it supplements information from the administrative database by allowing Canadian-born Chinese to be identified and including a measurement of depression. However, since the survey is not designed for studying minority populations, the sample of immigrant and Canadian-bom Chinese included among the respondents is too small to provide reliable answers to some of the research questions. Self-reported data, especially on sensitive ' topics such as mental health, are also subject to reporting biases. Discrepancies between self-reported health care utilization and administrative records are discussed in Chapter 5. Neither administrative data nor survey responses can be assumed to be more accurate than the other. While self-reported data are prone to errors, administrative data are also affected by restrictions in reporting systems (39,40). Telephone contacts and 157 Chapter 6 non-medical resources, for instance, are usually not included in the latter. The best choice of data source will have to depend on the purpose and population of the study. Agreement between data sources is generally higher for hospital care than for ambulatory visits whereas self-reports of the elderly and those with poorer health status are often found to be less reliable when compared to medical or administrative data (39,41-46). With respect to ambulatory mental health service utilization, agreement between survey and administrative data has been reported to be low to moderate in the general population and even lower among those with high level of distress (47). On the other hand, among those with serious mental illness, fairly high concordance between self-reports and administrative records has been observed for inpatient, emergency room and case management use over a recall period of one to nine months (48, 49). FUTURE DIRECTIONS This thesis is the first population-based study of mental health service utilization by a minority population in Canada. It has filled in some of the gaps in existing knowledge by quantifying the magnitude of disparities with respect to the largest visible minority population in BC. It has demonstrated that utilization varies within the population of Chinese immigrants and identified some of the sub-groups that are at high risk of mental health difficulties or of being under-served. It has examined some of the multitude of factors contributing to disparities and confirmed the roles of individual characteristics and contextual factors in health service utilization. Not only does it support Andersen's behavioural model of health care utilization by affirming the roles of individual and contextual predisposing, enabling and need factors in mental health service utilization, it has also introduced components salient to ethnic minority populations and mental health services (20-23). Future research on the subject can build on this foundation of knowledge. Issues that require clarification include the prevalence of mental health needs in minority populations, cultural factors associated with mental health, and the practice behaviour of physicians treating these populations. A l l three concern the need component in health services utilization. Prevalence of mental disorders relates to need on a population and contextual level. Recognizing that emotional difficulties are experienced and expressed 158 Chapter 6 differently in minority cultures, current psychiatric diagnostic instruments may not be appropriate measures of such needs in these cultures. Appropriate tools must take into account the cultural meaning of distress symptoms and the support systems available to respond to them (50). Cultural minorities' interpretation of mental health symptoms and their beliefs about effective remedies affect their perceived need on an individual level. They must be examined in order to develop new paradigms of mental health, mental illness and mental health services. In this regard, efforts should also be made to study the extent of use of alternative medicine and informal resources that minorities are more likely to turn to. With respect to the beliefs and values about mental health, it is imperative to examine also differences between immigrants and the Canadian-born. While both groups have been found to use few mental health services in this thesis, it should not be assumed that the dynamics behind the under-utilization are identical across generations. Understanding the generational differences will be informative as to the mutability or permanence of cultural factors relevant to access. Whereas minorities' understanding of mental health relates to perceived need for mental health services, the practice pattern of physicians providing care to minorities influences the evaluated need for individuals in this population. Although the results of this thesis are not conclusive, it seems likely that the cultural barriers that deter minorities from seeking formal mental health care may also impede physicians from identifying needs for such care. Intervention to reduce disparities in care will benefit from greater knowledge of the behaviours of these practitioners. While physicians who share the same cultural background as their patients are invaluable in a culturally responsive health care system, their role as providers and gatekeepers will be much enhanced by specific strategies that enable them to respond more effectively to mental health issues in their practices. Similar research into mental health service utilization should be replicated in other minority populations and geographical regions. Comparisons across different cultures will provide a more comprehensive picture of the interplay between immigration, cultural and health factors that influence health behaviours. Comparison with other geographical regions will contribute to greater understanding of the role of community structure and health service organization, including the availability of bilingual and 159 Chapter 6 bicultural health professionals, in minorities' health service utilization. Health behaviours such as use of health care undeniably involve a multitude of factors of which only a fraction is currently understood by academics and health administrators and professionals. Research into health behaviours in differing contexts serves to reveal more of the underlying dynamics and continually enrich the perspectives of the existing state of knowledge in the field. Finally, the program of research into equity in mental health for immigrants and minorities should also investigate health outcomes. Use of health services is a process indicator in Andersen's model of health care utilization, leading to health outcomes represented by health status and consumer satisfaction. Assumption cannot be made that utilization of medical care will bring about improved health. The correlation between the access to existing health services and evaluated and perceived mental health status is even less clear among minority groups, given the cultural differences in health care preferences and values. Knowledge of the health outcomes in different minority groups is essential to the development of a culturally responsive and effective mental health system. 160 Chapter 6 B I B L I O G R A P H Y 1. Cheung F K , Snowden L R . Community mental health and ethnic minority populations. Community Ment Health J 1990; 26: 277-291. 2. Harris K M , Edlund M J , Larson S. Racial and ethnic differences in the mental health problems and use of mental health care. Med Care 2005; 43: 775-784. 3. Intercultural Committee. Multiculturalism and Mental Health: Developing Culturally Competent Systems of Care. Vancouver: Greater Vancouver Mental Health Service: 1999. 4. Klimidis S, McKenzie DP, Lewis J, Minas M . Continuity of contact with psychiatric services: immigrant and Australian-born patients. Soc Psychiatry Psychiatr Epidemiol 2000;35:554-563. 5. Lai D W L , Tsang K T , Chappell N L , Lai D C Y , Chau SBY. Health and Well Being of Older Chinese in Canada. Calgary: Faculty of Social Work, University of Calgary: 2003. 6. Leong FTL. Asian Americans' differential patterns of utilization of inpatient and outpatient public mental health services in Hawaii. J Community Psychol 1994; 22: 82-96. 7. Matsuoka JK, Breaux C, Ryujin D H . National utilization of mental health services by Asian Americans/Pacific Islanders. J Community Psychol 1997; 25: 141-145. 8. Meinhardt K , Vega W ;. A method for estimating underutilization of mental health services by ethnic groups. Hosp Community Psychiatry 1987; 38: 1186-1190. 9. Peters R. Increasing the responsiveness of mainstream mental health services to ethnocultural minorities. In: Health and Cultures Vol . U Programs, Services and Care. Edited by Masi R, Mensah L , McLeod K A . Oakville: Mosaic Press: 1993: 187-203. 10. Roberts N , Crockford D. Psychiatric admissions of Asian Canadians to an adolescent inpatient unit. Can J Psychiatry 1997; 42: 847-851. 11. Snowden LR, Cheung F K . Use of inpatient mental health services by members of ethnic minority groups. A m Psychol 1990; 45: 347-355. 12. Review of the Literature on Migrant Mental Health. Canadian Task Force on Mental Health Issues Affecting Immigrants and Refugees: 1988. 13. Bowen S. Access to Health Services for Underserved Populations in Canada. "Certain Circumstances": Issues in Equity and Responsiveness in Access to Health Care in Canada. Ottawa, Ontario: Health Canada: 2000. 161 Chapter 6 14. Chen C-N , Wong J, Lee N , Chan-Ho M-W, Lau JT-F, Fung M . The Shatin Community Mental Health Survey in Hong Kong. Arch Gen Psychiatry 1993; 50: 125-133. 15. Hwu H-G, Yeh E-K, Chang L - Y . Prevalence of psychiatric disorders in Taiwan defined by the Chinese Diagnostic Interview Schedule. Acta Psychiatr Scand 1989; 79: 136-147. 16. Takeuchi DT, Chung R C - Y , Lin K - M et al. Lifetime and twelve-month prevalence rates of major depressive episodes and dysthymia among Chinese Americans in Los Angeles. A m J Psychiatry 1998; 155: 1407-1414. 17. WHO World Mental Health Survey Consortium. Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. J A M A 2004; 291: 2581-2590. 18. Clark L , Hofsess L . Acculturation. In: Handbook of Immigrant Health. Edited by Loue S. New York: Plenum Press: 1998: 37-59. 19. Miranda J, Lawson W, Escobar J. Ethnic minorities. Mental Health Services Research 2002; 4: 231-237. 20. Andersen R. A Behavioral Model of Families Use of Health Services. Center for Health Administration Studies, Research Series 25. The University of Chicago: 1968. 21. Andersen R, Newman JF. Societal and individual determinants of medical care utilization in the United States. Milbank Mem Fund Q 1973; 51: 95-124. 22. Andersen R M . Revisiting the behavioral model and access to medical care: Does it matter? J Health Soc Behav 1995; 36: 1-10. 23. Andersen R M , Davidson PL. Improving access to care in America. In: Changing the US Health Care System. Edited by Rice T H , Kominski GF. San Francisco: Jossey-Bass: 2001: 3-30. 24. Cooper L A , Roter D L , Johnson RL, Ford DE, Steinwachs D M , Powe NR. Patient-centered communication, ratings of care, and concordance of patient and physician race. Ann Intern Med 2003; 139: 907-915. 25. Flaskerud JH. The effects of culture-compatible intervention on the utilization of mental health services by minority clients. Community Ment Health J 1986; 22: 127-141. 26. Jerrell J M . Effect of ethnic matching of young clients and mental health staff. Cult Diversity Ment Health 1998; 4: 297-302. 162 Chapter 6 27. Maramba G G , Hall G C N . Meta-analyses of ethnic match as a predictor of dropout, utilization, and level of functioning. Cultural Diversity and Ethnic Minority Psychology 2002; 8: 290-297. 28. Saha S, Komaromy M , Koepsell TD, Bindman A B . Patient-physician racial concordance and the perceived quality and use of health care. Arch Intern Med 1999; 159:997-1004. ' 29. Sue S, Fujino DC, Hu L , Takeuchi DT, Zane NWS. Community mental health services for ethnic minority groups: A test of the cultural responsiveness hypothesis. J Consult Clin Psychol 1991; 59: 533-540. 30. Takeuchi DT, Sue S, Yeh M . Return rates and outcomes from ethnicity-specific mental health programs in Los Angeles. A m J Public Health 1995; 85: 638-643. 31. Zane N , Hatanaka H , Park SS, Akutsu P. Ethnic-specific mental health services: Evaluation of the parallel approach for Asian-American clients. J Community Psychol 1994; 22: 68-81. 32. Borowsky SJ, Rubenstein L V , Meredith LS , Camp P, Jackson-Triche M , Wells K B . Who is at risk of nondetection of mental health problems in primary care? J Gen Intern Med 2000; 15: 381-388. 33. Snowden LR. Bias in mental health assessment and intervention: theory and evidence. A m J Public Health 2003; 93: 239-243. 34. Snowden L R , Hu T-W. Outpatient service use in minority-serving mental health programs. Adm Policy Ment Health 1996; 24: 149-159. 35. Sue S. Psychotherapeutic services for ethnic minorities: Two decades of research findings. A m Psychol 1988; 43: 301-308. 36. Takeuchi DT, Mokuau N , Chun C-A. Mental health services for Asian Americans and Pacific Islanders. J Ment Health Adm 1992; 19: 237-245. 37. Yeh M , Takeuchi DT, Sue S. Asian-American children treated in the mental health system: A comparison of parallel and mainstream outpatient service centers. J Clin Child Psychol 1994; 23: 5-12. 38. Jones W. Cpding of Depression by Physicians in B.C. Mental Health Evaluation and Community Consultation Unit, University of British Columbia: October 2002. 39. Raina P, Torrance-Rynard V, Wong M , Woodward C. Agreement between self-reported and routinely collected health-care utilization data among seniors. Health Serv Res 2002; 37: 751-774. 163 Chapter 6 40. Tisnado D M , Adams JL, Liu H et al. What is the concordance between the medical record and patient self-report as data sources for ambulatory care? Med Care 2006; 44: 132-140. 41. Geary PD, Jette A M . The validity of self-reported physician utilization measures. Med Care 1984; 22: 796-803. 42. Glandon GL, Counte M A , Tancredi D. An analysis of physician utilization by elderly persons: Systematic differences between self-report and archival information. J Gerontol 1992; 47: S245-S252. 43. Ritter PL, Stewart A L , Kaymaz H , Sobel DS, Block D A , Lorig K R . Self-reports of health care utilization compared to provider records. J Clin Epidemiol 2001; 54: 136-141. 44. Roberts RO, Bergstralh EJ , Schmidt L , Jacobsen S J. Comparison of self-reported and medical record health care utilization measures. J Clin Epidemiol 1996; 49: 989-995. 45. Rbzario PA, Morrow-Howell N , Proctor E. Comparing the congruency of self-report and provider records of depressed elders' service use by provider type. Med Care 2004; 42: 952-959. 46. Wallihan D B , Stump TE, Callahan C M . Accuracy of self-reported health services use and patterns of care among urban older.adults. Med Care 1999; 37: 662-670. 47. Rhodes A E , Lin E, Mustard C A . Self-reported use of mental health services vs.. administrative: Should we care? Int J Methods Psychiatr Res 2002; 11: 125-133. 48. Cheung A H , Dewa CS, Wasylenki D. Economic grand rounds: impact on cost - estimates of differences in reports of service use among clients, caseworkers, and hospital records. Psychiatr Serv 2003; 54: 1328-1330. 49. Hennessy K D , Reed SK. Validating self-reports of mental health service use in a chronic population. J Nerv Ment Dis 1992; 180: 399-400. 