UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cancer and the family : distress and quality of life among Chinese-speaking patients and family caregivers Lee, Joyce Wai Kuan 2016

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

Item Metadata

Download

Media
24-ubc_2016_november_lee_joyce.pdf [ 2.32MB ]
Metadata
JSON: 24-1.0319081.json
JSON-LD: 24-1.0319081-ld.json
RDF/XML (Pretty): 24-1.0319081-rdf.xml
RDF/JSON: 24-1.0319081-rdf.json
Turtle: 24-1.0319081-turtle.txt
N-Triples: 24-1.0319081-rdf-ntriples.txt
Original Record: 24-1.0319081-source.json
Full Text
24-1.0319081-fulltext.txt
Citation
24-1.0319081.ris

Full Text

CANCER AND THE FAMILY: DISTRESS AND QUALITY OF LIFE AMONG CHINESE-SPEAKING PATIENTS AND FAMILY CAREGIVERS   by Joyce Wai Kuan Lee  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Population and Public Health)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   October 2016  © Joyce Wai Kuan Lee, 2016 ii  Abstract Background: Cancer is a family disease affecting the patient and family members. To date, few studies have explored the psychological distress of family caregivers and its relation with patient quality of life (QOL), particularly among culturally diverse cancer populations. This study seeks to understand the associations between patient distress, family caregiver distress and patient QOL in a Chinese-speaking population in British Columbia, Canada.    Methods: A cross-sectional survey of Chinese-speaking (study population) and Anglophone (comparison group) patients (N = 55) and their family caregivers (N = 40) was used to examine patient QOL and correlates. Multiple imputation of missing data for incomplete dyads resulted in data for 29 Chinese-speaking and 28 Anglophone dyads. Multiple linear regression and mediation analyses examined predictors of QOL, and its domains, and mediation effects of patient distress and family caregiver distress. Semi-structured interviews with a subset of the Chinese-speaking survey participants comprising ten patients and six family caregivers, including five patient-family caregiver dyads, were also conducted. Analysis of culturally-embedded experiences of patients and family caregivers regarding their distress and QOL was conducted, to provide further context and explanation for the findings from the quantitative analysis. Results: Patient distress emerged as the key factor in explaining patient QOL. The effects of patient age on patient emotional well-being were mediated by patient distress, such that lower distress in older patients explained better emotional functioning. A key theme that surfaced from the qualitative data analysis was the emotional regulation occurring in the family when coping with cancer. The negative impact of family caregiver distress on patient emotional well-being was acknowledged by both the patients and the family caregivers. Younger and older patient iii  interviewees alike indicated anxiety and worries about the impact of their cancer diagnosis on their families.     Conclusions: This study provides insights to Chinese-speaking patients’ and family caregivers’ illness experience and the interrelatedness of the dyads’ responses to cancer. In expanding our knowledge of cross-cultural cancer care, these findings highlight the important role of cultural background in shaping service needs and, in turn, service delivery to cancer patients and their families.   iv  Preface This dissertation is an original intellectual product of Joyce Wai Kuan Lee (J.L). The work reported in this document was conceived, conducted and written by J.L., who conceptualized the study design, collected all the data and performed the analysis of the data. The study conceptual framework was adapted from Northouse and colleagues’ stress appraisal model (Northouse et al., 2002). Questionnaires used to collect quantitative data were standardized instruments: Kessler Distress Scale (Kessler et al., 2003) and Functional Assessment of Cancer Therapy-G Scale (Cella et al., 1993). J.L developed the questions used in interviews with the Chinese-speaking patients and family caregivers, in consultation with thesis committee members. The research described in this dissertation was approved by the University of British Columbia Behavioural Research Ethics Board (UBC BREB Number: H12-00812) and Vancouver Coastal Health Research Institute (VCHRI Study Number: V12-00812).   v  Table of Contents Abstract ........................................................................................................................................... ii Preface............................................................................................................................................ iv Table of Contents .............................................................................................................................v List of Tables ...................................................................................................................................x List of Figures ................................................................................................................................ xi Acknowledgements ....................................................................................................................... xii Dedication .................................................................................................................................... xiii Chapter 1: Introduction ....................................................................................................................1 1.1 Background ..................................................................................................................... 1 1.2 Study objectives .............................................................................................................. 4 1.3 Research questions .......................................................................................................... 5 1.4 Dissertation overview ..................................................................................................... 6 Chapter 2: Literature Review ...........................................................................................................7 2.1 Cancer and the patient ..................................................................................................... 7 2.1.1 Patient distress and quality of life ............................................................................... 7 2.2 Cancer and the family ..................................................................................................... 9 2.2.1 Role of the family caregiver in social support ............................................................ 9 2.2.2 Family caregiver distress .......................................................................................... 12 2.2.3 Patient and family caregiver adaptation to cancer .................................................... 15 2.3 Positives of cancer ........................................................................................................ 17 2.3.1 The patient experience .............................................................................................. 17 2.3.2 The family caregiver experience ............................................................................... 18 vi  2.4 Cancer and a culturally diverse population ................................................................... 19 2.4.1 Health beliefs and help-seeking in cultural context .................................................. 19 2.4.2 Family caregiving in cultural context ....................................................................... 22 2.4.3 Caregiving in Chinese families ................................................................................. 23 2.5 Conceptual framework of the study .............................................................................. 26 Chapter 3: Methods ........................................................................................................................29 3.1 Study design .................................................................................................................. 29 3.2 Study participants.......................................................................................................... 31 3.3 Study procedure ............................................................................................................ 32 3.4 Quantitative methods .................................................................................................... 34 3.4.1 Objectives and hypotheses ........................................................................................ 35 3.4.2 Measurement instrument ........................................................................................... 36 3.4.2.1 Kessler Psychological Distress Scale (K6) ....................................................... 39 3.4.2.2 Functional Assessment of Cancer Therapy-General (FACT-G) ...................... 40 3.4.3 Variables formulation ............................................................................................... 41 3.4.4 Data structure and imputation ................................................................................... 43 3.4.5 Statistical analysis methods ...................................................................................... 47 3.4.5.1 Multiple linear regression ................................................................................. 47 3.4.5.2 Power analysis .................................................................................................. 50 3.5 Qualitative methods ...................................................................................................... 51 3.5.1 Objectives and phenomenon of interest .................................................................... 51 3.5.2 Analytical approach .................................................................................................. 52 3.6 Ethical considerations ................................................................................................... 56 vii  Chapter 4: Results ..........................................................................................................................58 4.1 Quantitative analysis ..........................................................................................................58 4.1.1 Sample description ........................................................................................................ 58 4.1.2 Bivariate analyses ......................................................................................................... 61 4.1.3 Multiple regression analyses ......................................................................................... 64 4.1.3.1 Direct effects of patient, illness-related, social/family and family caregiver factors on patient QOL ..................................................................................................................... 64 4.1.3.2 Indirect effects of patient, illness-related, social/family and family caregiver factors on patient QOL through patient and family caregiver distress ............................................. 71 4.1.4 Post hoc analyses .......................................................................................................... 79 4.1.5 Summary of quantitative analyses ................................................................................ 83 4.2 Qualitative analysis ............................................................................................................85 4.2.1 Description of interview sample ................................................................................... 85 4.2.2 Cultural understanding of cancer and disclosure of diagnosis ...................................... 88 4.2.2.1 Misconceptions about cancer ............................................................................ 88 4.2.2.2 Tensions in disclosure ....................................................................................... 91 4.2.2.3 Support through disclosure ............................................................................... 92 4.2.3 Distress and emotional regulation in the family ........................................................... 94 4.2.3.1 Sources of patient distress ................................................................................. 94 4.2.3.2 Sources of family caregiver distress ................................................................. 96 4.2.3.3 Emotional regulation between patient and family caregiver ............................ 98 4.2.4 Approaches to family caregiving and quality of life .................................................... 99 4.2.4.1 The “right” family support .............................................................................. 100 viii  4.2.4.2 Positives of cancer in the family ..................................................................... 101 Chapter 5: Discussion ..................................................................................................................104 5.1 Impact of patient distress on patient QOL .................................................................. 105 5.1.1 Sources of patient distress ....................................................................................... 106 5.1.2 Variation in the negative effects of patient distress across QOL domains ............. 107 5.1.3 Understanding the relation between cancer and social well-being ......................... 109 5.1.4 Connection of disclosure of cancer diagnosis and social support ........................... 111 5.2 Mediating role of distress in patient QOL .................................................................. 112 5.3 Impact of family caregiver distress on patient QOL ................................................... 114 5.3.1 Sources of family caregiver distress ....................................................................... 115 5.3.2 Distress and emotional regulation in the family ..................................................... 118 5.4 Benefits of family caregiving and the positives of cancer .......................................... 120 Chapter 6: Conclusions and Recommendations ..........................................................................125 6.1 Summary of key findings ............................................................................................ 125 6.2 Strengths and limitations of dissertation ..................................................................... 125 6.3 Implications and knowledge translation ..................................................................... 127 6.4 Future research ............................................................................................................ 