50. Kirmayer L i , Groleau D. Affective disorders in cultural context. The Psychiatric Clinics of North America 2001; 24: 465-478. • i 164 Appendix A DATA SOURCES TN LINKED DATABASES This appendix describes the administrative databases which are the sources of data for the analyses in Chapters 2 to 4. Landed Immigrant Data System (LIDS) The database from Citizenship and Immigration Canada (CIC) consists of individual information on all immigrants who landed between 1985 and 2000. Those who stated that their destination was British Columbia (BC) were linked to the B C Linked Health Data as part of a national immigrant health project. Access to this database was granted by CIC. Identifying information such as names and visa numbers was not released to the researcher. The data fields available to this project were: Date of birth Sex Educational qualification (2) Years of schooling Skill level (2) • Arrival date • Landing date • Immigration category (4) • Applicant family status (2) • Destination province, census metropolitan area and community (3) • Country of birth • Country and region of last permanent residence (3) • Marital status (2) Some data fields contain aggregated values of another data field or different coding methods. The data fields are grouped by the characteristic they describe and the number of fields relevant to that characteristic, i f more than one, is shown in parentheses. Intended occupation (4) Official language ability BC Linked Health Data (BCLHD) The B C L H D is a linked database of most administrative health records in the province of BC. Databases used in this project were the Medical Service Plan (MSP) Registry, the Vital Statistics-Death, the MSP Payments, the Hospital Separations, the Mental Health Services and the Pharmacare files. Access to data for the years 1985 to 165 Appendix A 2001 was granted by the Ministry of Health Services and Ministry of Health Planning (now the Ministry of Health). Identifying information such as names and personal health numbers was not released to the researcher. Descriptions of each database follow. MSP Registry The M S P is the publicly-funded provincial health insurance plan which covers all B C residents. Files for each year from 1986 to 2001 released for this project contained individual level registration data with the following data fields: • Birth date (month and year) • Postal code (first 3 digits) • Sex • Census metropolitan area • Day registration started for that year • Census division/subdivision • Number of days registered in that year • Census tract • Local health area and health authority • Socioeconomic quintile and (2) decile (2) Vital Statistics Death The only data field from this file for this project was the month and year of death of study subjects which occurred between 1985 and 2001, if applicable. MSP Payments These files consist of all payments in B C made to health practitioners for services covered by the health plan. Each entry represents one payment for a service or correction for a payment; one visit to the doctor may generate more than one entry. A study on physician resources reported that 96.0% of payments to general practitioners and 67.5% of payments to psychiatrists in the fiscal year 1996/97 were through such fee-for-service payments (1). The Payments file is thus expected to contain records of most of the general practitioner services rendered to B C residents and a smaller percentage of psychiatric services. The remainder of the physician payments is made through salary or sessional or service agreements and most of the utilization of psychiatric services remunerated this way is covered in the Mental Health Services database. Although the fee-for-service Payments file included only about approximately 2/3 of the contacts with psychiatrists in this project, the extent of the under-representation was expected to be the same between immigrants and the comparison group and did not affect the magnitude of the relative difference between immigrants and the comparison group, which was the 166 Appendix A objective of this project. Files for each fiscal year from 1985/86 to 2001/02 were released for this project. Data fields available were: • Birth date (month and year) . • Sex • Premium subsidy code • Date of service • Scrambled practitioner number • Specialty of practitioner • Scrambled number of referring practitioner Not all data fields were available in each year. For instance, diagnostic information was not entered until 1991/92. Hospital Separations The Hospital Separations files contain records of all separations from acute care and extended care facilities in B C , including day procedures. Files for each fiscal year from 1985/86 to 2001/02 were released for this project. Data fields available included: Diagnosis Fee item (specific service) Service units paid Type of service Location of service Location of practitioner Postal code (first 3 digits) and province of patient (2) Sex Age and age group (13) Admission date and separation date (2) Patient service and service group (2) Level of care Postal code (first 3 digits) In-hospital transfers - services and days (6) Transfer between hospitals (4) Province issuing health coverage Admission category Hospital code Length of stay (3) Scrambled number of most responsible physician Specialty of most responsible physician Location of service Use of occupational therapy Diagnoses (16) and diagnosis types(16) Diagnostic class and diagnostic short list (2) Pre-admission comorbidity Procedures (12) and dates (12) 167 Appendix A Entry code and exit code (2) Death codes (5) Hospital size Days in chronic behaviour disorder unit Assessment for long term care Alternate level of care days E codes and injury code (3) Procedures short list Use of ambulance Case mix indicators (8) Out of province hospitals -location and hospital code (2) Local health area Patient's residence in B C Days in acute care or rehab care • Tertiary care code The data fields in each year might be slightly different, owing to changes in record-keeping practices. Mental Health Services The Mental Health Services file contains records of activations in BC's mental health system. Each activation represents an episode of care in one of the system's inpatient or outpatient programs and may contain numerous contacts over a lengthy period of time. Information on individual contacts is not available. Data released for this project included: Employment status Marital status Living situation • Location of program • First and last contact dates (2) • Referral source • Diagnosis (3) Pharmacare The Pharmacare files contain information on all claims for medications dispensed outside of hospitals and which are covered by the publicly-funded Pharmacare program. Only certain population groups in B C receive comprehensive coverage of their medications through one of the Pharmacare plans. Before 2003, they included people aged 65 or over (Plan A), residents of long-term care facilities (Plan B), recipients of provincial income assistance (Plan C) and severely handicapped children in a community-based program (Plan F). In addition, patients with financial need also receive psychiatric medications for free when they register at community mental health services (Plan G). Most B C residents not in one of these plans are in Plan E and they receive 168 Appendix A coverage for medications only when their expenses exceed a pre-determined threshold. Drug prescription records for individuals on Plan E appear in the Pharmacare database only when the threshold is reached. Files for each year from 1985 to 2001 were released for this project. Data fields available included: • Pharmacare plan type • Product selection • Sex • Special authority codes (4) • CDIC code and therapeutic code (2) • Subsidy indictor • Date dispensed • Administrative codes (2) • Quantity and days of supply • Scrambled number of dispensed (2) prescribing practitioner • Quantity and days of supply accepted for payment (2) Not all data fields were available in each year. The number of days supply of medications, for example, was not available till 1996. College of Physicians and Surgeons of BC (CPSBC) Register The register of physicians in BC from 1985-2001 consists of directory active physicians. Permission to use the register for this thesis research was granted by CPSBC. The following data were obtained for each year from 1985 to 2001 on each general practitioner, family practitioner and psychiatrist registered: surname, given names, place of graduation and language ability. From this file, physicians who were able to communicate in Chinese were identified. The list of Chinese-speaking physicians was then linked to the B C Linked Health Data. A list of the scrambled physician numbers used in the data files for this project was then released to the researcher'to identify the Chinese language ability of physicians providing service to subjects in the project. The procedures used to determine the Chinese language skill of physicians are described in Appendix H . Data Linkage Linkage of all databases was performed by the Centre for Health Services and Policy Research at the University of British Columbia. The Landed Immigrant Data 169 Appendix A \ System was linked to the B C Linked Health Data by surname, given names, date of birth, sex and intended destination, using probabilistic linkage methods (2). After linkage, identifying information such as names, visa numbers and personal health numbers were removed and replaced by a study identification number. Records in the health databases of the linked immigrants were extracted and released to the researcher with the study identification number. Successful linkage rate for Chinese immigrants from China, Taiwan, Hong Kong and Macau was 73% overall. See Appendix B for linkage rates by various demographic characteristics. The College of Physicians and Surgeons register was linked to the B C Linked Health Data deterministically with each physician's MSP billing number. After linkage, the billing numbers were removed and only scrambled identification numbers were released to the researcher. References 1. Kazanjian A , Reid RJ, Pagliccia N , Apland L , Wood L. Issues in Physician Resources Planning in B.C.: Key Determinants of Supply and Distribution, 1991-96. H H R U 00:02. Vancouver: Centre for Health Services and Policy Research, University of British Columbia: 2000. 2. Newcombe HB. Handbook of Record Linkage: Methods for Health and Statistical Studies, Administration, and Business. New York: Oxford Medical Publications: 1988. 170 Appendix B LINKAGE RATES OF CHINESE IMMIGRANTS BETWEEN IMMIGRATION DATABASE AND H E A L T H DATABASE Of 213,809 records in the Landed Immigrant Data System (LIDS) of irrimigrants whose country of birth and country of last permanent residence were either Hong Kong, China, Taiwan or Macau, 155,988 records were successfully linked to the B C Linked Health Data, including some duplicate linkages. The overall linkage rate was 72.96%. The following are breakdowns of the linkage rates by sex and age group, landing year, marital status, education, official language ability, country of birth and country of last permanent residence. Table B.1 Linkage Rates by Age Group and Sex Male Female Unknown Total 0-14 73.02% 72.56% 0.00% 72.80% 15-24 74.92% 72.66% 0.00% 73.70% 25-44 71.87% 68.53% - 70.05% 45-64 81.26% 75.22% 100.00% 78.30% 65+ 83.30% 81.10% - 82.15% Unknown - 0.00% - 0.00% Total 74.61% 71.43% 33.33% 72.96% Table B.2 Linkage Rates by Landing Year and Sex Male Female Unknown Total 1985 62.14% 52.92% - 57.10% 1986 73.39% 61.95% - 67.20% 1987 70.27% 61.91% 100.00% 65.82% 1988 64.53% 56.18% - 60.11% 1989 67.62% 58.70% 0.00% 62.92% 1990 70.06% 59.69% - 64.74% 1991 73.38% 66.37% - 69.83% 1992 73.29% 67.57% - 70.29% 1993 78.80% 73.44% - 76.00% 1994 82.77% 79.62% 0.00% 81.13% 1995 81.94% 80.07% - 80.96% 1996 78.26% 77.38% - 77.81% 1997 73.82% 74.62% - 74.24% 1998 71.84% 72.33% 72.10% 1999 67.04% 68.73% - 67.92% 2000 66.00% 67.31% - 66.67% Total 74.61% 71.43% 33.33% 72.96% 171 Appendix B Table B.3 Linkage Rates by Marital Status and Sex Male Female Unknown Total Single 72.17% 71.54% 0.00% 71.85% Married 76.78% 70.43% 100.00% 73.49% Other 75.21% 79.96% - 79.13% Unknown 50.00% 57.14% - 53.85% Total 74.61% 71.43% 33.33% 72.96% Table B.4 Linkage Rates by Education and Sex Male Female Unknown Total 0 to 9 years of schooling 75.46% 73.89% 50.00% 74.62% 10 to 12 years of schooling 79.22% 73.04% 75.74% Some post-secondary 76.96% 69.61% 0.00% 72.91% University degrees 67.44% 65.53% 66.65% Total 74.61% 71.43% 33.33% 72.96% Table B.5 Linkage Rates by Official Language Ability and Sex Male Female Unknown Total English 72.79% 69.11% 0.00% 70.99% No English 76.09% 72.94% 50.00% 74.38% Unknown 100.00% 100.00% 100.00% Total 74.61% 71.43% 33.33% 72.96% Table B.6 Linkage Rates by Country of Birth and Sex Male Female Unknown Total Hong Kong 73.93% 70.75% 0.00% 72.26% China 72.88% 69.80% 50.00% 71.27% Taiwan 77.92% 74.82% - 76.34% Macau 79.24% 74.78% - 76.92% Total 74.61% 71.43% 33.33% 72.96% 172 Appendix B Table B.7 Linkage Rates by Country of Last Permanent Residence and Sex Male Female Unknown Total Hong Kong 75.76% 67.76% 78.31% 83.35% 71.86?/o 66.98% 74.83% 76.37% 50.00% 0.00% 73.72% 67.34% 76.56% 79.78% China Taiwan Macau Total 74.61% 71.43% 33.33% 72.96% Linkage rates for male and female immigrants were 74.61% and 71.43% respectively. Older immigrants had higher linkage rates than younger immigrants, with the lowest rate found in the 25-44 age group for both males and females. Linkage rates also varied by landing year, with immigrants landing in 1994-1995 haying the highest linkage rates. Higher linkage was found among the separated, divorced and widowed (Other) individuals. Those with university degrees tended to have lower linkage rates. There were minimal differences by ability to speak English. Immigrants from China were somewhat less likely to be linked than those from the other regions. The lower linkage rate of the 25-44 age group was possibly due to a larger proportion of the working age adults moving out of the province for employment opportunities elsewhere. Similarly, those with university degrees were also more likely to do the same. Older adults, on the other hand, were less mobile. The higher linkage of the separated, divorced and widowed could probably be attributed to their older age. Implications of the differential linkage rates on the validity of the research depend on the reasons for failure to link. There is evidence to support that many of those who are not linked likely moved out of the province before registering in the health plan and, therefore, have no effect on the research results., Issues related to data linkage rates are discussed in Appendix C. 173 Appendix C DATA LINKAGE ISSUES One of the potential limitations with regard to the linked immigration and health administrative database concerns the relatively low linkage rate (73%). The implications of non-linkage, the possible misclassification of comparison group members resulting from non-linkage and the estimated magnitude of each problem are discussed in this appendix. Non-linkage of Immigrants Who Landed in BC The reason that the remaining 27% of Chinese immigrants to British Columbia (BC) fail to be linked affects the validity of the findings. Following are the possible scenarios and their associated effects on the results. REASON FOR NON-LINKAGE EFFECT ON RESULTS A . Immigrant moved out of province Although immigrants who left B C might or out of country before registering have different utilization patterns had they in health plan stayed, the results of this study are representative of the immigrant population who chose to stay. B. Immigrant is in province but is not Over-estimates utilization rates of registered in health plan (and immigrants since a sub-population who therefore does not use health does not use health services is excluded services) from analysis C. Immigrant is registered in health Increases probability of misclassification plan but linkage fails for other of comparison group members since the reasons (e.g. discrepancies in name, immigrant becomes part of the general date of birth) population eligible for selection into the comparison group 174 Appendix C A Statistics Canada report on the Longitudinal Survey of Immigrants to Canada provides some clues as to the reason for non-linkage.1 The study traced immigrants who arrived between October 2000 and September 2001 about six months after landing. Of those immigrants who stated their intention to live in Vancouver on arrival, only 83.5% were found in the destination city six months later and 10.6% were located in Toronto or Montreal; another 5.9% were in other unspecified cities.2 Although the trace rate for this study was only 72% and involved a different cohort, this study suggested that a substantial number of immigrants to B C failed to be linked to provincial health plan records because they did not settle in the province. The actual percentage of out of province Chinese immigrants would likely be higher if those who moved to other cities outside of B C and those who moved out of the country were included. These immigrants would not affect the findings of this thesis project which focussed on those residing in BC. Coverage of Chinese Immigrants in BC Inter-provincial migration, however, creates an additional problem for this project - the problem of coverage. Chinese immigrants who intend to live in other provinces but who then decide to move to B C are not covered in the linked immigration database of this project. Even though they are expected to exhibit the same health service utilization patterns as the study immigrants, they are actually included in the general B C population and eligible to be selected as comparison subjects. Thus, the comparison group in this study is possibly contaminated by some recent Chinese immigrants who are unidentified because of linkage problems or migration from other provinces. The extent of the misclassification of Chinese immigrants as comparison subjects depends on several factors including in-migration from other provinces and the composition of the subject pool from which comparison subjects are selected. ' Longitudinal Survey of Immigrants to Canada: Process, progress and prospects. Statistics Canada Catalogue no. 89-611-XIE, October 2003. 2 ' Caveats about the Longitudinal Survey of Immigrants include the fact that it involves a somewhat different landing cohort in a different time period than the immigrant health project; in addition, the trace rate of the sample was only 72% and the overall response rate was 59.2%. 175 Appendix C A Citizenship and Immigration Canada report on inter-provincial migration of immigrants provides a glimpse into the coverage of the linked immigration database.3 The study tracked immigrants who landed between 1980 and 1995 over that time period by their tax returns. Of the 194,565 immigrants who intended to live in B C , only 90.6% remained in that province in 1995; in the meantime, B C received 52,675 immigrants from other provinces. Hence, in 1995, those who stated their intention to be in B C composed only 77% of all that cohort of immigrants found in B C The probability of these unidentified immigrants being selected as comparison subjects or the proportion of the comparison group being misclassified immigrants varies with the size of the immigrant population in the area from which comparison subjects are selected. The higher the concentration of immigrants in an area, the higher the probability that they are included in the comparison group. Assuming that only 77% of recent immigrants can be identified from the linked immigration database and using the 2001 census data for metropolitan Vancouver, a crude estimate of the percentage of misclassified recent Chinese immigrants for this project is derived as follows: • Population in Vancouver = 1,967,475 Total Immigrants Chinese Immigrants* Non-Chinese Immigrants Number who arrived 1991-2001 324,815 147,135 177,680 Number who arrived 1981-1990 142,580 44,535 98,045 Estimated number who arrived 1985-2001 324815+(142580x0.6) = 410,363 147135+(44535x0.6) = 173,856 177680+(98045x0.6) = 236,507 Estimated % of 1985-2001 immigrants in population 410363/1967475 = 20.9% 173856/1967475 = 8.8% 236507/1967475 = 12.0% immigrants born in China, Hong Kong, Taiwan or Macau 3 The Interprovincial Migration of Immigrants to Canada. IMDB Profile Series. Citizenship and Immigration Canada, January 2000. 176 Appendix C • Estimated % of true comparison population (non-immigrants or irrimigrants before 1985) in Vancouver = 79.1% • Estimated % of unlinked Chinese immigrants (included in comparison group) = 23% of 8.8% = 2.0% • Estimated % of unlinked non-Chinese irrimigrants (included in comparison group) = 23% of 12.0% = 2.8% The following graph illustrates the distribution of study immigrants, study comparison subjects and misclassified comparison subjects in the population. Figure C.1 Percentage of Chinese Immigrants, Immigrants and Comparison Subjects in Metropolitan Vancouver S Misclassified Chinese immigrants f2 Linked Chinese immigrants EH Linked non-Chinese immigrants M Misclassified non-Chinese immigrants • Comparison population • % of population eligible for selection as comparison subjects (true comparison subjects + misclassified immigrants) = 79.1% + 2.0% + 2.8% = 83.9% • % of Chinese immigrants in comparison group = 2% out of 83.9% = 2.4% Based on the above estimate, approximately 2.4% of the comparison subjects in this thesis research are misclassified recent Chinese immigrants. However, the estimate does not take into account variations in population composition by sex, age, year of landing or local health area, which are the bases on which the comparison group is selected. Moreover, the figures on inter-provincial migration of immigrants are based on tax returns, essentially excluding children and those with no income. Neither is 177 Appendix C information on migration of immigrants available by ethnicity. A precise estimate of the extent of misclassification is therefore not possible. In any case, misclassification generally reduces the effect of any associations in studies and is not a major threat to the validity of findings. 178 Appendix D DATA CLEANING AND PREPARATION IN LINKED DATABASES Original data files received for this thesis project from the Centre for Health Services and Policy Research included data of immigrants from Hong Kong, China, Taiwan, Macau, Singapore and Malaysia. Initial validation and cleaning were performed on all the data. Data specific to the study population were then extracted. This appendix describes the actions taken prior to data analysis. Landed Immigrant Data System (LIDS) • Total number of immigration records including those from Singapore and Malaysia was 169,281. • Data were checked for quality. • Duplicate linkages (more than one immigration record linked to a B C Linked Health Data registrant) were extracted for further review. A total of 980 duplicates involving 1960 records were extracted. These records were manually reviewed and it was decided that duplicates which shared the same demographic characteristics would be considered the same individuals apd retained for analyses. Demographic characteristics used to compare duplicates were sex, date of birth, landing date, country of birth, country of last permanent residence, education, marital status, official language skills. The first records of 101 duplicate pairs which shared the above characteristics were returned to the database. The second record of duplicate pairs and duplicates with different demographic features were deleted. The resulting immigration database consisted of 167,422 unduplicated records. • Immigrants whose country of birth and country of last permanent residence were both in Hong Kong, China, Taiwan or Macau were extracted for the study, yielding a study population of 154,242. i Medical Services Plan (MSP) Registry • Data were checked for quality. 179 Appendix D • Approximately 0.1% of linked immigrants from Hong Kong, China, Taiwan or Macau (167) had no MSP registration in 1986-2001; hence, information such as sex, date of birth and postal codes was also missing. • Of the Chinese immigrants in the study with no missing sex or date of birth in the registry, there were discrepancies in sex information between the MSP registry and the LIDS for 0.36% (550) of them and discrepancies in years of birth for 0.22% (346) of them. The discrepancies were evenly distributed in both directions. Only three of the discrepancies in year of birth were more than five years and these were manually reviewed. Tables D. 1 and D.2 describe in details the breakdown of sex and year of birth differences. Table D.1 Comparing Sex in Immigration Database and Health Database ( Sex in MSP Registry Sex in LIDS Female Male Unknown Missing TOTAL Female 78157 272 6 85 78520 (50.67%) (0.18%) (0.00%) (0.06%) (50.91%) Male 278 75355 6 82 75251 (0.18%) (48.86%) (0.00%) (0.05%) (49.09%) Unknown 0 1 (0.00%) 0 0 1 (0.00%) TOTAL 78435 75628 12 167 154242 (50.85%) (49.03%) (0.01%) (0.11%) (100%) V 180 Appendix D Table D.2 Comparing Year of Birth in Immigration Database and Health Database MSP Registry Year Number of Percentage minus LIDS Year Immigrants -10 1 0.00% -5 13 0.01% -4 9 0.01% -3 16 0.01% -2 32 0.02% -1 98 0.06% 0 153729 99.78% 1 84 0.05% 2 34 0.02% 3 35 0.02% 4 7 0.00% 5 15 0.01% 14 1 0.00% 100 1 0.00% • MSP payments of those with discrepant sex information were checked for "female" procedures, e.g. pelvic examinations, pregnancy-related services, abortion. It was determined that registry data contained fewer errors; therefore, decision was made that all immigrants were to be assigned the sex and date of birth information in the MSP registry. The LIDS information was used only when such information was missing in the registry files; however, this was immaterial since subjects with no registration in the study period were subsequently excluded from all analyses in the research. • The sex and age match between immigrants and comparison subjects was checked. Of the matched pairs, 0.05% (82) had different registry year of birth and 0.02% (37) had different sex. It was decided that these pairs would be excluded from the study. • The match in local health area (LHA) between immigrants and comparison subjects was checked. The comparison subjects were to be matched to the immigrants by the first L H A after the immigrants' landing; however, over 3% were found to have different LHAs. 181 Appendix D Observation time Since it could not be presumed that all subjects were resident in the province throughout the study period and therefore "at risk" of using health care, decision had to be made about how to define the observation time of each subject. While M S P registration could be taken as a positive indication of an individual's presence in the province, consideration was made of the mandatory 3-month waiting period for health plan registration that all immigrants and newcomers to the province had to undergo. The interval between immigrants' landing dates and their first registration in M S P was examined, as well as number of MSP payments and hospitalizations that occurred before and during this interval. With the exception of the 167 immigrants who had no registration in 1986-2001, 8.5% of the immigrants registered in M S P before their landing dates; the majority (74%) registered within the 3-month period after landing; another 12% registered between 4 to 12 months after landing; the remaining 5.5% registered more than a year after landing. Table D.3 provides a detailed breakdown of time lag between landing date and M S P registration date. Table D.3 Time Difference Between MSP Registration and Landing Date MSP Registration to Number of Percentage Landing Date Immigrants 10+ years before 49 . 0.03% 6-10 years before 539 0.35% 3-6 years before 2213 , 1.44% 1 -3 years before 5649 3.67% 4-12 months before 3600 2.34% 1 -3 months before 999 0.65% 0-3 months after 114621 74.44% 4-12 months after 17788 11.55% 1 -3 years after ; 6363 4.13% 3-6 years after 1683 ' 1.09% 6-10 years after 385 0.25% 10+ years after 79 ; 0.05% 182 Appendix D Since it was not possible to determine when an individual was living in the province without registration in cases of comparison subjects and of immigrants who did not immediately register after the waiting period expired, decision was made that "observation" would begin on the day of the first registration after an immigrant's landing and would take place only when the individual was registered. This was justified since the incidents of health care utilization outside of MSP registration were few. The immigrant's landing date was applied to the matching comparison subject. It was recognized that defining the observation time in this way would result in slight over-estimation of service utilization rates for both immigrants and the comparison group since individuals who did not take advantage of the public health plan, likely non-users or low-users of health care, were not counted. There were other challenges in the calculation of registration time and the inclusion/exclusion of outcome events owing to the limitations of the registration data. The registry files available to this research project contained only information on the first day of registration and the total number of days of registration in each year. Exact dates of registration were not known when subjects had interrupted periods of registration in a year. Therefore, all registration time in this research project was assumed to be continuous from the first day of registration, i.e. a subject who was registered on day 32 for 89 days was assumed to have uninterrupted registration from February 1 to April 30 of the year. A small number of outcome events might have been incorrectly included/excluded in cases where the assumption was not true. Overall, the number of events that occurred without valid registration based on this assumption was low. Vital Statistics-Death • Duplicates in the comparison group's file were removed. • Data were checked for quality. • Of the 1844 linked immigrants with documented death date, including those from Singapore and Malaysia, 1.95% (36) had MSP claims after the death date and 25.87% (477) had MSP registration more than one month after death date. Only 70% (1286) of the deaths were found in the hospital separations files. Of the 1317 hospital 183 Appendix D deaths, 30 were not substantiated by Vital Statistics data and one had a different death date. • In the comparison group, there were 5249 subjects with death records. Of these, 1.79% (94) had MSP claims after death and 16.13% (847) had MSP registration more than one month after death. Fifty-seven percent (2975) of the deaths were found in hospital separation files; 56 hospital deaths were not corroborated by Vital Statistics data and two had different dates. • A l l discrepant death cases were manually reviewed but it was not possible to determine if any one of the databases was more accurate in terms of death information. The death date in the Vital Statistics, i f present, was adopted as the end of observation for the subjects involved. M S P Payments • Data were arranged from fiscal years into calendar years. • Data were checked for quality. • Problems with diagnostic codes were noted, especially in relation to misplaced decimals resulting in "leading zeroes" in mental health diagnoses, i.e. diagnostic category of "308" was entered as "00308". Problem was associated with individual practitioners and was more prevalent among immigrants. Diagnoses of "00308" amounted to 15% of diagnoses of "308" among immigrants. It was decided that all diagnostic codes of "00290" to "00314" would be considered mental health diagnoses for this thesis research, since these five-digit codes were not valid ICD-9 or ICD-9-C M codes and the alternative interpretations ("002.9X" paratyphoid fever and "003.OX" salmonella gastroenteritis) were relatively unlikely. • Data for the study population were extracted and individual records were aggregated into visits. Each visit consisted of all the records pertaining to a practitioner and a patient in one day. The dates of visits were checked against MSP registration dates, landing dates and death dates and visits were classified into five categories according to the relationship between date of occurrence and the other dates: before landing, after landing but before first registration in MSP, after landing and with valid registration, after landing and