129 References ....................................................................................................................................132 Appendices ...................................................................................................................................141 Appendix A:  Demographics/Background information questions ...............................................142 Appendix B: Kessler Distress Scale (K6), English version .........................................................143 Appendix C: Kessler Distress Scale (K6), Chinese version ........................................................145 Appendix D: Functional Assessment of Cancer Therapy (FACT-G), English version 4.0 .........147 ix  Appendix E: Functional Assessment of Cancer Therapy (FACT-G), Chinese version 4.0 ........149 Appendix F: Semi-structured interview script for Chinese-speaking patient ..............................151 Appendix G: Semi-structured interview script for Chinese-speaking family caregiver ..............152 Appendix H: Descriptive summary of 19 dyads with incomplete information ...........................153 Appendix I: Pearson correlations in combined sample ................................................................154 Appendix J: Pearson correlations in Chinese-speaking group .....................................................155 Appendix K: Pearson correlations in Anglophone group ............................................................156 Appendix L Bivariate regression coefficients in combined sample ............................................157 Appendix M: Bivariate regression coefficients in Chinese-speaking group ...............................158 Appendix N: Bivariate regression coefficients in Anglophone group .........................................159  x  List of Tables Table 1: Variables description ...................................................................................................... 37 Table 2: Descriptive summary of patients and family caregivers in combined sample, Chinese-speaking and Anglophone group .................................................................................................. 60 Table 3: Summary of bivariate associations with QOL total and QOL subscales for combined sample (N=57) .............................................................................................................................. 62 Table 4: Summary of bivariate associations with QOL total and QOL subscales for Chinese-speaking group (n = 29) ................................................................................................................ 62 Table 5: Summary of bivariate associations with QOL total and QOL subscales for Anglophone group (n = 28) ............................................................................................................................... 63 Table 6: Main Hypothesis 1: Summary of multiple linear regression results............................... 67 Table 7: Main Hypothesis 2: Summary of multiple linear regression results............................... 72 Table 8: Summary of bivariate associations (correlations and regression coefficients) with QOL total and QOL subscales (n = 38) ................................................................................................. 80 Table 9: Main Hypothesis 1: Summary of multiple linear regression results (n=38) ................... 82 Table 10: Characteristics of interview participants ...................................................................... 86   xi  List of Figures Figure 1: Conceptual framework of distress and quality of life among cancer patients and family caregivers ...................................................................................................................................... 27 Figure 2: Power calculation for large and medium effect sizes in multiple regression ................ 50 Figure 3: Model 3a - Indirect effects of patient age on QOL total through patient distress, combined sample .......................................................................................................................... 75 Figure 4: Model 3a - Indirect effects of patient age on QOL total through patient distress, Anglophone group ........................................................................................................................ 75 Figure 5: Model 3b - Indirect effects of patient age on social well-being through patient distress, combined sample .......................................................................................................................... 76 Figure 6: Model 3c - Indirect effects of patient age on emotional well-being through patient distress, combined sample ............................................................................................................ 78 Figure 7: Model 3c - Indirect effects of patient age on emotional well-being through patient distress, Anglophone group .......................................................................................................... 78 xii  Acknowledgements The successful completion of this project is indebted to the institutions and individuals who have rendered immense support in the course of my dissertation endeavour. I am grateful for the financial support provided by the Western Regional Training Centre for Health Services Research, University of British Columbia, Canadian Institutes of Health Research and Canadian Frailty Network. I am also most thankful for the in-kind support I have received from Trinity Western University. My gratitude extends to the administrators and healthcare professionals for their partnership in recruitment throughout the data collection phase of the project. Particularly, I want to express my heartfelt appreciation to the patients and family caregivers for stepping forward to share their illness experiences, in the midst of their cancer journey. To my supervisor, Dr. Arminee Kazanjian, and my committee members, Drs. Carolyn Gotay, Kirsten Bell and Richard Sawatzky, thank you for investing your knowledge and wisdom in me through your affirming guidance.  To my mentors, thank you for your wise counsel and continuous encouragement, aspiring me to pursue excellence in all my undertakings.  To my friends and family, thank you for upholding me with your unwavering care and unceasing prayers.    To my parents, you are missed, and you will always be remembered for giving the best of yourselves to others.   To the Author and Perfector of my faith, thank You for Your surpassing goodness and faithfulness, bringing to completion that which You have begun in me.  xiii  Dedication  To Yip-Poh Lee and Su-Ngoh Soh, who loved life, and lived it to the fullest.  1  Chapter 1: Introduction 1.1 Background Cancer is a family disease, affecting not only individual patients, but also having a substantial impact on the family (Kim & Given, 2008; Northouse, 2005). The psychological distress of cancer patients is noted to be correlated with that of their family caregivers, even as the levels of distress in patients and in family caregivers are found to be associated with individual and illness factors such as age, gender, cancer stage and illness symptoms (Hodges, Humphris, & Macfarlane, 2005; Lissandrello, Respini, & Tralongo, 2006). While past studies have pointed to the positive correlations between cancer patient health-related outcomes and patient perception of family support availability, relatively little is known about the association between the quality of life (QOL) of patients and the psychological distress of their family caregivers. To date, there is limited evidence on the psychological and social dimensions of living with cancer, especially among culturally diverse populations. In view of the fact that the impact of cancer on the family is embedded in the cultural and social context in which the patients and the family caregivers are situated, understanding the cancer experience of patients and families from non-English speaking populations is necessary for delivering culturally appropriate care.  In Canada, visible minorities (persons, other than Aboriginal peoples, who are non-Caucasian in race or non-white in colour, as per Statistics Canada report) constitute approximately 16% of the national population, and this figure is projected to double to 31% by 2031 (Statistics Canada, 2011). Within British Columbia (BC), visible minorities comprise 26.5% of the province’s total population (Statistics Canada, 2006). Data from Census 2006 2  indicated that ethnic Chinese was the largest visible minority population group, constituting over 8% of BC’s total population. The data further suggested that in Metro Vancouver, this proportion is much higher, at 14.8% of the total population; in the City of Richmond, 40% of the residents are of Chinese ethnicity. Earlier Chinese immigrants to BC in the 1850s came from China through the United States, following the discovery of gold in the Fraser Valley (Chui, Tran, & Flanders, 2005). Changes in immigration law in the late 1970s were reflected in a growing proportion of wealthy entrepreneurs from Hong Kong and Taiwan, with a surge in migrants from Hong Kong particularly apparent in the decade prior to the handover to China in 1997. More recently, immigrants have been predominantly middle-class, highly-educated Chinese from Mainland China.  For Chinese-speaking patients and their families, adjusting to cancer may be distressing, particularly when culture and language are not congruent with the mainstream model of care delivery. Research regarding the impact of cancer on culturally diverse populations is urgently needed, especially in cultural contexts where the family, rather than the individual, is the "basic structural and functional unit" (Simpson, 2005, p. 671). Further, Simpson highlights that assessment and intervention strategies in cancer care, based on Western concepts, may not be appropriate for many such patients and their families, highlighting the critical need for culturally appropriate care for diverse cancer populations in BC and Canada.  To develop culturally appropriate cancer care interventions, it is of paramount importance to understand the cultural context of behaviours involved in both seeking and providing care. In view of the current gaps in psychosocial-oncology research on Chinese families’ experience of cancer, this study sought to investigate how cancer impacts Chinese-speaking families in the province of BC. To provide the backdrop for interpreting findings from the Chinese-speaking 3  group, a comparison group of Anglophone-Canadian cancer patients and family caregivers from BC was also included.  Notably, the definition of culture is complex, with a host of constructs including values, beliefs, customs, religion, attitude, behaviour, knowledge, and language. Further, the nature of culture is fluid and the process is dynamic (Almutairi, Dahinten, & Rodney, 2015). I have selected the linguistic aspect of culture to define the population groups investigated in my dissertation, in view of the challenges in operationalizing culture in cross-cultural studies. Notwithstanding the complexities of its conceptualization, it is important to be cognizant of neither essentializing nor trivializing culture, but to hold a flexible and balanced perspective of cultures in studies involving diverse populations.     Having had experience with cancer in my family over the past three years, this topic is of particular relevance and importance to my research endeavour. During the data collection phase of my study, my mother, who was Chinese-speaking, was diagnosed with advanced stage lung cancer. She passed away three months after her diagnosis. Two years after my mother’s death, my sister was diagnosed with ovarian cancer. She was undergoing chemotherapy, at the time when I was reporting my study findings. In my role as a daughter and a sister, I encountered first-hand the distress of living with cancer in the family, where the cultural beliefs and values regarding illness and help-seeking have served as facilitators as well as barriers in coping with the disease. At the same time, my family also experienced the positives of cancer, of personal growth and closer family relationships. The impact of cancer on the family is multi-facetted; it is through ongoing research that we glean insights into such complexities, in order to provide appropriate care for patients and their family caregivers. 4  1.2 Study objectives  This study investigated the impact of cancer on Chinese-speaking patients and their family caregivers in BC, using a conceptual framework adapted from Northouse and colleagues’ Stress-Appraisal Model that examined the effects of person, social and illness-related factors on the quality of life of patients and family members. In the study, the linguistic aspect was the key descriptor of the study population, distinguishing between the Chinese-speaking and Anglophone patients and family caregivers as representing two culture groups in the study. The overall goal of the study was to examine the relations among factors thought to influence patient QOL, and particularly the role of patient distress and family caregiver distress within the target group of Chinese-speaking cancer patients.  By means of quantitative analysis, the study first sought to examine the direct effects of factors relating to patient, illness-related, social/family and family caregiver on patient QOL. The mediation effects of patient distress and family caregiver distress on patient QOL were also investigated. To provide the background from which to interpret study results for the Chinese-speaking group, a comparison group consisting of Anglophone-Canadian cancer patients and family caregivers in BC were included in the study.  To examine the cancer experience of Chinese-speaking patients and their family caregivers in relation to their psychological distress and to patient quality of life, semi-structured interviews were conducted. The structural relationships of the key elements identified in the study's conceptual framework were also investigated through qualitative analysis. Particularly, the interrelatedness between patient distress and family caregiver distress on patient QOL was explored through semi-structured interviews. During these interviews, clarification was sought regarding family dynamics following the cancer diagnosis.  5  1.3 Research questions The key research question that underpins the quantitative analysis was: How do patient and family caregiver distress levels relate to the quality of life of cancer patients in Chinese-speaking populations? Following from this research question, the distress of patient and family caregivers was hypothesized to be directly related to patient QOL. In addition, distress was hypothesized to be indirectly related to patient QOL, mediating the effects of patient, illness-related, social/family, and family caregiver factors on QOL. The variables of the factors examined include age, gender, education, duration of migration, English fluency, cancer site/stage, treatment type, family caregiver relationship to patient, and duration of caregiving.   For the qualitative analysis, the associated research questions were: How does cancer impact patient and family caregiver psychological well-being in Chinese-speaking families? How are patient distress and family caregiver distress related, and how do they affect patient QOL? The findings from the qualitative analysis further explained the relations among patient and family caregiver distress and patient quality of life, deepening our understanding of the role family caregivers played while patients coped with the disease. While adding depth to the analysis, qualitative findings also provide insights into the discrepancies in findings across studies of cancer patient QOL (Leeman, Chang, Voils, Crandell, & Sandelowski, 2011). At the same time, the study results furnish information regarding cancer care needs of culturally and linguistically diverse populations, and serve as salient feedback to decision-makers and clinicians on the development of patient and family-centred care. 6  1.4 Dissertation overview This dissertation contains six chapters. Chapter 1 introduces the research topic, providing the background, study rationale, study objectives and the research questions investigated. In Chapter 2, a review of the literature pertinent to the research topic is presented. The chapter begins with a discussion of the implications of a cancer diagnosis on the patients, expanding the discourse to the impact of cancer on the family. It continues to explore issues patients and family caregivers encounter in coping with cancer, while also highlighting aspects of the illness that patients and family caregivers deem as beneficial experiences. The chapter transitions to the cultural dimension of family caregiving, drawing attention to issues of caregiving in Chinese families. In concluding Chapter 2, a conceptual framework depicting the relations among patients and family factors and quality of life is presented.  Chapter 3 explains the methods used in conducting the research, including a description of the study design, study population, data collection and data analysis protocol, where data analysis is delineated into quantitative and qualitative analysis approaches. Chapter 4 presents the study findings, reporting first on the results from quantitative analyses, followed by results from the qualitative analysis. Chapter 5 provides an interpretation of the study findings, integrating the results from the quantitative and the qualitative analyses. The study findings are discussed in the context of the prior knowledge of the research topic, as gleaned from the literature reviewed in Chapter 2. Chapter 6 concludes the dissertation, summarizing the key study findings, identifying the strengths and limitations of the study, while providing the implications for research and practice. Knowledge translation activities are also outlined, with recommendations for future research to advance knowledge on the study topic.  7  Chapter 2: Literature Review  2.1 Cancer and the patient  2.1.1 Patient distress and quality of life   Cancer is described as a “multiple traumatic event” (Golant & Taylor-Ford, 2010, p. 38) that is experienced over a lifetime; it is often characterized by numerous unpleasant experiences, including: diagnosis, treatment, side effects, fear of recurrence, and fear of dying. Golant and Taylor-Ford highlighted that distress of cancer could impact an individual’s cognitive, behavioral, social, emotional, and spiritual functioning, thereby interfering with the person’s perceived inability to cope with the illness. In health care literature, distress caused by acute and chronic illness is identified to encompass both physical discomfort and mental anguish, even as psychological distress is deemed the individualized patient response to illness in which the healthcare professionals intervene (Ridner, 2004). Ridner further defined psychological distress as the “unique discomforting, emotional state experienced by an individual in response to a specific stressor” (Ridner, 2004, p. 539). The levels of psychological distress in patients have been found to be significantly associated with age, gender, ethnicity, social economic status, cancer stage, and illness symptoms (Ashing-Giwa & Lim, 2011; Mehnert & Koch, 2008; Thomas, NandaMohan, Nair, & Pandey, 2011). Sinding and Wiernikowski note that the ill person's construction of their cancer experience is influenced by their demographic characteristics (such as age, social economic status), where cancer may be deemed more disruptive for some patients than for others (Sinding & Wiernikowski, 2008). Kim and Given (2008) posit that a person’s quality of life (QOL) is influenced by the availability of resources to the individual – including sociodemographic, medical, psychosocial and physical health factors 8  (Kim & Given, 2008). The authors further explain that when a demand is appraised to be exceeding the person’s resources, the demand is deemed a stressor, where the process of coping with the stressor will impact the individual’s QOL.  In the absence of consensus in defining what constitutes quality of life, QOL is conceptualized as a multidimensional construct that encapsulates different aspects or domains of life, including physical, psychological, social and functional well-being (Cella & Tulsky, 1993).  Ferrell and colleagues further expand the domains of QOL to include the measures of an individual’s emotional and spiritual well-being (Ferrell, Hassey Dow, & Grant, 1995). In the physical domain, the impact of cancer is related to many factors, which include the patient’s age and comorbidities, the specific diagnosis, location of the tumour, type of treatment and intensity of treatment (Miller, 2010). Miller further clarifies that the factors impacting the psychological and social domain (often referred to as psychosocial) include the cancer patient’s premorbid emotional life, social support, and factors related to the cancer type and the side effects of treatment. For the existential and spiritual dimension, the way in which cancer patients create meaning and purpose of life, death and dying, while making sense of the challenges that they are facing, will also influence their quality of life (Northouse, Kershaw, Mood, & Schafenacker, 2005). While QOL is used to denote outcomes as experienced by the patient, the variations in the definition of QOL include subjective evaluations (such as satisfaction), perceived health status (as in performance of daily activities), or both evaluation and health status (Ferrans, 2005). It follows that the perspective of the care recipient, of which the individual preferences and values are encapsulated, is fundamental in the measurement of patient QOL (Ferrans, Zerwic, Wilbur, & Larson, 2005).  9  The impact of cancer on each of the domains of patient QOL may also be interrelated, such as when a physical impairment after cancer treatment may increase individual psychological difficulties, and in turn impede health outcomes (Meyerowitz & Oh, 2009). While some cancer patients recover well after their treatment, others may continue to have medical, psychological, social and other consequences of living with the disease (Miller, 2010). In the range of issues that patients and families encounter, the challenges may also include financial concerns, availability of persons to provide care in the longer term, resuming the role as the care provider of dependents, and living with physiological and or functional loss after treatment (Golant & Taylor-Ford, 2010).  It follows that it is paramount to recognize and address the multidimensional effects of cancer on patients, and on the family members who assume the role of upholding the ill person’s QOL through caregiving. 2.2 Cancer and the family 2.2.1 Role of the family caregiver in social support Notwithstanding the changing pattern of the family unit, the family remains the central organizing social structure in many Western and non-Western cultures, particularly with the occurrence of critical illness (Moore & Butow, 2004). As a system that comprises interdependent people who can influence each other, a serious illness, such as cancer, affects the larger family or social network, not just the ill person (Broderick, 1993). While the nuclear family is regarded as the basic social unit in North American definitions of family, members of the extended family have also been included as key individuals of the Chinese family who provide care to their loved ones diagnosed with cancer (Bell, Lee, Foran, Kwong, & Christopherson, 2010; J. Lee & Bell, 2011). In family caregiving literature, the family caregiver is identified as an unpaid family member responsible for providing physical, informational, instrumental and emotional support to 10  aid the sick relative in the disease coping process (Bevan & Pecchioni, 2008). In this study, a family caregiver is defined by the cancer patients as a “family support person who provides physical, emotional and/or social care” after the cancer diagnosis. As congruent with the patient’s interpretation, persons sharing family-like relationships with the patients may be included as family caregivers in the study, where it would be important to consider relations beyond kinship when examining the impact of cancer on the family.  Family caregiving encompasses a series of relationships, where reconfiguration and renegotiation are continuously occurring (Dobbins, 2007). Dobbins further describes that family caregiving is “woven into the rich matrix of activities, events, roles, and relationships that make up lives” (Dobbins, 2007, p. 190). As the first line of support and caregiving, the family becomes the “buffer zone” (Lewis, 1986, p. 269) for patients living with the illness, while providing the context from which the patient attempts to derive personal meaning, purpose and self-worth. Particularly, a family's capacity to provide support in relation to finance, functional, social, emotional, and decision-making may serve as a base for stability in the uncertainty of cancer in the family and, in turn, may influence patient coping and health outcomes (Kotkamp-Mothes, Slawinsky, Hindermann, & Strauss, 2005).  Research on the role of the family among cancer populations has highlighted the critical role of family caregivers in providing emotional, instrumental and informational support to cancer patients (Bloom, 2000; Given, Given, & Kozachik, 2001). Bloom describes emotional support as the perceived availability of thoughtful and caring individuals, with whom patients can share their inner thoughts and feelings. Instrumental support or tangible support is defined as specific assistance that ill persons considered helpful. In providing informational support, Bloom explains that the knowledge relevant to the patient’s situation is shared. As the unpaid family 11  support person who aids in the patient's disease-process coping, the social support role of the family caregiver is complex and often spans across all dimensions of caregiving (Honea et al., 2008).  Defined as the “positive, affirmative and empathetic social interactions” (Golant & Taylor-Ford, 2010, p. 40), social support has been identified as a critical determinant of psychological well-being, health and coping behaviour, in the course of illness among cancer patients (Kotkamp-Mothes et al., 2005). While the size of social network may be indicative of the resources available to the patient in coping with the illness, Cheng and colleagues argue that the individual’s satisfaction with and decision to use the available resources is key in influencing various aspects of the patient’s quality of life (H. L. Cheng et al., 2013). Exploring social support and its relationship with QOL in breast cancer survivors in China, the authors observe that social support is situation-specific, where social support is perceived as helpful when it is congruent with the needs of a situation. By the same token, the influence of social support on the patient’s quality of life may also vary by the types of support received.  Where patients have identified receiving most benefit from the emotional support rendered by family and friends, women who expressed their emotions about breast cancer in receptive social contexts also reported improved physical health and reduced distress (Stanton et al., 2000). On the other hand, unsupportive behaviours from family members, including avoidance and non-disclosure of emotions, were found to be associated with increased distress among patients with breast and prostate cancer (Cordova, Cunningham, Carlson, & Andrykowski, 2001; Helgeson, Novak, Lepore, & Eton, 2004). The provision of instrumental support from family and close friends was also reported to be useful to both male and female cancer patients following surgery, particularly in relation to enhancing patient emotional well-12  being (Boehmer, Luszczynska, & Schwarzer, 2007). While informational support offered by health care providers is deemed beneficial, such information offered by family and friends may be perceived as unsolicited advice, adding to the distress of the patient (Helgeson, 2003).  Although social support may be conceptualized as constituting different components, it is important to note that these components may not operate in isolation from each other, and instead, each may serve overlapping functions (Pierce, Sarason, Sarason, Joseph, & Henderson, 1996). Pierce and colleagues opine that it is conceivable that instrumental supportive behaviours that serve to strengthen task-focused coping will lead the support recipient to conclude that he or she is valued and loved, hence feeling emotionally supported by the provider of the practical help. Further, the overlapping function of social support illustrates that supportive behaviour occurs in the context of interpersonal relationships. This type of social structure strongly increases appraisal of support by the patient.  2.2.2 Family caregiver distress  Among persons where family relationships are foundational in their social network, it follows that the impact of cancer is not confined to the ill person but also members of the family (Kim & Given, 2008; Northouse, 2005). Where cancer is deemed a family disease, it may be said that the diagnosis of cancer impacts the entire family network in myriad ways (Pecchioni, Thompson, & Anderson, 2006). The demands of caregiving may impact various aspects of the family caregivers’ lives, including personal routine, family and social relationships, career, financial, physical health and emotional wellbeing (Golant & Haskins, 2008; Hodgkinson, Butow, Hunt, et al., 2007; Kurtz, Kurtz, Given, & Given, 2004; Loke, Liu, & Szeto, 2003). While cancer patients bear the psychosocial stressors of aloneness, loss of control and loss of hope, the family caregivers face parallel psychological distress through the course of caring for 13  the ill family member (Golant & Haskins, 2010). In two meta-analyses of studies conducted in North America, Europe, Australia and Israel, family members, particularly female partners, have been said to report more psychological distress than their ill family member (Hagedoorn, Sanderman, Bolks, Tuinstra, & Coyne, 2008; Hodges et al., 2005). This finding concurred with Mellon and colleagues’ study of a US population-based sample of mixed cancer patients and their family caregivers, where the fear of recurrence was reported to be greater in the family caregivers than in the patients, as the family copes with the uncertainty following cancer treatment (Mellon, Kershaw, Northouse, & Freeman-Gibb, 2007). When family caregivers do not have outlets to express their anxieties, such suppressed emotions may lead to feelings of resentment towards the cancer patient whom they are caring for (Golant & Haskins, 2008). The role of the family caregiver is significantly impacted by patient prognosis, stages of illness and goals of care (Glajchen, 2004). The burden of caregiving on the family caregivers is also influenced by other role demands (being employed, concurrently caring for other family members) and the availability of social, financial and training resources to support their caregiving role (van Ryn et al., 2011). While it may be said that family caregivers’ experience of powerful emotions at the time of diagnosis and treatment may equal or exceed those of the patient, such emotionality may heighten as the illness advances and treatment becomes palliative in intent (Edwards & Clarke, 2004; Hodgkinson, Butow, Fuchs, et al., 2007). From Stenberg and colleagues’ review of literature on the effects of caring for a cancer patient in the family, the types of problems that family caregivers have reported include: physical health, social, information needs, emotional and burdens related to caregiving responsibilities (Stenberg, Ruland, & Miaskowski, 2010). While variations in problems and responsibilities are dependent on the severity of the patient’s illness, the authors further suggest that family caregiver problems 14  may be primarily psychological in the early stage of the patient’s treatment, shifting to be more psychosocial or physical in the later stages.  