first registration but not registered at time of visit, after 184 Appendix D death. The landing date of each immigrant was applied to the matched comparison subject. Table D.4 summarizes the results of the types of physician visits in relation to landing and M S P registration. Table D.4 MSP Visits and Health Plan Registration of Immigrant and Comparison Subjects in 1986-2001 Immiarant Comparison Total Number of MSP Visits 1986-2001 6,423,770 11,170,169 Before landing 2.28% 3.53% After landing Before first MSP registration 0.01% 0.04% With valid MSP registration 97.04% 95.27% No MSP registration 0.67% 1.16% After death <0.005% <0.005% In 1986-2001, over 95% of MSP visits by immigrants and comparison subjects occurred after the immigrants' landing and with valid MSP registration; only these visits were included in the analyses for this thesis research. Hospital Separations • Data were arranged from fiscal years into calendar years. • Data were checked for quality. Missing procedure dates were imputed as the date of the previous procedure and duplicated procedures were removed. • Admissions identified as "newborns" and related records were manually reviewed; eight records of children bom to immigrant mothers were removed and maternal records were retained. • The hospitalization dates were checked against M S P registration dates, landing dates and death dates. Hospitalizations were classified into seven categories according to the relationship between date of admission and separation and other dates: before landing (separation date before landing date), during landing (admission date before landing date but separation date after landing), before first MSP registration (admission and separation dates after landing but before first MSP registration), after 185 Appendix D landing and with valid registration (any MSP registration between the admission and separation dates), after landing and registration but not registered at time of hospitalization (no MSP registration for the duration of hospitalization), after death (admission date after the death date), not known (not determined if there was MSP registration during hospitalization). The last category involved hospitalizations where admission and separation were in different years; if separation was in a year when there was no MSP registration or separation occurred before registration began, it was not determined whether there was valid MSP registration in prior years after admission. The landing date of each immigrant was applied to the matched comparison subject. Table D.5 summarizes the types of hospital separations in relation to landing and MSP registration. Table D.5 Hospital Separations and Health Plan Registration of Immigrant and Comparison Subjects in 1986-2001 Immiarant Comparison Total Number of Hospital Separations 1986-2001 67,618 151,871 Before landing 3.72% 3.62% During landing 1 case 0 After landing Before first MSP registration 0.38% 0.05% With valid MSP registration 95.14% 95.15% No MSP registration ; 0.75% 1.16% After death 2 cases 9 cases Not known 2 cases 17 cases • Overall, over 95% of hospital separations by immigrants and comparison subjects occurred after the irrimigrants' landing and with valid MSP registration; only these records were included in the analyses for this thesis research. 186 Appendix D Mental Health Services • Data were checked for quality. A large percentage of records had no or deferred or missing diagnoses - 73% for immigrants and 45% for the comparison group. • Records were separated into inpatient and outpatient activations by location code. • First and last contact dates for each episode were revised to comply with landing, MSP registration and death dates and study period (1992-2001). Records with no last contact date were presumed to be active at the end of 2001. Activations that occurred before landing, outside of the study period, after death or when there was no valid MSP registration were removed. • Individual number of inpatient episodes was tallied to be included in counts of mental health hospitalizations. • Overlapping outpatient episodes were collapsed across all sites or programs. Length of activation time for each collapsed outpatient episode was calculated; length of activation was the time during the study period between first and last contact dates when there was valid MSP registration. Total length of outpatient activation for each subject was then calculated. Pharmacare • Data were checked for quality. Errors in date of dispensation were corrected. • Claims for psychiatric medications were extracted and records for medications of the same therapeutic class dispensed on the same day were combined. • The dates of dispensation were checked against MSP registration dates, landing dates and death dates. Claims were classified into five categories according to the relationship between date of dispensation and other dates: before landing, after landing but before first registration in MSP, after landing and with valid registration, after landing and registration but not registered at time of dispensation, after death. The landing date of each immigrant was applied to the matched comparison subject. Table D.6 summarizes the dates of prescriptions in relation to landing and MSP registration. 187 \ Appendix D Table D.6 Psychiatric Prescriptions and Health Plan Registration of Immigrant and Comparison Subjects in 1986-2001 Immiarant ComDarison Total Number of Psychiatric Prescriptions in 1986-2001 77530 710115 Before landing 0.45% 4.38% After landing Before first MSP registration 0.02% 0.03% With valid MSP registration 99.01% 94.90% No MSP registration 0.51% 0.70% After death 0.01% <0.005% In 1986-2001, at least 95% of psychotropic medications received by immigrants and comparison group members were dispensed after the immigrants' landing and with valid MSP registration; only these records were included in the analyses for this thesis research. 188 Appendix E SUBJECT SELECTION CRITERIA This appendix describes how study subjects were selected from two of the databases used in this thesis research. The first database was the immigration database and the second was the Canadian Community Health Survey database. Landed Immigrant Data System (LD3S) Selection of Immigrants For studies in Chapters 2 to 4, subjects were selected from the immigration database of immigrants who landed in British Columbia (BC) in 1985-2000. Data released to this thesis project consisted of those belonging to immigrants whose country of birth or country of last permanent residence was one of the following: China, Taiwan, Hong Kong, Macau, Singapore or Malaysia. An internet search of the respective census departments of the six jurisdictions yielded the following information in Table E. 1 about the proportion of ethnic Chinese in these regions: 189 Appendix E Table E.1 Percentage of Chinese in Selected Asian Populations Region Census Year Percentage of Chinese in Population China' 2000 Han Chinese = 91.6% Ethnic minority Chinese = 8.4% No information available on non-Chinese Taiwan2 2000 Foreign nationals = 1.8% No information available by ethnicity Hong Kong3 2001 Chinese = 94.9% Macau4 2001 Chinese = 96.8% (includes mixed descent) Singapore5 2000 Chinese = 76.8% Malaysia6 2000 Chinese = 26.0% According to a study with the Canadian Mortality Database which used surnames as the gold standard to determine ethnicity, using country of birth in China, Hong Kong or Taiwan as the criterion would identify 75% of all ethnic Chinese but include 4% non-Chinese (1). To optimize specificity in selecting ethnic Chinese in this research project, it was decided to include only immigrants whose country of birth as well as country of last permanent residence were from regions with high concentrations of ethnic Chinese, i.e., China, Taiwan, Hong Kong or Macau. In 2000, these territories accounted for 38%, of all immigrants to Vancouver. Since these four regions were inhabited predominantly by residents of ethnic Chinese origin and constituted the major source of immigrants to B C during the study period, it was estimated that the selection criterion would include most of the Chinese immigrants who came to BC. Applying the more restrictive criteria 1 National Bureau of Statistics of China http://www.stats.gov.cn/was40/detail?record=3&channelid=52984 2 Statistical Bureau, Republic of China http://eng.stat.gov.tw/ct.asp?xItem=8465&ctNode=1629 3 Demographic Statistics Section, Census and Statistics Department, Hong Kong http://www.censtatd.gov.hk/hong_kong_statistics/statistical_tables/index.jsp ?charsetID= 1 &subjectID= 1 &t ableID=139 4 Direccao dos Servicos de Estatfstica e Censos, Macao http://www.dsec.gov.mo/ 5 Singapore Department of Statistics http://www.singstat.gov.sg/keystats/c2000/topline2.pdf 6 Department of Statistics Malaysia http://www.sta tistics.gov.my/english/frameset_census.php?file=pressdemo 7 Citizen and Immigration Canada. Facts and Figures 2001. http://www.cic.gc.ca/english/pub/facts2001/4van-02.html 190 Appendix E of both country of birth and country of last permanent residence was intended to minimize the inadvertent inclusion of non-Chinese who were either born or lived in those regions. This project opted for higher specificity in subject selection in order to achieve estimates that would more accurately reflect the health service utilization behaviour of the target study population. Selection of Comparison Group For each immigrant linked, a comparison subject was randomly selected from the Medical Service Plan (MSP) registry according to the following criteria: • Excluding all immigrants in the LIDS database • Matched to the immigrant by sex, year of birth and first local health area after landing • Registered in MSP in the year that immigrant landed There was the same number of comparison subjects as immigrants. See Appendix C for estimated extent of misclassification in the comparison group. Exclusions A total of 155,988 linked immigration records were of individuals whose country of birth and country of permanent residence were both in China, Taiwan, Hong Kong or Macau. Of these, 1842 records (921 pairs) were found to be duplicate linkages; that is, two records were linked to an individual in the B C Linked Health Data. Based on the data available to the researcher, it was not possible to determine which of the duplicate linkages involved the same individual who had more than one landing record, and which involved different individuals who were considered equally reasonable matches according to the probabilistic linkage methods used (2). Assuming that at least one of the record in each duplicate pair was the correct match, it was decided that duplicates who shared the same demographic characteristics of sex, date of birth, landing date, country of birth, country of last permanent residence, educational attainment, marital status and official language ability would be regarded as one individual and retained in the analyses, since the research question concerned the relationship between these demographic characteristics and mental health service utilization. As a result, 96 pairs of duplicates (individuals) were retained and the rest of the duplicates were removed, resulting in a study population of 154,242 Chinese irrimigrants. 191 Appendix E Of the 154,242 immigrants, 107 were found to have comparison subjects who differed in sex or year of birth, contrary to the matching criteria. These immigrants were also excluded from analysis, leaving 154,135 individuals. Only comparison subjects matched to these immigrants were retained in the analyses. Since the study period was 1992-2001, immigrant and comparison subjects who had no MSP registration during the study period were also removed, resulting in a final study population of 152,184 immigrants and 151,303 comparison subjects. Additional exclusion criteria were applied in Chapters 2 and 4. In Chapter 2, the rates of mental health service utilization were compared between immigrants and their matched comparisons. To minimize the bias that could occur when comparing subjects observed during different points in time, the study included only the observation time and utilization events in years when both the immigrant and comparison subject in a matched pair were registered. Out of the 152,184 Chinese immigrants and 151,303 comparison subjects who were registered during the study period, 3211 (2.1%) immigrants and 2330 (1.5%) comparison subjects were eliminated because they had no common years of health plan registration in 1992-2001. Hence, the final study population for this study consisted of 148,973 pairs of immigrant and comparison subjects, or 297,946 individuals. In Chapter 4, subjects with severe and persistent mental illness were selected according to the following criteria which were adapted from another project (3). I. Any one of the following criteria: e) Two or more physician visits for a severe mental health condition in one year f) Any hospital charge where the most responsible diagnosis was a severe (mental health condition or the principal physician was a psychiatrist or the principal service was psychiatry or patient received electroconvulsive therapy g) Any mental health care activation where one of the diagnoses was a severe mental health condition in one year' h) Two or more prescriptions for medications used to treat severe mental health conditions in one year II. Meeting any one of the above criteria in at least two years during the study period A l l the records in the health databases, including events that occurred without health plan registration, were used to identify severe mental illness. The criteria were chosen to 192 Appendix E maximize specificity in identifying subjects with severe and persistent mental illness. It was recognized that the criteria would have excluded individuals who were resident in BC for fewer than two years, those who did not seek medical care and those who were treated with medications which had multiple indications for use (e.g. anti-convulsants). For the analysis on prescription of psychiatric medications in Chapter 4, only subjects who were enrolled in Pharmacare plans that fully covered psychiatric medications for the year were included, i.e. subjects in Plans A , B, C, F and G. A l l subjects who were in Plan E for all or part of the year were removed, even when they switched from Plan E to a full coverage plan during the year. This action was taken because the exact coverage dates of the individuals under most of the eligible plans were not known and the observation time for subjects who were covered only part of the year could not be determined. However, this action also assumed that individuals in the Pharmacare files who had no claims under Plan E were covered by one of the full coverage plans for the entire year, which was not necessarily the case. Since claims under Plan E appeared in the database only when the individual exceeded the minimum expense threshold for the year, some of the individuals retained in the analysis could have been under the full coverage plan for only part of the year but did not reach the expense threshold while under Plan E. Since observation time was based on health plan registration time for the entire year, the rate of prescription for these individuals would thus be under-estimated for that year. The number of instances of inadvertently including individuals on Plan E for part of the year is estimated to be small because the switch from Plan E to an eligible plan is expected to occur only once for most plans. For instance, individuals who turn 65 during the year are transferred into Plan A and will be under full coverage for the duration of the study. Canadian Community Health Survey (CCHS) Subjects for the study in Chapter 5 were the 18302 B C respondents for the CCHS Cycle 1.1. Lnmigrant and non-immigrant Chinese were identified by the self-reported ethnicity of ancestors, cultural and racial background and the language used for interview. A l l respondents in B C who reported Chinese in one of the above variables were identified as ethnic Chinese. Those born outside of Canada were considered 193 Appendix E immigrants (879) and those born in Canada (205) were the Canadian-born. Place of birth in China, Hong Kong or Taiwan was also investigated as an inclusion criterion for Chinese ethnicity but several of the respondents born in those regions were noted to report other ethnicities; therefore, it was decided that place of birth would not be used to define ethnicity. Comparison subjects consisted of all B C respondents who were not identified as ethnic Chinese and they were also classified as immigrants or Canadian-born. References 1. Sheth T, Nargundkar M , Chagani K , Anand S, Nair C, Yusuf S. Classifying ethnicity utilizing the Canadian Mortality Data Base. Ethn Health 1997; 2: 287-296. 