Studies on family caregiving have suggested that several factors are associated with caregiver distress, including gender, age, patient-caregiver relationship and patient symptoms (Dumont et al., 2006; Northouse, Katapodi, Schafenacker, & Weiss, 2012). In a review of literature examining the quality of life of family caregivers of cancer patients across ill trajectory, Kim and Given reported that the demographics and psychosocial characteristics found to be associated with higher caregiver distress include: being younger, being a woman, being the patient’s spouse, and having lower socioeconomic status (Kim & Given, 2008). From the findings of a study conducted in the US on the psychological adjustment of cancer caregivers with multiple roles, family caregivers who have to manage their employment and care for their children, while attending to the needs of their ill family member, were reported to be particularly likely to experience psychological distress (Kim, Baker, Spillers, & Wellisch, 2006). In another review of literature about the needs of family members of older adults with cancer, Given and Sherwood reported that while younger family caregivers were found to have challenges in affective domain when seeking to balance caregiving with their multiple societal roles, the difficulties older caregivers encountered included difficulty with tasks and physical care demands in providing care (Given & Sherwood, 2006). Notwithstanding that younger caregivers have predominantly reported more negative impact of caregiving than their older counterparts, Given and Sherwood highlight that older caregivers often have comorbid conditions, in addition to living on fixed incomes and facing a shrinking social support network. Such factors may put them at greater risk of distress, all of which may lead to poorer physical and mental health for the family caregivers.  15  Notably, family caregiving literature has highlighted female caregivers and partners to be the most vulnerable group of caregivers, with higher reported levels of distress in comparison to other informal caregiver groups. For female family caregivers, the risk of competing responsibilities may be greater than their male counterparts, resulting in feelings of entrapment in informal caregiving (Gaugler et al., 2005). While it may be said that female family caregivers are more attentive to their emotions and may be more likely to report negative effects of caregiving, spousal caregivers may also experience higher distress (Given et al., 2001; Kurtz et al., 2004). As primary attachment figures and care providers, spousal caregivers are required to maintain their role longer in providing the most extensive and comprehensive care to their ill partner (Nijboer et al., 2000). Even as age and gender may influence the family caregiving process, the dynamics in the relationship between caregiver and ill family member is also a significant consideration in caregivers’ perception of their caregiving role (Kim, Baker, et al., 2006). Particularly when the family caregiver and the patient are in less mutually satisfying relationships, such tension can create a very difficult caregiving situation, further heightening the caregiver psychological distress (Schumacher et al., 2008).  2.2.3 Patient and family caregiver adaptation to cancer   As a social system, the distress in one member of the family has a significant impact on other members, affecting patients’ and the family members’ adaptation to changes brought on by cancer, and subsequently their QOL (Northouse, Mellon, Harden, & Schafenacker, 2009). The psychological distress of cancer patients was found to be significantly correlated with that of their family caregivers, who included spouses and children as primary care providers (Hodges et al., 2005; Lissandrello et al., 2006). Family caregiving by its nature is fundamentally relational and often reciprocal (Feld, Dunkle, Schroepfer, & Shen, 2006; Lingler, Sherwood, Crighton, 16  Song, & Happ, 2008; Sanders & Kittay, 2005). As Chan and colleagues note in couple dyads, where the patient-caregiver relationship is regarded to be “reciprocal, and each party affect[ing] the other emotionally and behaviourally” (Chan, Molassiotis, Yam, Chan, & Lam, 2001, p. 389), patients are particularly sensitive to the worry and anxiety of their partners when they are diagnosed with cancer. Even as positive influence of family support extends to a patient's level of adjustment to the disease and quality of life, the family caregiver’s difficulties in adapting to cancer in the family could also adversely affect role adjustment of the patient, where adjustment problems may persist over time (Northouse, Mood, Templin, Mellon, & George, 2000).  Dealing with critical illness can result in significant psychological distress in families, interfering with normal developmental processes. As evident in research, families already experiencing signals of psychological distress, such as depression or anxiety, prior to cancer diagnosis will continue to manifest symptoms of psychological vulnerability throughout the illness trajectory (Northouse et al., 2009). As the family watches the patient's health deteriorate, helping the patient cope with emotional distress is identified as one of family caregivers highest unmet needs following the cancer diagnosis (Kim, Kashy, Spillers, & Evans, 2010). The worry of family members was also reported to persist after completion of the patient’s cancer treatment (Northouse, 2005). The uncertainty about the future and fear of cancer recurrence is the most commonly cited emotional distress for patients and family members, even as family caregivers have reported feeling more fearful than the patients that the cancer would recur (Mellon et al., 2007). Where family members experience higher uncertainty and perceive less support from others, the emotional toll puts them at higher risk of maladaptation to cancer in the family (Eton, Lepore, & Helgeson, 2005). 17  In the midst of high tension, the ill person and family members may adhere to a “conspiracy of silence" (Kotkamp-Mothes et al., 2005, p. 222), seeking to repress conflict and negative affect in order to protect each other in the family. Notwithstanding that open communication facilitates role changes and reduces role-related distress in the family system, Kotkamp-Mothes and colleagues note the challenge in evaluating communicative behaviour related to the cancer of a family member. Notably, the success in adapting to a serious illness in the family hinges on the members’ ability to change roles and rules in response to the variety of supportive needs in a constructive way. With a cancer diagnosis, the avoidance of communication about the disease may also reflect autonomy of members to detach themselves from the illness and threatening processes, although the conspiracy of silence may permanently inhibit open communication. While non-disclosure of emotions may be viewed by health professionals as maladaptive, Thorne highlights that such communication strategies may serve to maintain normalcy and dignity for the group, when they are consistent with the family philosophy and are understood by all in the family (Thorne, 1985). 2.3 Positives of cancer 2.3.1 The patient experience While the psychological distress of a cancer diagnosis and the treatment that follows can be intense, cancer patients are noted to manifest resilience in coping with the illness, and even find benefit from their experience with cancer (Miller, 2010). These benefits are classified according to three domains of life, namely, enhanced social resources, enhanced personal resources, and improved coping skills (Jim & Jacobsen, 2010). Jim and Jacobsen observe that those who report greater benefits are younger, members of ethnic minority groups, having greater disease severity and longer time since diagnosis. Other factors associated with greater 18  perceived benefits among cancer survivors include coping strategies of problem solving, positive reinterpretation and being involved in enjoyable activities.  However, association between perceived benefits and gender, socioeconomic status, education and type of treatment is unclear. Notwithstanding the challenges encountered, survivors may also experience a new emotional depth that characterizes their life after a cancer diagnosis (Miller, 2010). Indeed, the increased self-awareness and spirituality of cancer patients have often been described as the positives of cancer and the growth patients have experienced through their illness. As Bower points out, the recognition of the impermanence of life can prompt one to re-evaluate one’s values and goals, bringing about a new awareness of the self and personal growth (Bower et al., 2005). While the experience of cancer may serve as a teachable moment for some patients, this may not be the experience of others who have undergone similar illness. In the diversity of illness experience among survivors, there may be dual outcomes of both positive and negative transformations in individual perspective and identity in the aftermath of cancer (Kahana, Kahana, Deimling, Sterns, & VanGunten, 2011). As survivors are ultimately the interpreters of their cancer experience, Jim and Jacobsen further highlight that it would be prudent that healthcare providers maintain an open stance, neither rejecting nor insisting benefit-finding in cancer survivorship (Jim & Jacobsen, 2010). Although it is well intended to empower cancer patients to cope with the uncertainty of a life-threatening disease, it is important that individuals who do not experience growth are not pathologised, an outcome which only serves to further isolate patients in their illness experience.  2.3.2 The family caregiver experience Despite the challenges associated with giving care, cancer may also bring about opportunities for the families. While caregiving responsibilities can compete for time and energy 19  in the daily living of a family caregiver, caregiving interaction may be deemed a relational commitment where care providers show their love and support for the patient (Mok, Chan, Chan, & Yeung, 2003; Wong & Chan, 2007). Although the demands of caregiving may lead to psychological distress and negative physical health outcomes for the family caregivers, positive emotions arising from caregiving are also reported. The diagnosis of a life-threatening illness, as in cancer, is said to be a transformative experience, prompting family members towards better appreciation and perspective on life (Mellon, 2002; Rolland, 2005). Through such experiences, family members are led to a reorganization of life priorities, providing families with the opportunity to develop more immediate and caring relationships.  In addition to the deepening of relationship with the ill family member, caregivers have identified positive feelings of satisfaction and finding meaning in their caregiving experience, which serves to buffer the negative emotional and physical consequences of caring for an ill loved one (C. Cohen, Colantonio, & Vernich, 2002; Given, Sherwood, & Given, 2009). Further, the family caregivers’ ability to accept new possibilities of emotional and spiritual growth and appreciation for new relationships also helps them to thrive when facing the challenges of cancer in the family (Kim, Schulz, & Carver, 2007). Essentially, the negative and positive effects of caregiving may exist concurrently; the positive impact may serve to buffer the negative effects of caring for an ill family member (Kurtz et al., 2004; Pinquart & Sorensen, 2004).  2.4 Cancer and a culturally diverse population  2.4.1 Health beliefs and help-seeking in cultural context While the issues related to cancer may be universal, ramifications of problems that patients and their families face, are magnified or mitigated by their cultural and social environment (Ho, Saltel, Machavoine, Rapoport-Hubschman, & Spiegel, 2004). Particularly, 20  cultural factors related to individual perception of cancer, family role, and hierarchical role expectations of health care professionals are important considerations that influence the help-seeking behaviour of patients and family caregivers (Ashing-Giwa et al., 2004). Significant cultural differences regarding health beliefs typically emerge at the occurrence of a major health crisis, where such diversity would translate into different behavioural patterns (Rolland, 2005). Rolland notes that the areas in which variation in cultural norms may be observed include: beliefs about treatment and healing; the role of the patient as the ill person; communication about the disease; individuals to be involved in the caregiving system; persons to assume the primary caregiver role and gender norms; and the rituals viewed as normative across the illness trajectory. In instances where healthcare providers disclose to patients the truth about their cancer, without prior understanding of the underlying cultural implications, there may be the potential risk of demoralising patients and compromising their confidence in the treatment (Liu, Mok, & Wong, 2005).  Feelings of shame and guilt in patients may also arise among cultures that adhere to the belief that cancer is caused by having neglected or subjected one’s body to harmful habits, although such perception of the disease and focus on lifestyle may also be observed in North American cancer population (Ho et al., 2004). In some communities, such as among the Chinese, cancer may even be perceived as retribution of having committed injustice to others, resulting in the misfortune of a serious illness like cancer (Lui, Ip, & Chui, 2009). A cultural belief that cancer is "contagious" could also lead to social isolation (J. Lee & Bell, 2011). The impact of social stigma is particularly distressing when such beliefs of cancer are upheld in one’s community, further isolating the patient and reducing the social support network (H. L. Cheng et al., 2013).  21  Even as cultural beliefs influence a person’s emotions and concerns regarding cancer, family beliefs about normality also play a significant role in adaptation to illness. In presenting his conceptual framework- the Family Systems Illness model- developed to serve as a psychosocial map for families living with cancer over time, Rolland notes that where family values allowed having a problem, such permission enabled family members to seek help outside of the family and maintain a positive identity in coping with cancer (Rolland, 2005). From a study investigating the unmet needs and service barriers of Asian-American family caregivers, it was found that caregivers refused to accept help from outside the family, for reasons of pride and not wanting outsiders to be involved in family matters (Li, 2004). As Chan and colleagues observe, coping with cancer is primarily “a private and family matter” (Chan, Hon, Chien, & Lopez, 2004, p. 234), and the trusting relationships within the "in-group" are deemed enduring to fulfill the individual’s needs. While strong family ties and larger social support networks of extended family are beneficial to the family’s adaptation to cancer, such characteristics of the family may also prevent the caregivers from seeking help from formal care providers. In a meta-analysis of 116 empirical studies, Asian-Americans were found to provide more caregiving hours, receive lower levels of formal support services, have fewer financial resources and experience higher levels of depression than the other subgroups (Pinquart & Sorensen, 2005). Coupled with the cultural perspective of cancer as a family affair, the reluctance to share familial problems with outsiders may further exacerbate the family’s perceived burden of caregiving.  The expectations regarding patient and healthcare provider roles may also vary across cultures, where diversions from the Western philosophy of care can create tensions between the care recipients and care providers (Moore & Butow, 2004). In cultures where medical professionals are regarded to be the experts, any ambiguity in the presentation of diagnosis or 22  treatment recommendations may compromise patient and family confidence in care providers (Huang, Butow, Meiser, & Goldstein, 1999). In the North American care context, practices that uphold patient autonomy and shared decision-making may be perceived as incompetence by patients and families in cultural groups, where healthcare professionals are deemed the experts. It follows that care recipients who come from cultures where physicians are esteemed may be in tension with a care system that seeks to engage patients and families in the care process. Understandably, for patients whose health beliefs, value systems and linguistic background are distinct from the larger Canadian society, such differences may put them at odds with the Canadian health system (R. Lee, Rodin, Devins, & Weiss, 2001). 2.4.2 Family caregiving in cultural context  Notwithstanding the central role of the family in the management of the disease across cancer populations, the level of patient and family caregiver autonomy in decision making may vary according to the cultural context of the care recipients (Huang et al., 1999). In a survey of Chinese cancer patients and family caregivers in Australia, Huang and colleagues found that the majority of respondents concurred that the family was to assume the advisory role regarding the amount of information a physician should disclose to the patient, and the manner in which the information was conveyed. At the same time, the participants in the study also preferred that the family be fully informed of issues relating to the patient’s illness, in order for members in the family to best support the patient. With familial support as a coping mechanism, family members rely on social ties created and maintained when providing mutual economic and emotional support (Moore & Butow, 2004). Where the family, rather than the individual, is the unit of care, familial needs for information are central to family caregiving. In contexts of care where patient permission is still required before illness-related information may be disclosed to the family, it 23  may mean relatives will be denied information about diagnosis and prognosis, as per the patient’s request. Consequently, distress in the family may be heightened, as the difference in values of family-centred cultures and the Western individual-centred care context is accentuated. When recognizing the family as the primary source of social support for the cancer patient, it may be said that patterns of caregiving behaviour and family roles in the course of cancer care are culturally-conditioned (Gotay, 2000; Kagawa-Singer & Nguyen, 2000). At the same time, cultural variation in family structure exists, where the most critical family ties are spousal, parental and siblings. In cultural groups where the parent-child bond is most dominant, disease and cancer types that affect fertility could have profound negative implications on patients and their family members (Gotay, 2004). Further, in cultural contexts where there is an emphasis on maintaining stability and harmony in the family through members’ adherence to predefined roles, it is conceivable that a serious illness that disrupts and redefines roles in the family will create tension among its members (M. D. Lee, 2007; Loke et al., 2003). This could be particularly distressing for the cancer patient and the family caregiver, as each adjust to their new role in living with the illness. 2.4.3 Caregiving in Chinese families  While there are important regional differences in what is generally termed as Chinese culture, a common element is the emphasis placed upon the family as the “fundamental unit of society” (D. Y. H. Wu & Tseng, 1985, p. 7), where caring for the family structure and relationships is of paramount importance. With the emphasis on family solidarity and loyalty in the Chinese family system, concern and affection for family members often takes the form of caring for others’ physical needs, rather than the verbal articulation of emotion (Hsu, 1985). Cheng and colleagues note that traditional expectations in Chinese cultural contexts, which place 24  the needs of family members above one’s own, may result in family caregivers being fully occupied in just meeting the patients’ needs (Y. H. Cheng, Chi, Boey, & Ko, 1996). Family members also express caring through efforts that protect each other from physical as well as mental pain. Conceivably, the common practice of “protective truthfulness” (Liu et al., 2005, p. 268) would add to the demands of caregiving, as family caregivers bear the burden of concealing the cancer prognosis from the patient. Aligned with the concept of cohesion in the Chinese family, Hsu further clarifies that “collective uniformity” (Hsu, 1985, p. 99) of ideas and behavior is often demanded of its members, where the individual attributes are viewed as representing the collective qualities of the family. In the same vein, while a family member’s achievement is deemed a family’s honour, the individual’s adversity will constitute the family’s misfortune. This is particularly true of a cancer diagnosis, which is widely stigmatized in the Chinese community as a death penalty. Ho and colleagues note that in China, conversations about death may be interpreted as inviting it or condemning oneself to die (Ho et al., 2004). Where death and dying is a taboo topic, family members may be reluctant to discuss the disease, especially in the presence of the patient. In facing a life-threatening disease, such as cancer, problems are to be anticipated and professional help beyond the family is essential. In families where help-seeking is associated with being weak, such values may compromise the family’s resilience in coping with the illness (Li, 2004; Moore & Butow, 2004). Where the ability to tolerate momentary distress is deemed a virtue, seeking help outside the family may be regarded as shameful. Additionally in the Chinese community, the feelings of shame and guilt associated with the belief of retribution (where current sufferings are deemed to be the consequence of past actions) may prohibit patients and family members from disclosing the illness to others (Lui et al., 2009). Ho and colleagues 25  observe that the constraint on disclosing emotion or thoughts related to cancer in the Chinese population is largely driven by the desire to avoid disrupting social ties among the network of family and friends (Ho et al., 2004). Understandably, such communication gaps could complicate discussions about diagnosis and informed consent (Ngo-Metzger et al., 2003). For the family caregivers, it is conceivable that the constraints in expressing emotionality, coupled with the reluctance to disclose distress in order to preserve the collective presentation of the family to the outside world, may add to their psychological distress in the course of caregiving (J. Lee & Bell, 2011).  While having an understanding of cultural differences would be beneficial to healthcare providers, an awareness of diversity that exists within the cultural context is of critical importance (Moore & Butow, 2004).  Even as providing culturally appropriate care requires having information of the beliefs and practices within particular cultures and ethnic groups, exploring the preferences and needs of the individual remains central in care provision, and cannot be displaced. As healthcare providers remain cognizant of the cultural differences between themselves and the care recipients, such awareness will help forge a therapeutic alliance with patients and family caregivers in the course of care provision (Rolland, 2005). If cultural issues are disregarded, Rolland cautions that patients and families may wall themselves off from healthcare providers and community resources, resulting in noncompliance and treatment failure.  At present, there is a dearth of literature on the experiences of cancer among populations from culturally and linguistically diverse backgrounds (Given, Sherwood, & Given, 2011; Northouse et al., 2009; Stenberg et al., 2010). Particularly, Given and colleagues comment that “there is a noticeable absence in inclusion of diverse populations in any of the caregiver studies of the survivorship phase” (Given et al., 2011, p. 2019). With over 16% of the Canadian 26  population identified as ethnic minorities, and 15% of metro Vancouver population in BC indicating Chinese as their mother tongue, the need for culturally appropriate care for Chinese cancer patients and their families is evident. It follows that more diversity in both patient and family caregiver samples is critical, in order to better understand the role of racial and ethnic characteristics in the family’s adjustment to cancer and in the provision of appropriate care. 2.5 Conceptual framework of the study In studies of family caregiving of cancer patients, the frameworks most often described relate to stress (stimulus or response) and coping (adaptation), built on the works of Lazarus and Folkman. According to Lazarus and Folkman, emotional outcome for people is generated from their appraisal (evaluation or judgement) of the event, their coping processes, and the event outcome (Lazarus & Folkman, 1984). Based on Lazarus and Folkman’s framework, Northouse and colleagues derived the stress-appraisal model that explained the effects of person, social and illness-related factors on the quality of life of the patient and family members (Northouse et al., 2002). In Northouse and colleagues’ model, the individual’s appraisal of illness directly affects the ill person’s QOL and also mediates the effects of some of the antecedents on QOL. Within these approaches of giving care to a family member diagnosed with cancer, the caregiver responses are determined by a balance between care demands and availability to meet demands, where coping skills are central to adapting to the stress (Given et al., 2009; Sherwood et al., 2007). More importantly, the psychological distress experienced is not determined solely by the demands of the situation or by the resources of the person, but by the individual’s perception of the relation between the demands and resources in the situation. Adapting Northouse and colleagues’ theoretical model, the conceptual framework of this study is presented in Figure 1. In my dissertation, psychological distress (from hereon referred as 27  distress) is the central concept of my study framework. Therefore, I have focused on the component of emotional outcome (i.e. distress) of Northouse and colleagues’ model, rather than on the appraisal process. From the framework, four groups of factors are identified: patient, illness-related, social/family, and family caregiver. These factors affect patient QOL directly, and also indirectly through patient and family caregiver distress levels. The level of distress of the patient and the family caregivers may be interrelated, reflecting the reciprocal reaction of the ill person and the family caregiver to cancer as an “emotional system” (Northouse et al., 2012, p. 237). Patient and family caregiver distress levels are purported to be directly related to patient QOL and also indirectly through patient, illness-related, social/family, family caregiver factors, as mediated by the distress variables. In my study, variables of patient factors included age, sex, education, duration of migration and English fluency; variables of illness-related factors included cancer site, cancer stage and treatment type; variables of social/family factors included duration of caregiving and patient-family caregiver relationship; variables of family caregiver factors included age, sex and education of the family caregivers.   Figure 1: Conceptual framework of distress and quality of life among cancer patients and family caregivers            Patient QOL Patient distress Family caregiver distress Patient factors age, sex, education, migration duration, English fluency Illness-related factors cancer site/stage, treatment type  Social/family factors caregiving duration, patient-family caregiver relationship Family caregiver factors age, sex, education 28  Summary The implications of a cancer diagnosis are multifaceted, impacting not only the patient but also the family caregivers. To date, few studies have explored the psychological distress of family caregivers and its relation with patient quality of life (QOL), particularly among culturally diverse cancer populations. This study seeks to understand the associations between patient distress, family caregiver distress and patient QOL in a Chinese-speaking population in British Columbia, Canada. For this study, using the conceptual framework just illustrated, the direct effects and mediation effects of patient distress and family caregiver distress on patient QOL were first investigated. The Chinese-speaking patients and family caregivers were also interviewed, to glean insights into their experience of cancer in light of their cultural values and beliefs relating to illness and family role. The methods used in conducting the study are presented in the following chapter.  29  Chapter 3: Methods 3.1 Study design The study entailed a mixed methods design consisting of analyses of quantitative data from cross-sectional surveys and qualitative data from semi-structured interviews, in order to obtain a better understanding of the research issues than could be gleaned from either approach alone. As a methodology and a method, Creswell and Plano Clark note that mixed methods research consists of core characteristics that combine methods, philosophies and research design orientations (Creswell & Plano Clark, 2011). The mixed methods approach was said to be an outgrowth of the “triangulation of methods” (Dunning, Allison, Abonyi, & Crooks, 2008, p. 147) movement, where the main goal of triangulation was to confirm findings of a study by using qualitative and quantitative methods. In a mixed methods approach, the goal extends beyond triangulation for the purpose of confirming results. Rather, it encompasses the intent of enhancing comprehension, of gaining better understanding of results, discovering new perspectives, and/or developing a new measurement tool through the use of multiple methods (Tashakkori & Teddlie, 1998). Dunning and colleagues note the benefits and limitations of mixed methods research. The use of mixed methods serves to increase confidence in the data and the findings, while also providing the opportunity to revisit existing theories and enhance understanding of the phenomenon under study (Dunning et al., 2008). At the same time, using a mixed methods approach often requires increased skills, time and cost in conducting the studies. Further, Dunning and colleagues raise the question of whether it is appropriate to examine quantitative and qualitative data for confirmation and comprehension, in light of the distinct theoretical underpinnings specific to each method of inquiry.  