2. Newcombe HB. Handbook of Record Linkage: Methods for Health and Statistical Studies, Administration, and Business. New York: Oxford Medical Publications: 1988. 3. Reid RJ, Barer M L , McKendry R et al. Patient-Focused Care Over Time: Issues Related to Measurement, Prevalence and Strategies for Imporvement Among Patient Populations. Vancouver: Centre for Health Services and Policy Research, University of British Columbia: 2003. 194 Appendix F STUDY VARIABLES IN LINKED DATABASES This appendix explains the variables from the linked administrative databases used in analyses in Chapters 2 to 4. Some variables are drawn directly from the databases while some are derived from a combination of other variables. The first section describes the common variables used to compare outcomes, stratify analyses or adjust for effects: immigration status, sex, date of birth, age and landing date. The second section describes the outcome variables in Chapters 2 to 4 and the third section the explanatory variables specific to Chapter 3. Common Variables Variable Description Immigration Immigrants were subjects identified from the immigration Status database. Comparison subjects were B C residents selected from the health plan registry but not found in the immigration database. Sex Sex was the value documented in the health plan registry. See Appendix D for issues regarding determination of sex of the subjects Date of birth Date of birth was the value documented in the health plan registry. See Appendix D for issues regarding determination of date of birth. Only month and year of birth were available in the registry and were used in calculation of age variables. In SAS, the software program used for most of the data analysis in this research, the date of birth became the first day of the month of birth. 195 Appendix F Variable Description Landing Date Landing date was found in the immigration database and was the date an immigrant officially landed as legal resident in Canada. Although some immigrants had resided in Canada under other visas prior to landing, observation of all immigrants began only after their official landing. An immigrant's landing date was also applied to the matched comparison subject to determine his/her observation time. Age at Landing Age at landing was the number of years between the date of birth described above and the landing date Age at Entry to Age at entry to study was based on date of birth described above, Study although the method of calculation varied in each study. In Chapter 2, it was the age at the end of the first calendar year during the study period following the immigrant's landing when both immigrant and comparison subject were registered in the health plan.1 In Chapter 3, it was the age on the first day during the study period the subject was registered in the health plan after he/she or the matched immigrant had landed. In Chapter 4, it was the age on the first day of registration after an immigrant's landing in the year during the study period that the subject first met the criteria of having a severe mental illness. Since immigrants and comparison subjects were matched by year of birth rather than month of birth, using age at the end of the calendar year was necessary to ensure that both in a matched pair would belong to the same age group in the stratified analyses 196 Appendix F Variable Description ^ Observation Observation time was the total number of years that a subject was time registered in the health plan during the study period; only the time between an immigrant's landing and date of death (if applicable) was included. The value of this variable varied in each study. In Chapter 2, observation began on the first day of health plan registration after an immigrant's landing in a year when both the matched immigrant and comparison subject were registered in the health plan and included only time in years when both were registered. In Chapter 3, observation began on the first day of health plan registration after an immigrant's landing. In Chapter 4, observation began on the first day of registration after an immigrant's landing in the year the subject first met the criteria of having a severe mental illness. For most subjects, the difference between the age at landing and age at entry to study was small since the majority of immigrants arrived during the study period and became registered in the health plan soon after the landing date. Outcome Variables Outcome variables involved in Chapters 2 to 4 were number of physician visits and mental health visits, number of mental health hospitalizations, length of outpatient mental health care and amount of psychiatric medications prescribed. Variable Description Physician visit Data source: Medical Services Plan (MSP) Payments files A visit was defined as all payments to a practitioner for a patient in one day; that is, each subject could have at most one visit to a practitioner in one day. Records where the only claim was for no-charge referral or where service units paid was negative were excluded. 197 Appendix F Variable Description Mental health Data source: MSP Payments files visit Visits were defined as above. Inclusion criteria for mental health visits: 1) A l l visits where specialty of practitioner was psychiatry, or 2) Diagnostic category was psychiatry (see Appendix G for psychiatric diagnostic categories), or 3) Fee item was psychiatry (see Appendix G for list of psychiatric fee items) In cases where multiple claims with different diagnoses or fee items were submitted by a practitioner for a patient in one day, the visit was considered a mental health visit if at least one entry met the criteria above. Only visits where the specialty of the physician was general practice or psychiatry were included in the analyses. Mental health Data source: Hospital Separations files, Mental Health Services hospitalization file Inclusion criteria for Hospital Separations files: 1) Most responsible diagnostic category was psychiatry (see Appendix G for psychiatric diagnostic categories), or 2) Patient service code was psychiatry, or 3) Physician specialty was psychiatry, or 4) Procedure performed included electroconvulsive therapy Inclusion criterion for Mental Health Services files: - Location of care episode was an inpatient facility. Each record in the data files represented one hospitalization episode, including day procedures. 198 Appendix F Variable Description Outpatient Data source: Mental Health Services file mental health Outpatient mental health care was the length of time a subject was care active in the Mental Health Services database where location was not an inpatient facility. Records with overlapping activation time were collapsed such that there was no duplicate count of days. Psychiatric Data source: Pharmacare files medication Psychiatric medication was measured as the number of days of medications dispensed. Psychiatric medications were identified by the therapeutic codes and included major tranquilizers, minor tranquilizers, antidepressants, mood stabilizers and anti-convulsants. See Appendix G for list of psychiatric medication codes. v To ensure that the rates of utilization were reflective of amount of utilization while the subjects were under observation, utilization records that occurred before landing, after death or outside valid health plan registration periods were excluded. See Appendix D for percentage of records excluded. Determinants of Utilization Additional variables that appeared only in Chapter 3 as determinants of utilization were the following: 199 Appendix F Variable Description Socio- Socio-economic status of residence was indicated by the mean of economic the non-missing socio-economic quintiles of the subject's status of neighbourhood during the observation period. Socio-economic residence quintile information for each year was recorded in the health plan registry. It was assessed by Statistics Canada according to the average incomes of census enumeration areas in the nearest census cycle and assigned to each subject based on his/her postal code. The range of values was 1 to 5, with 1 being the lowest and 5 being the highest. Since the socio-economic quintile of two-thirds of the subjects' residence changed during the study period, the mean of the ordinal values was calculated for each individual to yield a continuous variable that represents a crude estimate of the average socio-economic status of that individual's environment during the study period. Of the immigrants who were registered in the health plan in 1992-2001, information on socio-economic quintile was completely missing for 0.5% (826) of them; the corresponding figure for the comparison group was 0.2% (322). 200 Appendix F Variable Description Health HSDA was assigned by the researcher using the most recent local Service health area (LHA) information available and according to the Delivery boundaries in 2003. L H A of each subject for each year was Area recorded in the health plan registry and was assigned based on (HSDA) postal code. Each L H A belonged to one of 16 HSDAs in BC. The boundaries of HSD As changed slightly over the years but the boundaries of LHAs were constant. 18% of immigrants and 23% of comparison subjects changed residence and associated HSDA during 1992-2001. Since there was no optimal method to select a "representative" HSDA, the HSDA based on last known L H A was chosen. HSDAs outside of the areas populated by the Chinese immigrants were aggregated into the "Other" category, which also included the 0.4% of the immigrants (594) who had no documented health jurisdiction. There was no missing H S D A information in the comparison group. Number of The number of visits in 1992-2001 to general practitioners where non-mental diagnosis and fee item were not related to mental health problems health visits was counted for each subject. Years since Years since landing was the number of years from landing date to landing* the mid-point of the observation period for each immigrant. Place of Place of origin was the country of last permanent residence in the origin* immigration database. Value was China, Hong Kong, Taiwan or Macau. Country of last permanent residence, rather than country of birth, was chosen because factors associated with where the immigrant resided before immigration, such as health system, were deemed to have stronger influence on health behaviours than factors associated with the birthplace. The two variables had the same value for the majority of immigrants in this study. 201 Appendix F Variable Description Education* The highest educational attainment at landing was recorded in the immigration database. Values were grouped into four categories: 0-9 years, 10-12 years, post-secondary diploma but no degrees and university degrees. There were no missing values. Marital status of immigrants at landing was recorded in the irnmigration database. Values were grouped into three categories: single, married, other (divorced, widowed, separated). Six immigrants had missing marital status information. Immigrants' self-reported ability to communicate in English and French at landing was documented in the immigration database. Values were grouped into two categories: English and no English. Ability to speak French was not considered in this study because of the small number of immigrants to B C able to speak this official language and its low utility in accessing health care in B C . Four immigrants had missing official language information. *These variables applied only to Chinese irnmigrants. Marital status* English ability* 202 Appendix G PSYCHIATRIC DIAGNOSES, FEE ITEMS AND MEDICATIONS This appendix describes all the diagnostic categories, physician claim fee items and therapeutic classes of medications used in Chapters 2 to 4 to define a utilization event as being for mental health purposes. Psychiatric Diagnoses The diagnostic category associated with an event was the first three digits of the diagnosis in the Medical Services Plan (MSP) Payments files or the first three digits (excluding the E codes) of the most responsible diagnosis in the Hospital Separations files. In this project, events where the diagnostic category belonged to one of the following were considered psychiatric. Most of the psychiatric diagnostic categories were based on the International Classification of Diseases 9 t h Revision (ICD-9); two of the categories were unique to British Columbia's MSP. The list of psychiatric diagnostic categories is as follows: Diagnostic Code Description 290 Senile and presenile organic psychotic conditions 291 Alcoholic psychoses 292 Drug psychoses 293 _. Transient organic psychotic conditions 294 Other organic psychotic conditions (chronic) 295 Schizophrenic psychoses 296 Affective psychoses 297 Paranoid states 298 Other nonorganic psychoses 299 Psychoses with origin specific to childhood 300 Neurotic disorders 301 Personality disorders 302 Sexual deviations and disorders 303 Alcohol dependence syndrome 203 Appendix G Diagnostic Code Description 304 Drug dependence 305 Nondependent abuse of drugs 306 Physiological malfunction arising from mental factors 307 Special symptoms or syndromes, not elsewhere classified 308 Acute reaction to stress 309 Adjustment reaction 310 Specific nonpsychotic mental disorders due to organic brain damage 311 Depressive disorder, not elsewhere classified 312 Disturbance of conduct, not elsewhere classified 313 Disturbance of emotions specific to childhood and adolescence 314 Hyperkinetic syndrome of childhood 04A* General psychiatric examination - no care required 50B* Anxiety/depression * MSP diagnostic codes used only in MSP Payments files See also Appendix D regarding problems with misplaced decimals in the MSP Payments files. Psychiatric Fee Items Certain fee items in MSP payments were used specifically by psychiatrists or indicated that a psychiatric problem was concerned. The list of these fee items is as follows: Fee Item Code Description 00065 Mental ill-health - Investigation 00066 Mental ill-health - Documentation 00067 Mental ill-health - Voluntary committal 204 Appendix G Fee Item Code Description 00605-00609 Visit, psychiatry (emergency, hospital, office, home) 00610 Consultation, psychiatry 00613-00614 Geriatric consultation 00615 Inpatient or home visit 00622-00627 Emotionally disturbed child and family, consultation and evaluation 00630-00632 Psychotherapy, individual, outpatient 00633, Psychotherapy, family 00635-00636 00641 Electroconvulsive therapy 00645 Patient management conference 00650-00652 Psychotherapy, individual, institution 00663-00670 Group psychotherapy 00699 Psychiatry, miscellaneous Psychiatric Medications Psychiatric medications were identified by their therapeutic class, which was indicated by the first eight digits of the therapeutic code. The following is the list of psychiatric medications. Some of the medications had multiple indications of use, including non-psychiatric conditions. Since the use of psychiatric medications was examined only in the sub-group identified as having severe and persistent mental illness, it was assumed in this study that these medications were prescribed for psychiatric reasons. Medication Type Therapeutic Class Anxiolytics 28120801,28240801-28240817 Anti-convulsants 28121205, 28129201, 28129203, 28129204, 28129207 205 Appendix G Medication Tvpe Therapeutic Class Apti-depressants 28160402-28160414, 28160416-28160423, 28169999 Anti-psychotics 28160803-28160806,28160808-28160811, 28160814-28160816, 28160818-28160824, 28160826-28160830, 28240818 < Mood stabilizers 28280001 Appendix H IDENTIFICATION OF CHINESE LANGUAGE ABDLITY OF PHYSICIANS This appendix describes the procedures involved in identifying the Chinese language ability of physicians who provided services to the study population in Chapters 3 and 4. It also discusses briefly the estimated proportion of Chinese-speaking physicians who were not correctly identified. Identification of Chinese-speaking Physicians The names, place of graduation and languages spoken of any physician who was registered in the College of Physicians and Surgeons of British Columbia (CPSBC) database as a general practitioner, family practitioner or psychiatrist in any year in 1985-2001 were obtained. The following algorithm was used to determine the Chinese language ability of these physicians in British Columbia (BC). I. Self-reported ability to speak Chinese or one of its dialects -> Chinese-speaking II. Graduation from an institution in China, Hong Kong or Taiwan -> Chinese-speaking III. Of those who were not identified in Steps 1 and 2, those with Chinese surnames were extracted. Ilia. Known by researcher to speak Chinese from personal information sources -> Chinese-speaking Known by researcher to be non-Chinese-speaking or judged to be non-Chinese (based on given names or place of graduation) -> non-Chinese-speaking Hlb. Of the remaining physicians with Chinese surnames, a telephone verification procedure was carried out: 1) Those who were still registered in 2003 and who had a valid telephone number listed in the CPSBC directory were contacted by telephone in December 2004-January 2005. 