In my study, a mixed methods design was used to provide different perspectives on the research questions concerning the role of family caregivers, in relation to patient adjustment to cancer within the  30  target group of a Chinese-speaking population. Specifically, quantitative regression analysis of survey data was conducted to examine how patient and family caregiver distress levels relate to patient quality of life among the Chinese-speaking cancer population living in Greater Vancouver, with the use of an Anglophone-Canadian group of patients and family caregivers to provide a backdrop for interpretation of the results for the target group. To shed light on the challenges of living with a cancer diagnosis in Chinese-speaking families, qualitative thematic analysis of interview data was used to draw insights about the impact of cancer on the psychological well-being of patients and their family caregivers, and the influence family caregivers have on the quality of life of the patients.  Aligned with the pragmatist paradigm that focuses on addressing issues by combining deductive and inductive thinking, I adopted mixed methods that encourage the “use of multiple worldviews” (Creswell & Plano Clark, 2011, p. 13) to answer research questions, rather than adhering to a single paradigm associated with either quantitative or qualitative research. Based on the quantitative results, I solicit general explanations regarding the associations among patient and family caregiver variables. Through qualitative analysis, we glean a better understanding of the processes by which cancer impacts the family. Essentially, the quantitative component of the study gives an account of the structures that are representative of the cancer phenomena observed in the family; the qualitative research provides us with a sense of the patient and the family caregiver perspectives of the disease, experienced individually and collectively in the family.  To operationalise the mixed methods approach, I used a convergent parallel design where quantitative and qualitative data were collected concurrently and analyzed during the same phase of the research process research; following this, the two sets of results were merged into an overall interpretation (Creswell & Plano Clark, 2011). The purpose was to “obtain different but complementary data” (Morse, 1991, p. 122) regarding the impact of cancer on the family, in order to best understand the  31  research issue. In this mixed methods study, both quantitative and qualitative methods were given equal weighting, as similarities and differences of the impact of cancer on patients and their family caregivers emerging from each approach were highlighted. The quantitative and qualitative results were directly compared and contrasted in relation to the research issues of distress and quality of life at the interpretation phase of the study, where both sets of results were synthesized to develop a more complete understanding of the phenomena under study. Such a research strategy provides a more complete and contextual portrayal of the impact of cancer on the patients and on their families, as informed by the integration of data from Chinese-speaking cancer patients and family caregivers, and an Anglophone comparison group, in British Columbia.           3.2 Study participants  I undertook primary data collection strategies, recruiting survey participants living in Greater Vancouver. Participants were primarily recruited through site display posters, distributed flyers and announcements at multiple recruitment sites. The sites included British Columbia Cancer Agency (BCCA), Canadian Cancer Society, local hospitals, support group networks, local broadcast media, community centres and libraries. I made presentations to care providers at their clinical rounds and team meetings; individual meetings were also initiated to facilitate referrals of potential participants. I further introduced the study to patients and families at the Chinese cancer support group facilitated at BCCA, which provided me with opportunities to make direct contact with potential participants.    Recruitment for Chinese-speaking and Anglophone cancer patients and family caregivers was conducted over a period of 1.5 years, between August 2012 and January 2014. The Chinese-speaking patient population in the study included persons who met the following criteria: 1) were 19 years or older; 2) identified Cantonese or Mandarin as their first language; 3) resided in BC; 4) were diagnosed with Stage I to III cancer; 5) were between 6 months to 2 years post diagnosis; 6) had identified at least  32  one family caregiver. Family caregivers in this study were identified by the patients as the family support persons who were primarily involved in the patients’ day-to-day care, in order to meet the physical, emotional and/or social needs of the patients. These included adults, 19 years or older, residing in BC, who shared family-like relationships with the patients. The comparison group consisted of Anglophone patients and family caregivers, who were non-Chinese native English-speakers to represent the broader Canadian society. With the exception of the language fluency inclusion criteria, the same eligibility conditions applied to the comparison group. For both groups, patients with metastatic cancer were excluded from the study, given that individuals with advanced cancer would likely be facing issues distinct from those on which the study is centred. Similarly, cancer patients diagnosed less than 6 months ago were excluded from the study population, in consideration of patient and family readiness to participate in studies during the crisis phase of a cancer diagnosis. Individuals having been diagnosed for more than two years were also excluded, for reasons of meaningful comparison of survivorship issues that patients would encounter, and the caregiving intensity family caregivers would likely face.  A subset of the Chinese-speaking survey participants also participated in semi-structured interviews. Interaction patterns within the family varied according to sex, age, cancer site and patient-family caregiver relationship. Accordingly, to provide varied interview data, interviewees were purposely sampled to include: female and male patients and family caregivers, with ages ranging from 27 to 69; patients of diverse cancer sites; family caregivers who assumed the role of spouses, children, or relatives/in-laws.  3.3 Study procedure A total of 99 patients and family caregivers provided informed consent to participate in the surveys; 95 were enrolled in the study. One Chinese-speaking and another Anglophone dyad withdrew  33  from the study for reasons of relocation and sickness. Individuals who had provided informed consent to participate in the study (55 patients and 40 family caregivers) were scheduled for quantitative questionnaires-based interviews administered in person, over the phone or by mailing, as per the participants’ preference for data collection modality. I predominantly used telephone or in-person meetings for data collection, to ensure completeness and clarity of data collected, taking into consideration time efficiency and participant accessibility. The phone surveys lasted approximately thirty minutes. In-person survey meetings were conducted with two patients, and written surveys were used for nine patients and family caregivers. To increase participation rate, I used a mixed mode of collecting survey data, where the influence of different methods of data collection in surveys has been reported to be minimal (de Leeuw, 2005).  Participation from both patients and their respective family caregivers was sought. The study sample constituted 38 dyads of patients and their family caregivers; however, participation from both individuals was not always possible, and non-dyadic data consisted of 17 patient participants and 2 family caregiver participants. Members of the dyads were interviewed separately so that the patient and his/her identified family caregivers could provide their perspectives independently of each other. During administration of the surveys, demographic data were first collected from all participants. Patient participants proceeded to report on their distress level and quality of life; family caregivers were only asked to report on their distress level.  After participants completed the questionnaires, consent was sought from Chinese-speaking survey participants identified as potential participants for the semi-structured interview, to provide additional in-depth responses on their experiences with cancer. The semi-structured interviews were scheduled within six months after the corresponding survey interviews. Sixteen Chinese-speaking survey participants, ten patients and six family caregivers, participated in a second interview. Of the ten  34  patients who participated in the interview, five of the patient’s family caregivers were also interviewed, providing five dyads for paired data analysis. The remaining five patients and one family caregiver interviewees were not related.  The questionnaires were used and the semi-structured interviews were conducted during the same data collection phase of the study. The duration of the interviews was between 1 and 2.5 hours. Written notes were taken during the semi-structured phone interviews; 11 interviews were audio recorded, where participants consented to the request for audio recording. I conducted all the interviews over telephone, as per the expressed preferences of participants. The interviews were also conducted in the language participants preferred: Cantonese, Mandarin, and English. As a native Chinese-speaker from Singapore who has received education in Canada and lived in North America for over 15 years, I am fluent in all three languages.    3.4 Quantitative methods  Multiple linear regression was used to investigate the direct effects on patient quality of life from factors within these categories: patient, illness-related, social/family, and family caregiver. Additionally, mediation analysis was conducted to examine the indirect effects of patient and family caregivers distress levels on patient QOL. Data on the independent and dependent variables were collected using self-report questionnaires. Background information about the patients and family caregivers was collected via interview prior to administering the questionnaires, and used to generate covariates. Derived variables were constructed to measure distress and QOL, and dummy variables were created to represent the categorical covariates. Multiple imputation was used to address missing data in the study sample.       35  3.4.1 Objectives and hypotheses There were two primary goals of the quantitative analysis: i) To examine the direct effects of factors relating to patient, illness-related, social/family and family caregiver on patient QOL, and ii) To investigate the mediation effects of patient and family caregiver distress levels on patient QOL. Based on the study’s conceptual framework, patient QOL is asserted to be influenced by factors related to the patient and family caregiver background (age, sex, education, English fluency, migration history), the illness (cancer site, cancer stage, treatment) and the nature of caregiving (caregiving duration and type of patient-family caregiver dyadic relationship). While these factors may directly impact patient QOL, the factors could also influence patient QOL indirectly through the patient and family caregiver appraisals of their respective distress levels, as demonstrated in the stress appraisal model.  Congruent with the notion that cancer is a family disease, the role of the family caregiver in patient adjustment to cancer may be studied in the context of the impact of the family caregiver distress on patient QOL. It follows that it is important to assess the association between patient distress and patient QOL, and that of family caregiver distress and patient QOL, when examining the factors that are predicted to have direct impact on the quality of life of the patients. While the association between patient QOL and distress may be bi-directional, patient and family caregiver distress are both assumed to precede patient QOL in the study.  The main hypotheses tested in the study are: 1. Patient factors (age, sex, education, migration duration, English fluency), illness-related factors (cancer site, cancer stage, treatment), social/family factors (caregiving duration, patient-family caregiver relationship), and family caregiver factors (age, sex, education, migration duration, English fluency) are directly related to patient QOL.  36  2. Patient factors (age, sex, education, migration duration, English fluency), illness-related factors (cancer site, cancer stage, treatment), social/family factors (caregiving duration, patient-family caregiver relationship), and family caregiver factors (age, sex, education, migration duration, English fluency) are also directly related to patient distress and family caregiver distress. 3. Patient distress and family caregiver distress are directly related to patient QOL and thus, mediate the associations between patient QOL and the factors of patient, illness-related, social/family and family caregiver. The direct effects of factors of patient, illness-related, social/family and family caregiver on patient QOL were examined through the first hypothesis. Hypotheses 2 and 3 investigated the mediation effects of patient and family caregiver distress on patient QOL. 3.4.2 Measurement instrument Data from three sets of variables were collected for this study that used a conceptual framework to examine the impact of patient and family caregiver distress on patient quality of life. The first set of data relates to the covariates, and consists of background information about the patients and family caregivers. The variables are: age, sex, education level (high school, post-secondary, university, advanced degree), migration duration (years of living in Canada), self-assessed level of English fluency (beginner, intermediate, advanced), cancer site, cancer stage, types of treatment received, caregiving duration (months of caregiving to the patient), and types of patient-family caregiver relationship. The self-reported data were obtained from the patients and family caregivers via interview prior to completion of the questionnaire (Appendix A). The data were subsequently classified according to the categories shown in the study variable description table (Table 1). 37  Table 1: Variables description Variable names / Measurement type Variables Dependent / Continuous 1) QOLTotal - PWB - SWB - EWB - FWB   Total scores of FACT-G subscales  Physical well-being subscale score  Social/Family well-being subscale score  Emotional well-being subscale  Functional well-being subscale  Independent / Continuous  1) P_Distress 2) C_Distress   Patient distress total score  Family caregiver distress total score  Covariates I) Continuous 1) P_Age 2) C_Age 3) P_Canada 4) C_Canada 5) Caregiving II) Categorical  6) Patient sex Referent P_Male 7) Family caregiver sex Referent  C_Male 8) Patient education i. Referent ii. PEd1  iii. PEd2  9) Caregiver education i. Referent ii. CEd1  iii. CEd2  10) Patient English fluency i. Referent ii. PEng 11) Family caregiver English fluency i. Referent ii. CEng  12) Patient-Caregiver relationship i. Referent ii. Relation1 iii. Relation2   Patient age  Family caregiver age Years patient lived in Canada  Years family caregiver lived in Canada  Months of caregiving    Female patient Male patient  Female family caregiver Male family caregiver  Patient with post-secondary education Patient with high school education Patient with university or advanced degree  Family caregiver with post-secondary education Family caregiver with high school education  Family caregiver with university or advanced degree  Patient with beginner or intermediate English Patient with advanced English  Patient with beginner or intermediate English Family caregiver with advanced English  Extended family/relatives or close friends Family caregiver a spouse/partner of patient Family caregiver a child of patient  38  Variable names / Measurement type Variables 13) Cancer Site i. Referent ii. Site1  iii. Site2  iv. Site3  14) Cancer Stage i. Referent ii. Stage1  iii. Stage3  15) Treatment type i. Referent ii. Treatment   Esophagus, lung, lymphoma or nasopharyngeal cancer Breast cancer   Bladder or colorectal cancer Ovarian, prostate or uterine cancer  Stage II cancer Stage I cancer Stage III cancer  Chemotherapy, drugs, radiation or surgery More than 1 treatment type   The second set of data relates to the independent variables: patient level of distress and family caregiver level of distress. The responses to the patient and family caregiver self-assessment of their respective level of distress were solicited from the participants using the Kessler Psychological Distress Scale questionnaire during the survey interviews (Appendices B and C). Details on the Kessler distress scale is found in the section below. The third set of data constitutes the dependent variable: patient quality of life. The self-reported quality of life data were collected from the patients using the Functional Assessment of Cancer Therapy-General (FACT-G) Scale questionnaire, after the patients had responded to the Kessler distress scale (Appendices D and E). The FACT-G scale is also described in the following section. Following the Kessler distress K6 development in 2002, and FACT-G in 1993, both scales have been validated and widely used in the Chinese populations in Hong Kong and China. While translation and back-translation is an approach adopted in the development of a new scale, I have not translated and back-translated the instruments, given the scales have already been tested and validated in several studies of the Chinese populations.        39  3.4.2.1 Kessler Psychological Distress Scale (K6) The short-form Kessler Psychological Distress Scale (K6) of non-specific psychological distress was used to assess patient and caregiver level of distress in the study. Consisting of six core questions that centred on non-specific psychological distress, the measure has been widely recommended as a simple and valid means of assessing mental health (Furukawa, Kessler, Slade, & Andrews, 2003). K6 asked that respondents rate how frequently they experienced symptoms of general anxiety and depression during the past thirty days on a five-point scale (from 1= all of the time to 5 = none of the time) (Kessler et al., 2003). The items were reverse coded, where a higher score indicated greater distress. Responses to the questions were summed to form a scale, ranging from 6 (no distress) to 30 (severe distress). In consultation with the K6 Scale developer, the scores were re-scaled to the range of 0 to 4 for this study, where 0 represented “none of the time” and 4 represented “all of the time.” The sum of the scores would range from 4 to 24, from no distress to severe distress. The rescaling was implemented to facilitate transition of participants from the K6 to the longer FACT-G quality of life scale during data collection, where the longer FACT-G scale adhered to the 0-4 scoring system.   The Kessler Psychological Distress Scale has been translated into 35 different languages (including Cantonese and Mandarin) and has been used in WHO World Mental Health Surveys carried out in 30 countries, including Hong Kong and Mainland China. K6 was tested to have good precision in the 90th-99th percentile range of the population distribution. The scale was also reported to have strong discrimination ability, with high probability to accurately distinguish a randomly chosen case from a randomly chosen non-case of psychological distress based on their screening scale scores (area under the curve (AUC) = 0.88), (Kessler et. al, 2002). The Chinese version of the instrument has been reported to have good reliability and validity for evaluation of mental health status in the Chinese population in China, with test-retest reliability of 0.7, split-half reliability of 0.71 and Cronbach alpha of 0.8 (Zhou,  40  Chu, & Wang, 2008). The English and traditional Chinese versions of K6 are attached, as per Appendices B and C.  3.4.2.2 Functional Assessment of Cancer Therapy-General (FACT-G) The Functional Assessment of Cancer Therapy-General (FACT-G) scale was used to measure patient quality of life (QOL). The FACT-G is a self-report questionnaire that consists of 27 items that measure the following four domains of patient quality of life/well-being, as a function of treatment: physical (7 items), social/family (7 items), emotional (6 items), functional (7 items) (Cella & Tulsky, 1993; Cella et al., 1993). Respondents were asked to rate each item on a five-point scale (from 0 =Not at all to 4 = Very much) based on a recall period of seven days. Items that were negatively stated were reverse coded, where a higher score indicated better QOL. The scores in each domain were summed to obtain subscale scores, which were subsequently summed to a total score (referred to as “QOL total” from hereon). The last item in the social/family subscale relates to the patient’s current level of satisfaction of sexual activity (“I am satisfied with my sex life”); instruction for proceeding with the survey for those participants opting not to respond to this item was indicated in the questionnaire. I am cognizant that query on sexuality from a younger female person, not from the family, may be perceived as intrusive and lacking in propriety, and had omitted the item during the survey interview. The item was treated as missing data, as per the FACT-G scoring protocol. The Functional Assessment of Cancer Therapy-General instrument has been translated into over 60 languages, including Cantonese and Mandarin. The FACT-G has been used in the Chinese cancer population to assess quality of life of patients living in Taiwan and Hong Kong (M. L. Chen, Chu, & Chen, 2004; Yu et al., 2000). The FACT-G was reported to have high coefficients of reliability and validity, and sensitivity to clinical change (Cronbach alpha = 0.89, test-retest reliability = 0.92) (Cella et al., 1993; Yu et al., 2000). The Chinese version of the FACT-G has reported good reliability, with  41  Cronbach alphas ranging from 0.85 to 0.87. The English and traditional Chinese versions of FACT-G are attached, as per Appendices D and E. 3.4.3 Variables formulation  From the conceptual framework, the dependent variable assessed in the study was patient quality of life (QOL). In addition to overall QOL (based on the FACT-G total scores), the FACT-G subscales were examined separately as dependent variables, to ascertain patient QOL with respect to each subscale of well-being, where the subscales may have varying degree of importance in relation to the independent variables. In the study, patient distress and family caregiver distress were identified as the key independent variables in relation to patient QOL. The covariates in the study were: Patient age, family caregiver age, patient sex, family caregiver sex, patient education level, family caregiver education level, patient English fluency, family caregiver English fluency, patient length of stay in Canada, family caregiver length of stay in Canada, patient-caregiver relationship, family caregiving duration, cancer site, cancer stage and number of treatment types. These covariates were used as independent variables when the associations between patient QOL and factors relating to patient, illness-related, social/family and family caregivers were explored. To examine the potential mediating effects of patient distress and family caregiver distress on patient QOL, patients and family caregivers respective distress levels were assessed as independent variables in the regression analyses, concurrently with the covariates in the study. The patient language group variable was used to compare the Chinese-speaking and Anglophone participants groups.  Variables measured on continuous scales were patient QOL total and QOL subscales, patient and family caregiver distress level, patient and family caregiver age, years patients have lived in Canada, and years family caregivers have lived in Canada and months of caregiving. The binary variables were sex (1 for male; 0 for female) of patients and family caregivers, English fluency (1 for advanced English; 0  42  for beginner or intermediate English) of patients and family caregivers and treatment type (1 for more than one treatment type; 0 for chemotherapy, drugs, radiation or surgery). Categorical variables coded as dummy variables in the regression analysis included education level, relationship, cancer site and cancer stage.  The education level of the participants was reported to span from high school to advanced degree. Two dummy variables were introduced to represent two levels of patient education: PEd1 for high school and PEd2 for university or advanced degree. PEd1 was coded 1 for patients with high school education, and 0 for patients with post-secondary, university or advanced degree; PEd2 was coded 1 for patients with university or advanced degree, and 0 for patients having high school or post-secondary education. Consequently, post-secondary education was the reference category in the regression analysis. Similarly, two dummy variables were introduced to represent the education level of family caregivers: CEd1 was coded 1 for family caregivers with high school education, and 0 for those with post-secondary, university or advanced degree; CEd2 was coded 1 for family caregivers with university or advanced degree, and 0 for those with high school or post-secondary education.  Family caregivers in the study included spouses/partners, (adult) children, parents, extended family/relatives and close friends of the patients. To represent the type of relationship between the patients and family caregivers in the study sample, two dummy variables were introduced: Relation1 for spouse or partner of the patient and Relation2 for child of the patient. Relation1 was coded 1 for family caregivers who are spouses or partners of the patients, and 0 for children, parents, extended family/relatives or close friends of patient; Relation2 was coded 1 for family caregivers who are children of the patients, and 0 for spouses/partners, parents, extended family/relatives or close friends of the patients. The latter (extended family/relatives or close friends) was accordingly regarded as the reference category in the regression analysis.  43  The cancer sites reported in the sample included: bladder, breast, colorectal, esophagus, lung, lymphoma, nasopharyngeal, ovarian, and prostate. Three dummy variables were introduced to represent the cancer sites of the patients: Site1 was coded 1 for breast cancer, and 0 for all other cancer sites; Site2 was coded 1 for bladder and colorectal cancer, and 0 for all other cancer sites; Site3 was coded 1 for ovarian, uterine and prostate cancer, and 0 for all other cancer sites. The cancer sites in the reference category were esophagus, lung, lymphoma and nasopharyngeal. To represent the 3 stages of cancer related to the respective patient’s diagnosis, two dummy variables were introduced: Stage1 and Stage3. Stage1 was coded 1 for patients with Stage I cancer, and 0 for Stage II or III cancer. Stage3 was coded 1 for patients with Stage III cancer, and 0 for Stage I or II cancer. The reference category was Stage II cancer. Description of study variables is presented in Table 1. 3.4.4 Data structure and imputation  To examine the direct and indirect effects of patient distress and family caregiver distress, regression analyses were performed on the combined data of the Chinese-speaking and Anglophone participant groups, as well as separately for each language group. The combined data of both language groups yielded 57 observations: 38 patient-family caregiver dyads, 17 patients without family caregiver data, and 2 family caregivers without patient data. The number of observations in each language group was approximately equal: 29 in the Chinese-speaking group and 28 in the Anglophone group. Given the small sample size from each language group, the data were combined to provide sufficient power to detect statistical significance. The dataset for quantitative analysis was set up using a dyadic structure, where each observation constituted patient data and the corresponding family caregiver data.          As dyadic participation was not a condition for participation in the study, it was anticipated that there would be incomplete data resulting from non-participation of either the patient or the family  44  caregiver. Several options for dealing with the missing values resulting from incomplete dyads were considered: 1. Analyze only the available data, that is, using only 38 dyads with complete data and exclude the 19 non-dyadic observations that constituted the missing cases; 2. Impute a small proportion of missing values; possibly conducting imputation for missing patient data from the two family caregivers data collected. That is, there were two cases when only caregiver data were available, and patient data could be imputed for these two cases enabling analysis of 40 dyads; 3. Revise hypotheses to examine non-dyadic responses, conducting analyses on 55 patients and 40 family caregivers as individual observations; 4. Conduct multiple imputations for missing values of the 19 individuals with incomplete dyads.  Option 1 may be comparable to casewise deletion that assumes the missing data are trivial and are missing completely at random. In real world data, the assumption of missing completely at random (MCAR) is a rare occurrence (Graham, 2012). The dyadic partners of the individuals with missing data may have decided not to participate for reasons related to the actual values missing (e.g. choosing not to participate because of high distress or low quality of life). With almost a third of the dyads with incomplete data, the missing data on 19 individuals cannot be considered a trivial omission, and substantial reduction in cases will undermine validity of the study results. When these assumptions are violated, it cannot be said that the estimates are efficient or unbiased.  The same concern may be said of option 2, when 17 missing cases are ignored. For option 3, a point to note is that the data are not independent, in cases where the patient and family caregiver participants are from the same family, or are related. It follows that analysis on the combined sample of  45  55 patients and 40 family caregivers will violate the assumption of independence of data. A further argument for adhering to analysis of dyadic responses is that there is limited existing evidence on the study topic of the interrelatedness of patient and their family caregiver distress and quality of life among culturally diverse populations. As such, it will be important to investigate the complexities of patient and family caregiver interaction by examining dyadic responses, with the goal to contribute to findings as well as lessons for future research of such nature.  