2) The first person who responded to the phone call (usually the receptionist in the physician's office) was asked to verify if the 207 Appendix H physician was able to speak Chinese or understand patients who speak Chinese. i) If the answer was affirmative Chinese-speaking ii) If the answer was negative -> non-Chinese-speaking IV. Physicians who do not have Chinese surnames and who did not report Chinese ability were considered non-Chinese-speaking. The following table summarizes the number of physicians identified as Chinese-speaking in each step of the algorithm. Table H.1 Number of Chinese-speaking Physicians Identified from CPSBC Register Chinese No Chinese Not Known Total I. Self-reported Chinese ability 735 735 II. Graduation from Chinese institutions 68 68 III. Number of Chinese surnames extracted 597 Ilia. Determined by researcher 9 28 37 1Mb. Number with Chinese surname but were not identified in previous steps 560 lllb-1. Registered in 2002-03 224 lllb-2. Successfully contacted 86 64 12 162 Number not registered or not contacted 398 398 TOTAL 898 Approximately 30,000 physician records were received from the CPSBC register. Some of the records were duplicates owing to changes in the spelling of the physicians' names or other information. The exact number of directory active general practitioners 208 Appendix H and psychiatrists in 1985-2001 was not determined. Out of this grand total, 898 were identified to be Chinese-speaking. Estimate of Unidentified Chinese-speaking Physicians • Out of 1372 physicians with Chinese surnames, 707 reported Chinese language ability and 665 did not. • Of those who did not report Chinese language ability, the ability of 105 was determined from place of graduation or researcher's knowledge. • Out of the remaining 560 whose language ability was not known, 162 were successfully contacted; the Chinese language ability of 398 remained unknown. Of the 162 contacted, 86 were confirmed to be Chinese-speaking (53%). Assuming that the distribution of Chinese language ability was the same between those successfully contacted and those not contacted, the estimate of the number of Chinese-speaking physicians among the 398 physicians not contacted was 211. • The number of unidentified Chinese-speaking physicians among those without Chinese surnames was not formally estimated but was expected to be very low since the percentage of non-Chinese surnames among those who self-reported Chinese language ability was approximately 3%. • There was a potential of 1109 (898+211) general practitioners and psychiatrists with Chinese language ability in B C in 1985-2001. Of these, an estimated 211 (19%) were not identified as such. The magnitude of the bias these unidentified Chinese-speaking physicians created in the analyses where the language ability of the physicians was a factor depended on the number of such physicians who were in active practice in a year and the number of Chinese immigrants or comparison subjects they provided service to. In general, physicians whose language ability could not be confirmed were thought to be less likely in active practice (since they were no longer registered by 2003 or were not able to be contacted) and less likely to draw Chinese immigrants to their practice (since they did not self-report Chinese language ability). J 209 Appendix I COMPARING MODEL ESTIMATES AND BOOTSTRAP STANDARD ERRORS In Chapter 3, the association of each determinant to consultations with general practitioners or psychiatrists was estimated as a rate ratio using a generalized linear regression model assuming a negative binomial distribution. To assess the sensitivity of the standard errors to the negative binomial model assumption, the estimated regression coefficients in the multivariate models were bootstrapped with 1000 resamples. Eight multivariate models were compared - the determinants of rate of visits to general practitioners and psychiatrists by immigrant males and females stratified by two age groups. Since the bootstrap standard errors were similar to those from the original regression models, only the results from the original models were reported in the chapter. The results of the estimated rate ratios and standard errors from bootstrapping are displayed in Tables 1.1 to 1.8, together with the model estimates. ) 210 Appendix I Table 1.1 Negative Binomial Model Est imates of Rate Ratios of Mental Health Visi ts to General Practice and Results of Bootst rapping (Female Immigrants Under 25 at Landing) Model Estimates Bootst rap Results Rate Ratio SD 95% CI Rate Ratio SD 95% CI Age 1 1.10 1.00 (1.09,1.11) 1.10 1.00 (1.09,1.11) Years Since Land ing 1 1.07 1.01 (1.05 ; 1.15) 1.07 1.01 (1.04,1.09) Rate of Non-Mental Health V is i ts 2 1.20 1.01 (1.19,1.22) .1.20 1.01 (1.18,1.22) Mean Socio-economic Quint i le 3 1.01 1.02 (0.98 : 1.04) 1.01 1.02 (0.96,1.05) Place of Or ig in China 1.10 1.07 (0.97,1.24) 1.10 1.09 (0.92,1.31) Hong Kong 1.20 1.05 (1.09,1.30) 1.20 1.06 (1.06,1.35) Macau 0.79 1.23 (0.52,1.19) 0.78 1.23 (0.50,1.15) Taiwan Health Service Delivery Area (HSDA) Fraser North 1.03 1.05 (0.94 : 1.13) 1.03 1.06 (0.92,1.16) Fraser South 1.01 1.08 (0.87,1.18) 1.01 1.11 (0.83,1.25) North Shore 0.83 1.12 (0.67 1.03) 0.82 1.15 (0.63,1.07) Richmond 0.76 1.05 (0.69,0.83) 0.75 1.08 (0.65,0.87) South Vancouver Island 1.32 1.16 (0.99 1.76) 1.29 1.21 (0.87,1.86) Other 1.92 1.17 (1.41 2.62) 1.90 1.27 (1.17,3.01) Vancouver 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 211 Appendix I Table 1.2 Negative Binomial Model Est imates of Rate Ratios of Visi ts to Psychiatry and Results of Bootst rapping (Female Immigrants Under 25 at Landing) Model Estimates Boots t rap Results Age1 Years Since Land ing 1 1.15 Rate of Non-Mental Health V is i t s 2 1.09 Mean Socio-economic Quint i le 3 0.95 Place of Or ig in China 1.27 Hong Kong 1.64 Macau 0.07 Taiwan Health Service Delivery Area (HSDA) Rate Ratio SD 1.04 1.01 95% CI (1.01,1.07) 1.05 (1.05:1.25) 1.03 (1.02,1.16) 1.08 (0.82,1.10) Rate Ratio 1.05 1.16 1.10 0.95 SD 95% CI 1.02 (1.01,1.09) 1.05 (1.06,1.27) 1.03. (1.04,1.17) 1.09 (0.80,1.13) 1.34 1.23 3.11 (0.72,2.26) (1.10,2.45) (0.01 ,0.68) 1.30 1.47 (0.63,2.68) 1.70 1.33 (1.01 ,3:06) 0.00 11077.6 (0.00,0.29) Fraser North 0.72 1.25 (0.47,1.12) 0.72 1.32 (0.41 1.23) Fraser South 0.43 1.43 (0.22,0.88) 0.41 1.66 (0.13 1.03) North Shore 0.63 1.61 (0.25,1.61) 0.62 1.60 (0.23 1.42) Richmond 0.67 1.25 (0.44,1:04) 0.66 .1.35 (0:36 1.17) South Vancouver Island 1.88 1.98 (0.49,7.17) 1:10 4.18 (0.06 5.56) Other 1.07 2.20 (0.23,5.00) 0.73 3.52 (0.07 3.73) Vancouver 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 212 Appendix I Table 1.3 Negative Binomial Model Est imates of Rate Ratios of Mental Health Visits to General Practice and Results of Boots t rapp ing (Male Immigrants Under 25 at Landing) Model Est imates Boots t rap Results Rate Ratio SD 95% CI Rate Ratio SD 95% CI Age 1 1.08 1.00 (1.07,1.08) 1.08 1.01 (1.07,1.09) Years Since Land ing 1 1.13 1.01 (1.11 ,1.16) 1.12 1.05 (1.03,1.21) Rate of Non-Mental Health V is i ts 2 1.19 1.01 (1.17,1.21) 1.19 1.01 (1.16,1.22) Mean Socio-economic Quin t i le 3 0.92 1.02 (0.88,0.96) 0.93 1.05 (0.86,1.02) Place of Or ig in China 1.89 1.08 (1.62,2.22) 1.86 1.22 (1.28,2.67) Hong Kong 0.99 1.06 (0.89,1.11) 1.00 1.08 (0.87,1.16) Macau 1.01 1.26 (0.64,1.58) 1.03 1.25 (0.65,1.53) Taiwan Health Service Delivery Area (HSDA) Fraser North 1.10 1.06 (0.98,1.24) 1.11 1.11 (0.91,1.39) Fraser South 0.94 1.11 (0.77,1.14) 0.94 1.14 (0.73,1.23) North Shore 1.01 1.14 (0.78,1.32) 1.01 1.16 (0.76,1.35) Richmond 0.80 1.06 (0.71 ,0.90) 0.80 1.16 (0.60,1.07) South Vancouver Island 2.90 1.20 (2.04,4.13) 2.84 1.35 (1.53,4.95) Other 1.55 1.23 (1.03,2.32) 1.47 1.45 (0.70,2.96) Vancouver 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 213 Appendix I Table 1.4 Negative Binomial Model Estimates of Rate Ratios of Visits to Psychiatry and Results of Bootstrapping (Male Immigrants Under 25 at Landing) Model Estimates Bootstrap Results Rate Ratio SD 95% CI Rate Ratio SD 95% CI Age1 1.05 1.02 (1.01 1.08) 1.04 1.02 (1.01 ,1.08) Years Since Landing1 1.09 1.05 (0.98,1.20) 1.09 1.06 (0.97,1.21) Rate of Non-Mental Health Visits2 1.13 1.04 (1.04,1.23) 1.14 1.04 (1.07,1.23) Mean Socio-economic Quintile3 0.97 1.08 (0.83,1.13) 0.98 1.10 (0.81 ,1.18) Place of Origin China 1.11 1.35 (0.62,1.99) 1.19 1.43 (0.58,2.42) Hong Kong 0.97 1.25 (0.63,1.50) 1.01 1.30 (0.60,1.70) Macau 0.27 2.49 (0.04,1.59) 0.25 1.98 (0.06,0.79) Taiwan Health Service Delivery Area (HSDA) Fraser North 1.04 1.27 (0.65,1.65) 1.04 1.31 (0.61 ,1.76) Fraser South 0.49 1.44 (0.24,1.00) 0.47 1.66 (0.15,1.14) North Shore 2.17 1.60 (0.87,5.42) 2.06 1.56 (0.83,4.77) Richmond 1.00 1.25 (0.65,1.55) 1.00 1.30 (0.58,1.65) South Vancouver Island 2.54 2.08 (0.60,10.65) 2.02 2.37 (0.38,7.98) Other 1.16 2.35 (0.22,6.21) 0.30 114.89 (0.00,4.49) Vancouver 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 214 Appendix I Table 1.5 Negative Binomial Model Estimates of Rate Ratios of Mental Health Visits to General Practice and Results of Bootstrapping (Female Immigrants Aged 25 or Over at Landing) Model Estimates Bootstrap Results Rate Ratio SD 95% CI Rate Ratio SD 95% CI Age1 1.00 1.00 (1.00; 1.00) 1.00 1.00 (1.00,1.00) Years Since Landing1 1.08 1.00 (1.07:1.09) 1.08 1.01 (1.06,1.09) Rate of Non-Mental Health Visits2 1.18 1.00 (1.17,1.18) 1.18 1.00 (1.17,1.18) Mean Socio-economic Quintile3 1.02 1.01 (1.00,1.03) 1.02 1.01 (0.99,1.04) Place of Origin China 1.11 1.03 (1.05,1.18) 1.11 1.04 (1.03,1.20) Hong Kong 1.01 1.03 (0.96,1.07) 1.01 1.03 (0.95,1.08) Macau 0.80 1.11 (0.64,0.98) 0.79 1.13 (0.62,1.01) Taiwan Health Service Delivery Area (HSDA) Fraser North 1.05 1.02 (1.00,1.10) 1.05 1.03 (0.99,1.12) Fraser South 1.06 1.04 (0.98,1.15) 1.06 1.05 (0.96,1.17) North Shore 0.80 1.06 (0.71 ,0.90) 0.80 1.07 (0.70,0.91) Richmond 0.74 1.03 (0.71 ,0.78) 0.75 1.04 (0.70,0.80) South Vancouver Island 1.33 1.08 (1.15,1.53) 1.33 1.11 (1.08,1.64) Other 0.98 1.10 (0.82,1.18) 0.98 1.09 (0.82,1.16) Vancouver Education 0-9 Years 0.86 1.04 (0.80,0.92) 0.86 1.05 (0.79,0.94) 10-12 Years 1.03 1.03 (0.97,1.09) 1.03 1.04 (0.95,1.11) Non-University 0.97 1.03 (0.92,1.03) 0.97 1.04 (0.91 ,1.04) University Degree Marital Status Single 0.94 1.03 (0.89,0.99) 0.94 1.05 (0.86,1.03) Separated/Divorced/Widowed 0.92 1.04 (0.85,0.98) 0.92 1.05 (0.83,1.02) Married English Ability English 1.02 1.02 (0.98,1.07) 1.02 1.03 (0.96,1.09) No English 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 215 Appendix I Table 1.6 Negative Binomial Model Est imates of Rate Ratios of Visi ts to Psychiatry and Resul ts of Boots t rapp ing (Female Immigrants Aged 25 or Over at Landing) Model Est imates Age 1 Years Since Land ing 1 Rate of Non-Mental Health V is i t s 2 Mean Soc io-economic Quin t i le 3 Place of Or ig in China Hong Kong Macau Taiwan Health Service Del ivery Area (HSDA) Rate Ratio 0.99 1.13 1.15 0.97 2.44 2.68 2.97 SD 1.01 1.03 1.02 1.06 1.21 1.20 1.94 95% CI (0.98 ,1.00) (1.06 1.21) (1.12,1.19) (0.88,1.08) (1.67,3.54) (1.87,3.82) (0.81 ,10.87) Rate Ratio 0:99 1.15 1.16 0.97 2.47 2.70 1.93 Boots t rap Results SD 1.01 1.03 1.02 1.07 s 1.29 1.26 3.35 No English 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 95% CI (0.98,1.01) (1.08,1.22) (1.12,1.20) (0.86,1.11) (1.50,4.17) (1.76,4.31) (0.19,10.69) Fraser North 0.64 1.16 (0.48 0.86) 0.63 1.22 (0.43,0.92) Fraser South 0.73 1.32 (0.43 1.26) 0.71 1.54 (0.28,1.62) North Shore 1.33 1.45 (0.64 2.77) 1.17 1.82 (0.34,3.35) Richmond 0.83 1.16 (0.62 1.12) 0.81 1.21 (0.55,1.17) South Vancouver Island 1.08 1.57 (0.44 2.61) 0.93 1.95 (0.20,2.59) Other 0.14 2.10 (0.03 0.59) 0.03 119.20 (0.00,0.64) Vancouver Educat ion 0-9 Years 0.73 1.25 (0.47 1.13) 0.75 1.38 (0.40,1.39) 10-12 Years 0.83 1.22 (0.57 1.22) 0.89 1.29 (0.54,1.49) Non-University 0.83 1.20 (0.58 1.19) 0.86 1.29 (0.53,1.47) University Degree Marital Status Single 2.30 1.18 (1.66 3.20) 2.31 1.24 (1.51 3.48) Separated/Divorced/W idowed 0.96 1.27 (0.60 1.53) 1.02 1.26 (0.64,1.63) Married Engl ish Abi l i ty English 0.72 1.16 (0.54 0.95) 0.73 1.19 (0.52,1.05) 216 Appendix I Table 1.7 Negative Binomial Model Estimates of Rate Ratios of Mental Health Visi ts to General Practice and Results of Boots t rapping (Male Immigrants Aged 25 or Over at Landing) Model Estimates Bootst rap Results Rate Ratio SD 95% CI Rate Ratio SD 95% CI Age 1 0.99 1.00 (0.99,1.00) 0.99 1.00 (0.99,1.00) Years Since Land ing 1 1.07 . 1.01 (1.05,1.08) . 1.07 1.01 (1.05,1.09) Rate of Non-Mental Health V is i ts 2 1.23 1.00 (1.22,1.24) 1.23 1.01 (1.22,1.25) Mean Socio-economic Quint i le 3 1.01 1.01 (0.98,1,03) . 1.01 1.02 (0.97,1.04) Place of Or ig in China 1.45 1.04 (1.33,1.58) 1.45 1.06 (1.29,1.62) Hong Kong 1.18 1.04 (1.10,1.27) 1.18 1,06 (1.06,1.31) Macau 0.94 1.16 (0.70,1.25) 0.93 1.17 (0.69,1.26) Taiwan Health Service Delivery Area (HSDA) Fraser North 0.94 1.03 (0.88,1.01) 0.94 1.05 (0.86,1.04) Fraser South 1.05 1.06 (0.94,1.17) 1.04 1.09 (0.88,1.23) North Shore 1.25 1.08 (1.07,1.45) 1.22 1.23 (0.85,1.82) Richmond 0.71 1.04 (0.66,0.76) 0.71 1.05 (0.64,0.79) South Vancouver Island 1.47 1.10 (1.22,1.77) 1.47 1.11 (1.20,1.84) Other 1.29 1.13 (1.01,1.64) 1.27 1.20 (0.89,1.77) Vancouver Educat ion 0-9Years 0.89 1.05 (0.81,0.97) 0.89 1.07 (0.77,1.01) 10-12 Years 0.98 1.04 (0.90,1.06). 0.98 1.06 (0.88,1.09) Non-University 1.09 1.04 (1.02,1.17) 1.09 1.05 (0.99,1.21) University Degree -Marital Status Single 1.22 1.04 (1.14,1.32) 1.21 1.08 (1.04,1.42) Separated/DivorcecWV idowed 1.10 1.10 (0.92,1.32) 1.09 1.16 (0.81,1.46) Married Engl ish Abi l i ty English 1.10 1.03 (1.04,1.17) 1.11 1.06 (0.99,1.23) No English 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1 to 5) 217 Appendix I Table 1.8 Negative B inomia l Model Est imates of Rate Ratios of Visi ts to Psychiatry and Resul ts of Boots t rapp ing (Male Immigrants Aged 25 or Over at Landing) Model Est imates Boots t rap Results Rate Ratio SD 95% CI Rate Ratio SD 95% CI Age 1 1.00 1.01 (0.98,1.01) 1.00 1.01 (0.98,1.02) Years Since L a n d i n g 1 1.02 1.04 (0.95,1.10) 1.04 1.05 (0.94:1.15) Rate of Non-Mental Health V is i t s 2 1.26 1.03 (1.19,1.34) 1.28 1.03 (1.21 ,1.37) Mean Soc io-economic Quin t i le 3 . 1.05 1.08 (0.91:1.22) 1.05 1.10 (0.87,1.28) Place of Or ig in China 1.76 1.30 (1.05 2:94) 1.98 1.47 (0.95,4.24) Hong Kong 2.08 1.24 (1.36:3.17) 2.27 1.37 (1.29,4.16) Macau 1.26 2.29 (0.25:6.37) 0.44 125.35 (0.00,5.06) Taiwan Health Service Del ivery Area (HSDA) Fraser North 0.54 1.22 (0.36:0.80) 0.56 1.29 (0.34,0.94) Fraser South 0.96 1.40 (0.49:1.85) 0.94 1.73 (0.29,2.62) North Shore 0.30 1.65 (0.11 :0.79) 0.30 1.86 (0.08,0.94) Richmond 0.40 1.24 (0.27:0.61) 0.41 1.28 (0.25,0.66) South Vancouver Island 0.68 1.84 (0.21:2.23) 0.41 10.10 (0.03,1.95) Other 0.22 2.24 (0.04:1.07) 0.16 8.08 (0.03,0.54) Vancouver Educat ion 0-9 Years 0.97 1.32 (0.56:1.68) 0.95 1.42 (0.49,1.89) 10-12 Years 0.78 1.26 (0.50,1.24) 0.81 1.32 (0.46,1.40) Non-University 0.98 1.23 (0.66:1.46) 1.02 1.30 (0.61 ,1.74) University Degree Marital Status Single 5.39 1.26 (3.45,8.44) 5.33 1!37 (2.88,9.68) Separated/Divorced/W idowed 0.35 1.79 (0.11:1.09) 0.34 1.83 (0.09,0.97) Married Engl ish Abi l i ty English 0.92 1.21 (0.63,1.34) 0.91 1.23 (0.61 1.39) No English 1 Rate ratio is change per year 2 Rate ratio is change per non-mental health visit per year 3 Rate ratio is change per unit (range = 1. to 5) 218 Appendix J EFFECTS OF DETERMINANTS ON RATES OF MENTAL H E A L T H VISITS TO PHYSICIANS In Chapter 3, the effect of each determinant on the rate of mental health visits to general practitioners or psychiatrists of Chinese immigrants was examined in a univariate model and in a multivariate model including all the determinants for that sex and age group. The effects of some selected determinants were further analyzed in models containing different combinations of the other variables in order to explore the contribution of these determinants after controlling for the effects of the other covariates. The selection of determinants for additional analyses was based on the magnitude of the difference in effect between the univariate and the multivariate models of the particular determinant. The selection of the combination of controlling variables included in these additional models was guided by the conceptual and empirical relationships among these variables. A l l models were general linear regression models with negative binomial distribution and stratified by sex and age group (under 25 at landing and 25 or over at landing) and type of physicians (general practitioners or psychiatrists). Effects of determinants were represented by rate ratios. In the end, all determinants for mental health visits to general practitioners by male and female immigrants of both age groups were examined in additional analyses. For visits to psychiatrists, only the determinants of place of origin and health service delivery area were examined for female irrimigrants under 25 at landing and the determinants of place of origin, health service delivery area and English skill were examined for female immigrants 25 or over at landing. Additional analyses were not performed for determinants of psychiatric visits of male immigrants of both age groups because of the relatively small amount of change between the univariate and multivariate models in these two groups. Tables J . l to J.6 present the results of the additional analyses. 219 Appendix J Table J.1 Effects of Determinants on Rate of Mental Health Visi ts to General Practice (Female Immigrants under 25 at Landing) Determinant Cont ro l l ing Variables Rate Ratio 95% CI Age at None (Univariate Model) 1.13 (1.12,1.13) Entry to Years 1.12 (1.12,1.13) Study Non-MH Rate 1.10 (1.10,1.11) Years, Non-MH Rate 1.10 (1.09,1.11) SES 1.13 (1.12,1.13) Years, Non-MH Rate, SES 1.10 (1.09,1.11) Origin 1.13 (1.12,1.13) Years, Non-MH Rate, Origin 1.10 (1.09,1.11) HSDA 1.13 (1.12,1.13) ALL 1.10 (1.09,1.11) Years None (Univariate Model) 1.18 (1.16,1.21) Since Age 1.10 (1.08,1.12) Landing Non-MH Rate 1.13 (1.11,1.15) Age, Non-MH Rate 1.08 (1.06,1.10) SES 1.19 (1.16,1.21) Age, Non-MH Rate, SES 1.08 (1.06,1.09) Origin 1.18 (1.16,1.20) Age, Non-MH Rate, Origin 1.07 (1.05,1.09) HSDA 1.18 (1.16,1.20) ALL 1.07 (1.05,1.09) Rate of Non- None (Univariate Model) 1.26 (1.25,1.28) Mental Age 1.21 (1.19,1.22) Health Years 1.25 (1.23,1.27) Visits to GP Age, Years 1.20 (1.19,1.22) SES 1.26 (1.24,1.27) Age, Years, SES 1.20 (1.19,1.22) Origin 1.25 (1.23,1.27) Age, Years, Origin 1.20 (1.18,1.21) HSDA 1.26 (1.24,1.28) ALL 1.20 (1.19,1.22) Mean Socio- None (Univariate Model) 0.87 (0.84,0.90) economic Age 0.96 (0.93,0.99) Quint i le Years 0.86 (0.83,0.88) Age, Years 0.95 (0.92,0.98) Non-MH Rate 0.95 (0.92,0.98) Age, Non-MH Rate 1.01 (0.99,1.04) Age, Years, Non-MH Rate 1.01 (0.98,1.04) Origin 0.93 (0.90,0.96) Age, Origin 0.99 (0.96,1.03) HSDA 0.87 (0.85,0.90) Age, HSDA 0.96 (0.93,0.99) ALL 1.01 (0.98,1.04) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all six variables 220 Appendix J Table J.1 (Continued) Effects of Determinants on Rate of Mental Health Visits to General Practice (Female Immigrants under 25 at Landing) Determinant Controlling Variables Place of Origin Health Service Delivery Area Health Service Delivery Area (Ref = Taiwan) None (Univariate Model) Age Years Age, Years Age, SES Non-MH Rate Age, Non-MH Rate Age, Years, Non-MH Rate Age, Years, Non-MH Rate, HSDA ALL (Ref = Vancouver) None (Univariate Model) Age Years Age, Years SES Age, SES Non-MH Rate Age, Non-MH Rate Age, Years, Non-MH Rate Age, Origin Age, Years, Non-MH Rate, Origin ALL (Ref = Vancouver) None (Univariate Model) Age Years Age, Years SES Age, SES Non-MH Rate Age, Non-MH Rate Age, Years, Non-MH Rate Age, Origin Age, Years, Non-MH Rate, Origin ALL China Hong Kong Macau Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI 2.19 (1.96:2.45) 1.66 (1.52,1.80) 1.00 (0.66,1.52) 1.49 (1.33,1.67) 1.51 (1.39,1.64) 0.98 (0.65,1.48) 2.24 (2.00,2.51) 1.47 (1.34,1.60) 0.98 (0.64,1.50) 1.55 (1.38,1.73) 1.42 (1.30,1.54) 0.98 (0.64,1.48) 1.99 (1.76,2.24) 1.60 (1.46,1.74) 0.93 (0.61 ,1.42) 1.48 (1.31,1.67) 1.51 (1.39,1.64) 0.97 (0.64,1.47) 1.43 (1.28,1.60) 1.26 (1.15,1.37) 0.75 (0.50,1.12) 1.11 (0.99,1.24) 1.21 (1.11 ,1.31) 0.78 (0.52,1.17) 1.13 (1.01 ,1.27) 1.15 (1.06,1.25) 0.78 < (0.52,1.18) 1.08 (0.96,1.22) 1.19 (1.09,1.30) 0.79 (0.52,1.19) 1.10 (0.97,1.24) 1.20 (1.09,1.30) 0.79 (0.52,1.19) Fraser North Fraser South North Shore Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI 0.85 (0.77,0.93) 0.74 (0.64,0.86) 0.58 (0.46,0.73) 0.96 (0.88,1.06) 0.82 (0.71 ,0.95) 0.71 (0.57,0.88) 0.90 (0.82,0.99) 0.78 (0.67,0.91) 0.61 (0.49,0.76) 1.00 (0.91,1.10) 0.86 (0.74,1.00) 0.72 (0.58,0.90) 0.87 (0.79,0.96) 0.80 (0.69,0.93) 0.73 (0.58,0.92) 0.97 (0.88,1.06) 0.84 (0.72,0.97) 0.75 (0.60,0.94) 0.90 . (0.82,0.99) 0.82 (0.71 ,0.95) 0.71 (0.57,0.88) 1.00 (0.91,1.09) 0.92 (0.79,1.06) 0.81 (0.65,1.00) 1.03 (0.94,1.12) 0.95 (0.82,1.11) 0.82 (0.66,1.01) 1.00 (0.91 ,1.10) 0.97 (0.84,1.13) 0.77 (0.62,0.96) 1.03 (0.94,1.13) 1.01 (0.87,1.18) 0.84 (0.68,1.04) 1.03 (0.94,1.13) 1.01 (0.87,1.18) 0.83 (0.67,1.03) Other Richmond South Vancouver Island Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI 1.21 (0.88,1.66) 0.66 (0.60,0.73) 1.37 (1.01,1.86) 1.51 (1.12,2.04) 0.78 (0.71,0.86) . 1.17 (0.88,1.56) 1.29 (0.93,1.79) 0.68 (0.62,0.75) 1.42 (1.04,1.94) 1.56 (1.15,2.13) 0.79 (0.72,0.87) 1.22 (0.90,1.63) 1.42 (1.01 ,1.98) 0.66 (0.60,0.73) 1.39 (1.02,1.88) 1.66 (1.21 ,2.27) 0.78 (0.71,0.86) 1:18 (0.88,1.57) 1.45 (1.07,1.97) 0.66 (0.60,0.73) 1.46 (1.10,1.95) 1.72 (1.28,2.31) 0.76 (0.69,0.83) 1.27 (0.96,1.69) 1.77 (1.31,2.38) 0.76 (0.70,0.84) 1.31 (0.98,1.74) 1.59 (1.17,2.15) 0.78 (0.71 ,0.86) 1.20 (0.90,1.61) 1.82 (1.35,2.45) 0.75 (0.69,0.83) 1.32 (0.99,1.76) 1.92 (1.41,2.62) 0.76 (0.69,0.83) 1.32 (0.99,1.76) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all six variables 221 Appendix J Table J.2 Effects of Determinants on Rate of Mental Health Visi ts to General Pract ice (Male Immigrants under 25 at Landing) Determinant Contro l l ing Variables Rate Ratio 95% CI Age at None (Univariate Model) 1.08 (1.08,1.09) Entry to Years 1.07 (1.07,1.08) Study Non-MH Rate 1.09 (1.09,1.10) Years, Non-MH Rate 1.09 (1.08,1.09) SES 1.08 (1.07,1.09) Years, Non-MH Rate, SES 1.08 (1.07,1.09) Origin 1.08 (1.07,1.09) Years, Non-MH Rate, Origin 1.08 (1.07,1.09) HSDA 1.08 (1.07,1.09) ALL 1.08 (1.07,1.08) Years None (Univariate Model) 1.18 (1.15,1.20) Since Age 1.13 (1.10,1.15) Landing Non-MH Rate 1.16 (1.14,1.19) Age, Non-MH Rate 1.11 (1.09,1.14) SES 1.17 (1.15,1.20) Age, Non-MH Rate, SES 1.11 (1.08,1.13) Origin 1.18 (1.16,1.21) Age, Non-MH Rate, Origin 1.13 (1.11 ,1.16) HSDA 1.17 (1.14,1.20) ALL 1.13 (1.11 ,1.16) Rate of Non- None (Univariate Model) 1.15 (1.13,1.17) Mental Age 1.19 (1.16,1.21) Health Years 1.15 (1.13,1.17) Visits to GP Age, Years 1.18 (1.16,1.21) SES 1.15 (1.13,1.18) Age, Years, SES 1.18 (1.16,1.21) Origin 1.15 (1.12,1.17) Age, Years, Origin 1.18 (1.16,1.21) HSDA 1.15 (1.13,1.17) ALL 1.19 (1.17,1.21) Mean Socio- None (Univariate Model) 0.83 (0.80,0;86) economic Age 0.86 (0.83,0.89) Quint i le Years 0.84 (0.81 ,0.87) Age, Years 0.87 (0.84,0.90) Non-MH Rate 0.82 (0.79,0.85) Age, Non-MH Rate 0.85 (0.82,0.88) Age, Years, Non-MH Rate 0.86 (0.83,0.90) Origin 0.91 (0.87,0.94) Age, Origin 0.92 (0.89,0.96) HSDA 0.85 (0.81,0.88) Age, HSDA 0.86 (0.84,0.90) ALL 0.92 (0.88,0.96) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all six variables 222 Appendix J Table J.2 (Continued) Effects of Determinants on Rate of Mental Health Visits to General Practice (Male Immigrants under 25 at Landing) Determinant Controlling Variables Place of (Ref = Taiwan) Origin Health Service Delivery Area Health Service Delivery Area China Hong Kong Macau Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 2.59 (2.26,2.98) 1.53 (1.37 ,1.69) 1.88 (1.21 ,2.94) Age 2.17 (1.89,2.50) 1.35 (1.22 ,1.50) 1.55 (0.99 ,2.41) Years 2.59 (2.25,2.98) 1.21 (1.09 ,1.35) 1.64 (1.05 .2.55) Age, Years 2.23 (1.94,2.57) 1.14 (1.03 ,1.27) 1.42 (0.91 ;2.21) SES 2.22 (1.91,2.59) 1.43 (1.28 .1.59) 1.70 (1.09 ,2.66) Age, SES 1.91 (1.64,2.22) 1.28 (1.15 1.42) 1.43 (0.92 ,2.24) Non-MH Rate 2.57 (2.24,2.95) 1.41 (1.27 1.56) 1.86 (1.20 ,2.88) Age, Non-MH Rate 2.10 (1.83,2.41) 1.21 (1.09 1.34) 1.40 (0.91 ,2.16) Age, Years, Non-MH Rate 2.16 (1.88,2.49) 1.02 (0.92 1.14) 1.29 (0.84 ,1.99) HSDA 2.35 (2.04,2.72) 1.48 (1.32 1.65) 1.44 (0.91 ,2.27) Age, HSDA 2.05 (1.77,2.37) 1.33 (1.20 1.48) 1.31 (0.83 ,2.07) Age, Years, Non-MH Rate, HSDA 2.15 (1.86,2.48) 1.03 (0.92 1.16) 1.08 (0.69 , 1-69) ALL 1.89 (1.62,2.22) 0.99 (0.89 1.11) 1.01 (0.64 ,1.58) (Ref = Vancouver) Fraser North Fraser South North Shore Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 0.78 (0.69,0.87) 0.53 (0.44 0.64) 0.60 (0.46 .0.78) Age 0.84 (0.75,0.95) 0.63 (0.52 0.75) 0.68 (0.53 0.89) Years 0.89 (0.79,1.00) 0.64 (0.53 0.78) 0.67 (0.51 0.87) Age, Years 0.93 (0.82,1.05) 0.72 (0.59 0.87) 0.73 (0.56 0.95) SES 0.82 (0.73,0.93) 0.63 (0.52 0.76) 0.81 (0.62 1.06) Age, SES 0.88 (0.79,0.99) 0.71 (0.59 0.86) 0.88 (0.67 1.15) Non-MH Rate 0.81 (0.72,0.91) 0.57 (0.47 0.68) 0.63 (0.49 0.82) Age, Non-MH Rate 0.88 (0.79,0.99) 0.68 (0.56 0.82) 0-72 (0.56 0.93) Age, Years, Non-MH Rate 0.97 , (0.87,1.09) 0.78 (0.64 0.94) 0.78 (0.60 1.01) Age, Non-MH Rate, SES 0.93 (0.83,1.04) 0.78 (0.65 0.94) 0.96 (0.74 1.24) Origin 0.90 (0.80,1.02) 0.72 (0.59 0.88) 0.74 (0.57 0.96) Age, Origin . 0.94 (0.84,1.06) 0.78 (0.64 0.95) 0.80 (0.62 1.03) Age, Years, Non-MH Rate, Origin 1.17) 1.10 (0.98,1.24) 0.91 (0.75 1.11) 0.90 (0.70 ALL 1.10 (0.98,1.24) 0.94 (0.77 1.14) 1.01 (0.78 1.32) (Ref = Vancouver) Other Richmond South Vancouver Island Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio. 95% CI None (Univariate Model) 1.62 (1.10,2.40) 0.63 (0.56 0.71) 2.27 (1.57 3.29) Age 1.28 (0.86,1.89) 0.69 (0.61 0.77) 2.03 (1.41 2.91) Years 2.09 (1.42,3.09) 0.70 (0.62 0.79) 2.74 (1.88 4.00) Age, Years 1.57 (1.06,2.32) 0.74 (0.65 0.84) 2.37 (1.64 3.42) SES 1.65 (1.09,2.50) 0.64 (0.56 0.72) 2.44 (1.68 3.54) Age, SES 1.31 (0.87,1.98) 0.69 (0.61 0.78) 2.19 (1.52 3.15) Non-MH Rate 1.85 (1.27,2.70) 0.63 (0.56 0.71) 2.55 (1.78 3.65) Age, Non-MH Rate 1.44 (0.98,2.10) 0.68 (0.61 0.77) 2.25 (1.59 3.18) Age, Years, Non-MH Rate 1.75 (1.19,2.55) 0.74 (0.66 0.84) 2.61 (1.84 3.71) Age, Non-MH Rate, SES 1.46 (0.98,2.18) 0.69 (0.61 0.78) 2.44 (1.72 3.46) Origin 1.43 (0.96,2.12) 0.69 (0.61 0.78) 2.57 (1.77 3.74) Age, Origin 1.17 (0.78,1.74) 0.73 (0.65 0.83) 2.21 (1.54 3.18) Age, Years, Non-MH Rate, Origin 1.49 (1.01,2.20) 0.81 (0.72 0.91) 2.86 (2.01 4.06) ALL 1.55 (1.03,2.32) 0.80 (0.71 , 0.90) 2.90 (2.04 4.13) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all six variables 223 Appendix J Table J.3 Effects of Determinants on Rate of Mental Health Visits to General Practice (Female Immigrants Aged 25 or Over at Landing) Determinant Controlling Variables Rate Ratio 95% CI Age at None (Univariate Model) 1.00 (1.00,1.00) Entry to Years 1.00 (1.00:1.00) Study Non-MH Rate 1.00 (1.00:1.00) Years, Non-MH Rate 1.00 (0.99:1.00) Years, Non-MH Rate, SES 1.00 (0.99:1.00) Years, Non-MH Rate, Origin 1.00 (0.99:1.00) ALL 1.00 (1.00:1.00) Years None (Univariate Model) 1.10 (1.09:1.11) Since Age 1.10 (1.09:1.11) Landing Non-MH Rate 1.07 (1.06:1.08) Age, Non-MH Rate 1.07 (1 .06 : 1.08) Age, Non-MH Rate, SES 1.07 (1 .06 : 1.08) Age, Non-MH Rate, Origin 1.08 (1.07,1.09) ALL 1.08 (1.07,1.09) Rate of Non- None (Univariate Model) 1.18 (1.17,1.18) Mental Age 1.18 (1.17,1.18) Health Years 1.17 (1.17,1.18) Visits to GP Age, Years 1.17 (1.17,1.18) Age, Years, SES 1.17 (1.17,1.18) Age, Years, Origin 1.17 (1.17,1.18) ALL 1.18 (1.17,1.18) Mean Socio- None (Univariate Model) 0.97 (0.96,0.99) economic Age 0.97 (0.96,0.99) Quintile Age, Years 0.96 (0.95,0.98) Age, Non-MH Rate 1.02 (1.01 ,1.04) Age, Origin 1.00 (0.98,1.01) Age, HSDA 0.97 (0.96,0.99) ALL 1:02 (1.00,1.03) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all nine variables 224 Appendix J Table J.3 (Continued) Effects of Determinants on Rate of Mental Health Visits to General Practice (Female Immigrants Aged 25 or Over at Landing) Determinant Controlling Variables Place of (Ref = Taiwan) China Hong Kong Macau Origin Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 1.34 (1.26,1.41) 1.26 (1.20,1.32) 1.02 (0.82 1.26) Age 1.34 (1.27 1.42) 1.24 (1.18,1.30) 1.00 (0.81 ,1.24) Years 1.37 (1.29:1.45) 1.15 (1.09,1.21) 0.90 (0.73:1.12) Age, Years 1.37 (1.29:1.45) 1.14 (1.09,1.20) 0.90 (0.72,1.12) Age, SES 1.34 (1.26:1.42) 1.24 (1.18,1.30) 1.00 (0.81 ,1.25) Non-MH Rate 1.05 (0.99,1.11) 0.99 (0.94,1.04) 0.80 (0.65,0.98) Age, Non-MH Rate 1.05 (0.99:1.11) 1.00 (0.95,1.05) 0.80 (0.65,0.99) Age, Years, Non-MH Rate 1.06 (1.01 ,1.12) 0.94 (0.90,0.99) 0.74 (0.60,0.92) Age, Non-MH Rate, SES 1.09 (1.03,1.16) 1.02 (0.97,1.06) 0.82 (0.67,1.01) Age, HSDA 1.33 (1.26,1.41) 1.28 (1.22,1.35) 1.01 (0.81,1.25) Age, Eduction 1.36 (1.29,1.44) 1.22 (1.16,1.29) 1.00 (0.81 ,1.24) Age, Marital Status 1.36 (1.28,1.44) 1.26 (1.20,1.32) 1.03 (0.83,1.27) Age, English 1.34 (1.27,1.42) 1.24 (1.18,1.30) 1.00 (0.81,1.24) ALL 1.11 (1.05,1.18) 1.01 (0.96,1.07) 0.80 (0.64,0.98) Health (Ref = Vancouver) Fraser North Fraser South North Shore Service Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI Delivery None (Univariate Model) 0.99 (0.95,1.04) 0.92 (0.84,0.99) 0.77 (0.69,0.87) Area Age 1.00 (0.96,1.05) 0.93 (0.85 1.01) 0.78 (0.69 0.87) Age, Years 1.04 (0.99,1.10) 0.97 (0.90,1.06) 0.79 (0.70,0.89) Age, SES 1.01 (0.96,1.06) 0.95 (0.87,1.03) 0.82 (0.72 0.92) Age, Non-MH Rate 1.02 (0.97,1.07) 1.03 (0.95,1.11) 0.81 (0.73,0.91) Age, Origin 1.04 (0.99,1.09) 1.04 (0.95:1.13) 0.82 (0.73,0.92) Age, Education 1.00 (0.95,1.05) 0.93 (0.86,1.01) 0.78 (0.69,0.87) Age, Marital Status 0.99 (0.95,1.04) 0.92 (0.84,0.99) 0.77 (0.69,0.87) Age, English 1.00 (0.95,1.05) 0.93 (0.85,1.01) 0.78 (0.69,0.87) ALL 1.05 (1.00:1.10) 1.06 (0.98,1.15) 0.80 (0.71 ,0.90) Health (Ref = Vancouver) Other Richmond South Vancouver Island Service Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI Delivery None (Univariate Model) 0.91 (0.77,1.09) 0.77 (0.73,0.81) 1.13 (0.97,1.32) Area Age 0.92 (0.77:1.09) 0.77 (0.74,0.81) 1.14 (0.98,1.33) Age, Years 0.92 (0.77,1.10) 0.79 (0.75,0.83) 1.19 (1.02,1.39) Age, SES 0.95 (0.79,1.14) 0.78 (0.74,0.81) 1.15 (0.99,1.33) Age, Non-MH Rate 0.99 (0.84,1.17) 0.73 (0.69,0.76) 1.32 (1.14,1.52) Age, Origin 0.94 (0.79,1.12) 0.78 (0.74,0.82) 1.15 (0.99,1.34) Age, Education 0.93 (0.78,1.10) 0.77 (0.73,0.81) 1.16 (1.00,1.35) Age, Marital Status 0.91 (0.76,1.08) 0.77 (0.73,0.81) 1.14 (0.98,1.32) Age, English 0.92 (0.77,1.09) 0.77 (0.74,0.81) 1.14 (0.98,1.33) ALL 0.98 (0.82,1.18) 0.74 (0.71 ,0.78) 1.33 (1.15,1.53) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all nine variables 225 Appendix J Table J.3 (Continued) Effects of Determinants on Rate of Mental Health Visits to General Practice (Female Immigrants Aged 25 or Over at Landing) Determinant Controlling Variables Education (Ref = University degree) 0 - 9 Years 10- 12 Years Non-University Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 1.12 .(1.06,1.19) 1.18 (1.12,1.26) 1.09 (1.03,1.15) Age 1.06 (0.99,1.13) 1.16 (1.10,1.23) 1.08 (1.01,1.14) Age, Years 0.98 (0.92,1.05) 1.09 (1.03,1.16) 1.00 (0.94,1.06) Age, Non-MH Rate 0.88 (0.83,0.94) 1.03 (0.97,1.09) 1:00 (0.95,1.06) Age, Years, Non-MH Rate 0.84 (0.79,0.89) 0.98 (0.93,1.04) 0.94 (0.89,1.00) Age, Origin 1.01 (0.95,1.08) 1.16 (1.09,1.23) 1.07 (1.01 ,1.14) Age, HSDA 1.06 (1.00,1.13) 1.19 (1.12,1.26) 1.09 (1.02,1.15) Age, English 1.06 (0.99,1.14) 1.17 (1.10,1.24) 1.08 (1.02,1.14) ALL 0.86 (0.80,0.92) 1.03 (0.97,1.09) 0.97 (0.92,1.03) Marital Status (Ref = Married) Single Separated/Divorced/ Widowed Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 0.88 (0.84,0.93) 0.97 (0.91 ,1.04) Age 0.91 (0.86,0.96) 0.90 (0.84,0.97) Age, Years 0.84 (0.79,0.88) 0.93 (0.86,1.00) Age, Non-MH Rate 0.99 (0.94,1.04) 0.89 (0.83,0.96) Age, Years, Non-MH Rate 0.93 (0.89,0.98) 0.91 (0.84,0.97) Age, Origin 0.88 (0.84,0.93) 0.88 (0.82,0.95) Age, English 0.91 (0.86,0.96) 0.90 (0.84,0.97) ALL 0.94 (0.89,0.99) 0.92 (0.85,0.98) English Skill Rate Ratio 95% CI None (Univariate Model) 0.97 (0.93 1.00) Age 1.00 (0.96 1.04) Age, Non-MH Rate 1.05 (1.01 1.09) Age, Origin 0.99 (0.95 1.04) Age, Origin, HSDA 1.00 (0.96 1.04) Age, Education 1.01 (0.96 1.06) Age, Non-MH Rate, Education 1.01 (0.97 1.06) Age, Martial Status 1.01 (0.97 1.05) ALL 1.02 (0.98 1.07) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all nine variables 226 Appendix J Table J.4 Effects of Determinants on Rate of Mental Health Visits to General Practice (Male Immigrants Aged 25 or Over at Landing) Determinant Controlling Variables Rate Ratio 95% CI Age at None (Univariate Model) 1.01 (1.01 ,1.01) Entry to Years 1.01 (1.00,1.01) Study Non-MH Rate 0.99 (0.99,0.99) Years, Non-MH Rate 0.99 (0.99 0.99) Years, Non-MH Rate, SES 0.99 (0.99,0.99) Years, Non-MH Rate, Origin 0.99 (0.99,0.99) ALL 0.99 (0.99,1.00) Years None (Univariate Model) 1.13 (1.11,1.14) Since Age 1.12 (1.11 ,1.14) Landing Non-MH Rate 1.06 (1.05,1.08) Age, Non-MH Rate 1.07 (1.06,1.09) > Age, Non-MH Rate, SES 1.07 (1.06,1.09) Age, Non-MH Rate, Origin 1.08 (1.07,1.09) ALL 1.07 (1.05,1.08) Rate of Non- None (Univariate Model) 1.23 (1.22,1.24) Mental Age 1.24 (1.23,1.25) Health Years 1.22 (1.21 1.23) Visits to GP Age, Years 1.24 (1.22,1.25) Age, Years, SES 1.23 (1.22,1.25) Age, Years, Origin 1.23 (1.22,1.24) ALL 1.23 (1.22,1.24) Mean Socio- None (Univariate Model) 0.96 (0.94,0.98) economic Age 0.95 (0.93,0.97) Quintile Age, Years 0.93 (0.91,0.95) Age, Non-MH Rate 0.99 (0.97,1.01) Age, Origin 0.98 (0.96,1.01) Age, HSDA 0.95 (0.92,0.97) ALL 1.01 (0.98,1.03) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all nine variables 227 Appendix J Table J.4 (Continued) Effects of Determinants on Rate of Mental Health Visits to General Practice (Male Immigrants Aged 25 or Over at Landing) Determinant Controlling Variables Place of Origin Health Service Delivery Area Health Service Delivery Area (Ref = Taiwan) China Hong Kong Macau Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 1.56 (1.44,1.68) 1.45 (1.36,1.55) 1.05 (0.79,1.40) Age 1.60 (1.48,1.73) 1.42 (1.33,1.51) i.08 (0.79,1.41) Years 1.69 (1.56,1.83) 1.33 (1.25,1.42) 0.97 (0.73,1.30) Age, Years 1.72 (1.59,1.86) 1.31 (1.23,1.40) 0.98 (0.73,1.31) Age, SES 1.57 (1.45,1.71) 1.40 (1.31,1.50) 1.04 (0.78,1.39) Non-MH Rate 1.46 (1.36,1.58) . 1.20 (1.13,1.28) 0.92 (0.69,1.21) Age, Non-MH Rate 1.41 (1.31 1.52) 1.21 (1.13,1.29) 0.91 (0.68,1.20) Age, Years, Non-MH Rate 1.47 (1.36,1.59) 1.15 (1.07,1.22) 0.87 (0.65,1.15) Age, HSDA 1.52 (1.41,1.65) 1.43 (1.34,1.53) 1.01 (0.76,1.35) Age, Eduction 1.67 (1.54,1.80) 1.42 (1.33,1.52) 1.11 (0.83,1.48) Age, Marital Status 1.57 (1.46,1.70) 1.39 (1.30,1.48) 1.04 (0.78,1.39) Age, English 1.61 (1.49,1.74) 1.38 (1.30,1.48) 1.10 (0.82,1.46) ALL 1.45 (1.33,1.58) 1.18 (1.10,1.27) 0.94 (0.70,1.25) (Ref = Vancouver) Fraser North Fraser South North Shore Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 0.88 (0.82,0.94) 0.85 (0.77,0.95) 0.99 (0.85,1.16) Age 0.88 (0.82,0.94) 0.87 (0.78,0.98) 0.95 (0.82,1.11) Age, Years 0.91 (0.85,0.97) 0.92 (0.82,1.02) 0.95 (0.81,1.11) Age, SES 0.90 (0.84,0.96) 0.92 (0.82,1.03) 1.06 (0.90,1.24) Age, Non-MH Rate 0.89 (0.83,0.95) 0.94 (0.84,1.04) 1.20 (1.04,1.39) Age, Years, Non-MH Rate 0.90 (0.85,0.96) 0.96 (0.86,1.07) 1.19 (1.03,1.38) Age, Origin 0.93 (0.87,0.99) 1.00 (0.90,1.12) 1.02 (0.87,1.19) Age, Education 0.88 (0.82,0.94) 0.87 (0.78,0.97) 0.96 (0.82,1.12) Age, Marital Status 0.90 (0.85,0.97) 0.92 (0.82,1.02) 0.98 (0.84,1.15) Age, English 0.87 (0.81 ,0.93) 0.86 (0.77.0.96) 0.94 (0.80.1.09) ALL v 0.94 (0.88,1.01) 1.05 (0.94,1.17) 1.25 (1.07,1.45) (Ref = Vancouver) Other Richmond South Vancouver Island Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 1.06 (0.83,1.34) 0.72 (0.67,0.77) 1.37 (1.12,1.67) Age 1.08 (0.85,1.37) 0.72 (0.67,0.77) 1.39. (1.14,1.69) Age, Years 1.11 (0.87,1.42) 0.72 (0.67,0.77) 1.38 (1.13,1.69) Age, SES 1.21 (0.94,1.56) 0.73 (0.68,0.78) 1.39 (1.14,1.70) Age, Non-MH Rate 1.16 (0.93,1.46) 0.68 (0.63,0.72) 1.52 (1.27,1.83) Age, Years, Non-MH Rate 1.18 (0.93,1.48) 0.68 (0.64,0.73) 1.51 (1.26,1.82) Age, Origin 1.11 (0.88,1.41) 0.75 (0.70,0.80) 1.35 (1.11.1.65) Age, Education 1.11 (0.87,1.41) 0.71 (0.66,0.76) 1.45 (1.19,1.76) Age, Marital Status 1.09 (0.86,1.39) 0.73 - (0.68,0.78) 1.43 (1.17,1.74) Age, English 1.09 (0.86,1.38) 0.71 (0.66,0.76) 1.38 (1.13,1.68) ALL 1.29 (1.01 ,1.64) 0.71 (0.66,0.76) 1.47 (1.22,1.77) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all nine variables 228 Appendix J Table J.4 (Continued) Effects of Determinants on Rate of Mental Health Visits to General Practice (Male Immigrants Aged 25 or Over at Landing) Determinant Controlling Variables Education (Ref = University degree) 0 • 9 Years 10 -12 Years Non-University Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 1.17 (1.09 1.27) 1.17 (1.09 1.26) 1.27 (1.19,1.36) Age 1.03 (0.95 1.12) 1.12 (1.04 1.20) 1.23 (1.15,1.32) Age, Years 0.96 (0.88,1.05) 1.04 (0.96 1.12) 1.11 (1.03,1.19) Age, Non-MH Rate 0.91 (0.85,0.99) 0.96 (0.89 1.03) 1.12 (1.05,1.19) Age, Years, Non-MH Rate 0.88 (0.81 ,0.95) 0.91 (0.85 0.98) 1.05 (0.98,1.12)-Age, Origin 0.96 (0.88,1.04) 1.09 (1.01 1.18) 1.24 (1.16,1.33) Age, Years, Non-MH Rate, Origin 0.85 (0.79,0.92) 0.93 (0.86 1.00) 1.07 (1.00,1.15) Age, HSDA 1.02 (0.94,1.11) 1.14 (1.06 1.23) 1.26 (1.17,1.35) Age, English 1.15 (1.05,1.25) 1.17 (1.09 1.27) 1.26 (1.18,1.35) Age, Years, Non-MH Rate, English 0.92 (0.84,1.01) 0.93 (0.87 1.01) 1.06 (0.99,1.13) ALL 0.89 (0.81 ,0.97) 0.98 (0.90 1.06) 1.09 (1.02,1.17) Marital Status English Skill (Ref = Married) Single Separated/Divorced/ Widowed Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 1.19 (1.11,1.27) 1.33 (1.11 1.60) Age 1.46 (1.35,1.57) 1.16 (0.97 1.40) Age, Years 1.36 (1.26,1.47) 1.26 (1.04 1.51) Age, Non-MH Rate 1.30 (1.21,1.40) 1.07 (0.89 1.28) Age, Years, Non-MH Rate 1.25 (1.16,1.35) 1.11 (0.93 1.33) Age, Origin 1.43 (1.32,1.54) 1.10 (0.91 1.32) • Age, English 1.48 (1.38,1.60) 1.17 (0.97 1.40) ALL 1.22 (1.14,1.32) 1.10 (0.92 1.32) Rate Ratio 95% CI None (Univariate Model) 1.07 (1.02,1.13) Age 1.18 (1.12,1.25) Age, Years ' 1.16 (1.10,1.23) Age, SES 1.20 (1.13,1.27) Age, Non-MH Rate 1.13 (1.07,1.19) Age, Origin 1.17 (1.11 ,1.24) Age, HSDA 1.20 (1.13,1.27) Age, Education 1.20 (1.13,1.28) Age, Martial Status 1.20 (1.14,1.27) ALL 1.10 (1.04.1.17) Age = Age at entry to study; rate ratio is change per year . ' ' Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all nine variables 229 Appendix J Table J.5 Effects of Determinants on Rate of Visits to Psychiatry (Female Immigrants under 25 at Landing) Determinant Controlling Variables Place of (Ref = Taiwan) China Hong Kong Macau Origin Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 1.97 (1.21 ,3.18) 2.74 (1.91,3.95) 0.10 (0.01,0.89) Age 1.62 (1.00,2.62) 2.32 (1.61 ,3.34) 0.10 (0.01 ,0.89) Years 2.09 (1.29,3.40) 2.31 (1.60,3.33) 0.10 (0.01 ,0.92) Age, Years 1.85 (1.14,3.02) 2.14 (1.49,3.09) 0.10 (0.01,0.92) SES 1.76 (1.03,3.00) 2.64 (1.82,3.82) 0.10 (0.01 ,0.90) Non-MH Rate 1.76 (1.08,2.84) 2.39 (1.66,3.44) 0.09 (0.01 ,0.84) Age, Non-MH Rate 1.51 (0.93,2.44) 2.07 (1.44,2.99) 0.09 (0.01 ,0.85) HSDA 1.44 (0.86,2.43) 2.27 (1.54,3.35) 0.07 (0.01 ,0.63) Age, HSDA 1.29 (0.77,2.16) 1.97 (1.34,2.91) 0.07 (0.01 ,0.67) Age, Non-MH Rate, HSDA 1.20 (0.72,2.01) 1.76 (1.19,2.60) 0.07 (0.01 ,0.64) ALL 1.27 (0.72,2.26) 1.64 (1.10,2.45) 0.07 (0.01 ,0.68) Health (Ref = Vancouver) Fraser North Fraser South North Shore Service Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI Delivery None (Univariate Model) 0.56 (0.37,0.86) 0.23 (0.12,0.45) 0.36 (0.14,0.91) Area Age 0.64 (0.42,0.98) 0.27 (0.14,0.52) 0.50 (0.20,1.25) Years 0.59 (0.39,0.89) 0.25 (0.13,0.49) 0.40 (0.16,1.01) Age, Years 0.66 (0.43,1.01) 0.28 (0.15,0.54) 0.49 (0.20,1.23) SES 0.56 (0.37.0.85) 0.24 (0.12,0.45) 0.43 (0.17,1.12) Age, SES 0.64 (0.42,0.98) 0.27 (0.14,0.52) 0.56 (0.22,1.44) Non-MH Rate 0.55 (0.36,0.83) 0.27 (0.14,0.51) 0.39 (0.16,0.98) Age, Non-MH Rate 0.66 (0.43,1.01) 0.32 (0.17,0.61) 0.54 (0.21 ,1.34) Age, Years, Non-MH Rate 0.69 (0.45,1.06) 0.33 (0.17,0.63) 0.52 (0.21,1.30) Origin 0.60 (0.39,0.91) 0.36 (0.18,0.71) 0.44 (0.17,1.12) Age,'Origin 0.68 (0.44,1.04) 0.39 (0.19,0.78) 0.57 (0.23,1.42) Age, Years, Non-MH Rate, Origin 0.73 (0.48,1.13) 0.45 (0.22,0.90) 0.59 (0.23,1.47) ALL 0.72 (0.47,1.12) 0.43 (0.22,0.88) 0.63 (0.25,1.61) Health (Ref = Vancouver) Other Richmond South Vancouver Island Service Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI Delivery None (Univariate Model) 0.51 (0.13,2.06) 0.55 (0.36,0.85) 1.70 (0.44,6.62) Area Age 0.75 (0.19 2.95) 0.68 (0.44,1.03) 1.54 (0.40,5.90) Years 0.59 (0.15,2.35) 0.64 (0.42,0.98) 1.55 (0.40,5.98) Age, Years 0.74 (0.19 2.93) 0.70 (0.46,1.07) 1.57 (0.41 5.96) SES 0.72 (0.15,3.36) 0.56 (0.37,0.86) 1.75 (0.45,6.84) Age, SES 0.98 (0.21 ,4.52) 0.68 (0.45,1.04) 1.61 (0.42,6.19) Non-MH Rate 0.59 (0.15,2.36) 0.56 (0.37,0.85) 2.07 (0.54,7.95) Age, Non-MH Rate 0.86 (0.22,3.37) 0.67 (0.44,1.03) 2.05 (0.54,7.80) Age, Years, Non-MH Rate 0.82 (0.21,3.23) 0.70 (0.46,1.06) 2.04 (0.54,7.73) Origin 0.58 (0.14,2.33) 0.52 (0.34,0.81) 1.42 (0.36,5.50) Age, Origin 0.81 (0.20,3.25) 0.64 (0.41,0.99) 1.33 (0.35,5.08) Age, Years, Non-MH Rate, Origin 0.81 (0.20,3.26) 0.68 (0144,1.05) 1.81 (0.48,6.89) ALL 1.07 (0.23,5.00) 0.67 (0.44,1.04) 1.88 (0.49,7.17) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all variables 230 Appendix J Table J.6 Effects of Determinants on Rate of Visits to Psychiatry (Female Immigrants Aged 25 or Over at Landing) Determinant Controlling Variables Place of (Ref = Taiwan) Origin English Skill China Hong Kong Macau Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 3.14 (2.26 4.37) 4.15 (3.09 5.56) 2.15 (0 59,7.84) Age 3.13 (2.25 4.35) 4.28 (3.17 5.78) 2.27 (0 62 8.33) Years 3.25 (2.34 4.54) 3.52 (2.61 4.74) 1.64 (0 45,5.96) SES 3.00 (2.10 4.27) 4.05 (3.00 5.46) 2.12 (0 58,7.76) Non-MH Rate 2.75 (1.99 3.82) 3.35 (2.50 4.48) 2.66 (0 75,9.46) HSDA 2.67 (1.88 3.78) 3.57 (2.61 4.89) 1.86 (0 51,6.82) Non-MH Rate, HSDA 2.33 (1.65 3.28) 2.87 (2.10 3.93) 2.30 (0 64,8.28) Eduction 3.07 (2.17 4.32) 3.99 (2.91 5.46) 2.09 (0 57,7.69) Marital Status 3.04 (2.18 4.22) 3.96 (2.94 5.32) 2.35 (0 65,8.51) English 3.12 (2.24 4.34) 4.20 (3.12 5.64) 2.12 (0 58,7.75) Non-MH Visits, Marital Status 2.66 (1.92 3.68) 2.91 (2.17 3.90) 3.33 (0 95,11.66) ALL 2.44 (1.67 3.54) 2.68 (1.87 3.82) 2.97 (0 81 ,10.87) Rate Ratio 95% CI None (Univariate Model) 1.00 (0.79 1.26) Age 1.02 (0.80 1.29) Years 0.91 (0.72 1.15) SES 1.05 (0.83 1.32) Non-MH Rate 1.05 (0.83 1.31) Origin 0.92 (0.73 1.16) HSDA 0.99 (0.78 1.25) Education 1.17 (0.89 1.53) Martial Status 0.93 (0.74 1.18) Years, Origin 0.88 (0.70 1.12) Years, Marital Status 0.86 (0.68 1.09) Origin, Marital Status 0.82 (0.65 1.05) Years, Origin, Marital Status 0.81 (0.63 1.02) Years, Origin, Marital Status, Education 0.79 (0.60 1.03) ALL 0.72 (0.54 0.95) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all variables 231 Appendix J Table J.6 (Continued) Effects of Determinants on Rate of Visits to Psychiatry (Female Immigrants Aged 25 or Over at Landing) Health Service Delivery Area (Ref = Vancouver) Determinant Controlling Variables Health Service Delivery Area Fraser North Fraser South North Shore Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 0.53 (0.40 0.72) 0.36 (0.22 0.59) 0.73 (0.37 1.45) Age 0.53 (0.39 0.71) 0.36 (0.22 0.59) 0.73 (0.37 1.45) Years 0.52 (0.39 0.70) 0.36 (0.22 0.59) 0.84 (0.42 1.67) SES 0.54 (0.40 0.73) 0.38 (0.23 0.62) 0.86 (0.42 1.76) Non-MH Rate 0.51 (0.38 0.69) 0.42 (0.26 0.69) 0.81 (0.41 1.60) Origin 0.62 (0.46 0.83) 0.69 (0.41 1.16) 0.82 (0.41 1.63) Education 0.55 (0.41 0.74) 0.37 (0.23 0.62) 0.84 (0.42 1.70) Marital Status 0.57 (0.42 0.76) 0.37 (0.22 0.61) 0.83 (0.42 1.66) English 0.53 (0.40 0.72) 0.36 (0.22 0.59) 0.73 (0.37 1.46) SES, Education 0.55 (0.41 0.74) 0.39 (0.24 0.65) 0.97 (0.47 2.01) SES, Education, Origin 0.61 (0.45 0.83) 0.71 (0.42 1.21) 0.90 (0.44 1.86) SES, Education, Years 0.54 (0.40 0.72) 0.40 (0.24 0.66) 1.14 (0.55 2.38) SES, Education, Years, Marital Status 0.57 (0.42 0.77) 0.41 (0.24 0.68) 1.28 • (0.61 2.66) SES, Education, Marital Status, Non-MH Rate, 0.64 (0.48 0.86) 0.78 (0.45 1.33) 1.11 (0.54 2.27) Origin Non-MH Rate, Origin, Marital Status 0.64 (0.48 0.86) 0.76 (0.45 1.29) 1.10 (0.56 2.17) ALL 0.64 (0.48 0.86) 0.73 (0.43 1.26) 1.33 (0.64 2.77) (Ref = Vancouver) Other Richmond South Vancouver Island Rate Ratio 95% CI Rate Ratio 95% CI Rate Ratio 95% CI None (Univariate Model) 0.04 (0.01 0.17) 0.82 (0.61 1.10) 0.59 (0.24 1.47) Age 0.04 (0.01 0.16) 0.81 (0.60 1.10) 0.59 (0.23 1.47) Years 0.05 (0.01 0.21) 0.86 (0.64 1.16) 0.60 (0.24 1.49) SES 0.05 (0.01 0.19) 0.81 (0.60 1.10) 0.61 (0.24 1.51) Non-MH Rate 0.06 (0.01 0.22) 0.79 (0.59 1.05) 0.83 (0.34 2.03) Origin 0.08 (0.02 0.31) 0.83 (0.62 1.12) 0.70 (0.28 1.74) Education 0.05 (0.01 0.20) 0.81 (0.60 1.09) 0.67 (0.27 1.69) Marital Status 0.05 (0.01 0.19) 0.85 (0.63 1.14) 0.70 (0.28 1.74) English 0.04 (0.01 0.17) 0.82 (0.61 1.10) 0.59 (0.24 1.47) SES, Education 0.06 (0.01 0.23) 0.81 (0.60 1.09) 0.69 (0.27 1.74) SES, Education, Origin 0.09 (0.02 0.38) 0.82 (0.61 1.11) 0.72 (0.29 1.80) SES, Education, Years 0.06 (0.01 0.25) 0.86 (0.64 1.16) 0.67 (0.27 1.69) SES, Education, Years, Marital Status 0.07 (0.02 0.28) 0.92 (0.68 1:24) 0.75 (0.30 1.90) SES, Education, Marital Status, Non-MH Rate, 0.13 (0.03 0.55) 0.82 (0.61 1.10) 1.22 (0.51 2.95) Origin Non-MH Rate, Origin, Marital Status 0.11 (0.03 0.44) 0.83 (0.62 1.11) 1.22 (0.51 2.93) ALL 0.14 (0.03 0.59) 0.83 (0.62 1.12) 1.08 (0.44 2.61) Age = Age at entry to study; rate ratio is change per year Years = Years since landing; rate ratio is change per year Non-MH Rate = Rate of non-mental health visits to general practice; rate ratio is change per visit per year SES = Mean socio-economic quintile; rate ratio is change per unit (1 to 5) Origin = Place of origin HSDA = Health Service Delivery Area ALL = Multivariate model including all variables 232 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share