While there are various ways of handling missing values, I have chosen option 4, to maximize efficient use of the data by correcting for missing data. To retain 57 dyads in the analysis, missing values to 17 caregivers and 2 patients were imputed using multiple imputation methods. Multiple imputation (MI) technique assumes that data are missing at random (MAR), which means that the missing responses depend on (or can be explained by) other variables included in the dataset. However, the missing data are assumed to be independent of any (unobserved or unknown) variables that are not part of the dataset. This is less restrictive than the assumption of missing completely at random (MCAR), as in casewise deletion, where the missing data are assumed to be independent from any other observed or unobserved variables (i.e., the absence of data, or missingness) cannot be explained by other variables included in the dataset). With a less restrictive assumption on missingness, MI allows for additional variability due to the unknown missing values by imputing more than one value for the missing item (Graham, 2012). This is obtained by replacing each missing data by a vector composed of a number of possible values predicted over the multiple imputations of data sets. By taking uncertainty in the imputation into account, MI is said to be a more conservative approach with less risk of a Type I error (with greater standard errors for the imputed values)(Rubin, 1987). With missing data on 19 observations (17 patients without caregiver’s participation; 2 caregivers without patient’s participation), some variables in the study were reported to have up to 30% of missing  46  data. MI can address the missing data problem, even with sample sizes as low as 50, and with as much as 50% missing from most variables in the dataset (Graham & Schafer, 1999). While MI cannot compensate for the limitations of a small sample, it can allow for effective use of all data that were collected. Nonetheless, caution is to be exercised when interpreting the imputed values (Gyimah, 2001). Therefore, the analyses were repeated using the complete data of 38 dyads to compare results and determine the extent to which conclusions may have differed as a result of the two different approaches to dealing with missing data. Prior to imputation of missing values, mean computation was performed on each of the FACT-G subscales (Physical, Social/Family, Emotional, Functional) and on the Kessler distress total scores scales for patients and for family caregivers. The mean scores were computed based on an individual’s responses to the items that do not have missing responses. Using the NORM 2.03 program (Schafer 1997), MI was performed to impute missing values of all variables with incomplete data. While the general practice in MI involves imputation of item scores in order to minimize estimation bias, MI was performed on FACT-G subscales scores for this study due to its small sample size; total scores were computed after MI of subscale scores. Similarly, MI was performed on Kessler distress total scores for the patients and for the family caregivers, and not on individual items of the distress scales. With only 57 observations in the study, imputing individual item scores will substantially expand the number of variables. In the case of the FACT-G scale comprising of 26 items across 4 subscales, each item would constitute an additional variable for MI, if item-level imputation was used. The consequence of a relatively large number of variables to be estimated in a small sample size is the failure to produce results, following iterations of the maximum likelihood estimation to obtain convergence of solutions.  Consistent with experts’ recommendations, 10 datasets were imputed in the study, as imputing more than 10 datasets will have marginal gain in efficiency (Rubin, 1987). Imputed values were not  47  rounded because rounding constitutes adding unnecessary random error variance to the imputed values (Graham, 2012). Considering the small sample size in the study, a small variation may have a significant impact on the statistical significance of the parameter estimates (e.g. one outlier may influence the results substantially.  3.4.5 Statistical analysis methods SPSS 22 software was used for statistical analysis. The assumption of normality for all continuous variables within the combined data sample and separately in the Chinese-speaking and Anglophone groups was tested. Kurtosis and skewness statistics of +2.58 were used as the criterion of normality for small sample size (Field, 2005); boxplots were also examined for cases of extreme outliers. The assumption of linearity between the QOL dependent variables (QOL total scores, Physical well-being subscale, Social well-being subscale, Emotional well-being subscale, Functional well-being subscale) and all other continuous variables across the sample groups was tested through scatterplots. Pearson correlations were computed, to assess the bivariate associations between the continuous variables. To examine the associations between the dependent variables and the categorical variables, each categorical variable was regressed on the QOL total scores and on the FACT-G subscales in separate regression models. T-tests were conducted to test group difference for the continuous variables. The chi-square test was used to investigate difference between the Chinese-speaking and Anglophone groups for the categorical variables. 3.4.5.1 Multiple linear regression  Multiple linear regression was performed using the combined sample, and the respective Chinese-speaking and Anglophone samples for comparison across groups, to test for: i) direct effects of patient, illness-related, social/family, and family caregiver factors as well as patient and family caregiver  48  distress on patient QOL, and ii) indirect effect of the factors on patient QOL through patient and family caregiver distress. First, each QOL dependent variables (QOL total scores and QOL subscales scores) was regressed on the patient and family caregiver distress variables and on the significant covariates (as informed by the bivariate analyses). These regressions were conducted to assess for effects of the distress variables on QOL, while controlling for the covariates. Second, the distress variables were regressed on the covariates that had statistically significant bivariate associations, to assess for the direct effect of the factors on patient distress. Third, the indirect effects of the associated factors on QOL were assessed, first by regressing the QOL variables on the covariates that were significant in step 1 of the multiple regression, followed by computation of the indirect effects using results from step 1 and step 2 of the regressions. Indirect effects were computed by taking the effect of distress on QOL from the multiple regression at step 1 and multiplying it by the effect of the identified factor on distress, as informed by results from step 2 of the multiple regression (MacKinnon, 2008). Tests for group difference in the effects of the independent variables on QOL (and subscales) in each multiple regression model were conducted, using interaction terms of the grouping variable and each independent variable (1=Chinese-speaking; 0=Anglophone). From the study framework, the bidirectional arrows between patient distress and family caregiver distress variables suggest the possibility that the distress variables could influence one another, representing a reciprocal effect. The interrelatedness of the distress variables was first examined through the bivariate analysis, and followed up with multiple regression where the effects of the factors on QOL were controlled. Assessing the impact of the reciprocal associations between the distress variables on QOL would involve examining the looping effects on the direct and indirect effects on QOL, where sufficient degrees of freedom would be required to estimate the model (MacKinnon, 2008). Considering  49  the limitation in estimating such effects in a small sample, repeated mediation arising from reciprocal associations between the distress variables was not examined in this study.   The assumptions for normality, linearity, outliers, homoscedasticity, outliers, independence of errors (residuals), and multicollinearity of residuals were tested for each regression model across all groups (Chinese-speaking, Anglophone and combined sample, where, n = 29, 28 and 57 respectively), using the criteria outlined in Tabachnick and Fidell (Tabachnick & Fidell, 2007). Kurtosis and skewness statistics of +2.58 were used as the criterion of normality for small sample size (Field, 2005). Residuals scatterplots of each model were examined for compliance to the assumptions of normality, linearity and homoscedasticity between predicted dependent variable scores and errors of prediction. With reference to the criteria indicated in Tabachnick and Fidell (2007), the assumptions tested included: (i) residuals (differences between obtained and predicted dependent variable scores) are normally distributed about the predicted dependent variable scores; (ii) residuals have a linear association with the predicted dependent scores; (iii) variance of the residuals about the predicted dependent variable scores is the same for all predicted scores. In addition to screening for extreme outliers by examining boxplots of the independent variables and covariates, residual plots were also assessed for compliance to the assumption that outliers were influencing the regression solution. Durbin-Watson statistics were examined, where values of less than 1 or greater than 3 would denote concerns for non-independence of errors, or presence of autocorrelation of errors. Multicollinearity and singularity (variables redundancy) was also checked. When the correlation between independent variables exceeded .7, and when the largest variance inflation factor (VIF) was greater than 10 (or tolerance (1/VIF) below .1), there would be reason for concern of multicollinearity (Field, 2005).   50  Scatterplots were used to examine the associations between the dependent variables with each of the continuous independent variables. Negative linear associations were largely observed between QOL dependent variables and patient distress. There was, however, no clear linear association between QOL dependent variables and any other independent variables or covariates observed in the scatterplots.  3.4.5.2 Power analysis Based on Cohen (1992) criteria of effect size, the number of participants needed to detect a large (f 2=.35) and medium (f 2=.15) effect size when taking the standard alpha level of .05 and the recommended power of .8 would be 74 and 149 participants, respectively for multiple regression with 18 independent variables. While all covariates in the study are considered in the power calculation, it is plausible that a number of the variables will not be statistically significant in the regression. For instance, with 10 independent variables, and similar convention of alpha (.05) and power level (.8), the number of participants required to detect Cohen’s criteria of large and medium effect size would be 57 and 118, respectively (see Figure 2)(J. Cohen, 1992). The power analysis based on different numbers of independent variables in the multiple regression of the study is charted, as per the power calculation graph below. The final study sample of 57 observations was largely determined by feasibility in primary data collection involving participation from the cancer patients and their family caregivers. With a sample size of 57 observations, there would be sufficient power to detect large effect size for multiple regression with 10 independent variables, at alpha level of .05 and power of .8 (J. Cohen, 1992). Figure 2: Power calculation for large and medium effect sizes in multiple regression  51   3.5 Qualitative methods The qualitative analysis is based on interviews with a subset of the Chinese-speaking survey participants. Adapting Emerson et al.’s (1995) ethnographic approach to analyzing field notes, the initial stage of the data analysis involved open coding of the interview notes to identify themes. Through an iterative process involving repeated viewings of the interview notes and audio recordings, the coding scheme of the study was developed. To compare patient and family caregiver responses to issues of distress and quality of life following a cancer diagnosis, the interview data were examined within the patient-family caregiver dyad, in addition to the coding at the individual level. Further insights drawn from the analysis of the dyadic data were integrated with the findings that emerged from the data at the individual level. 3.5.1 Objectives and phenomenon of interest To examine the cancer experience of Chinese-speaking patients and their family caregivers in relation to their psychological distress and to patient quality of life, semi-structured interviews were conducted. The structural relations of the key elements identified in the conceptual framework were also  52  investigated through the qualitative analysis. In particular, the interrelatedness between patient distress and family caregiver distress on patient QOL was explored through the semi-structured interviews, where clarification on family dynamics following a cancer diagnosis was sought.  Adopting a funnel approach, the interviews began with broad questions on the participants’ overall experience with cancer, followed by more focused questions on their experience of distress and quality of life. Specifically, the interviews were centred on clarifying the phenomena identified in the cross-sectional survey, to provide insights into the following key issues: what constitutes distress for patients and for family caregivers; why might it be more for some and less for others; how might these experiences of distress play out in patient QOL measures; to what extent does the experience of cancer differ, based on cultural background and migrant history. The semi-structured interview guides for patients and for family caregivers are attached, as per Appendices F and G.    The qualitative approach used in the study does not constitute an ethnography, which would involve researcher’s participation as well as observation of the participants’ environment, from the view point of the individuals. In my study, I have used a set of guiding questions during the semi-structured interviews, which were deployed as prompters in the conversations with the participants about their cancer experience. While the questions may appear prescriptive, the approach I have adopted allowed flexibility in the way that the questions were posed. Further, the questions were open to the interpretation by the interviewees, providing an inductive means of soliciting responses from each individual interviewed.  3.5.2 Analytical approach Between February and October 2013, ten Chinese-speaking patients and six family caregivers were interviewed. Of the ten patients who participated in the interview, family caregivers of five patients were also interviewed, providing data of five dyads for the analysis. Where interviewees provided their  53  consent, interviews were audio-recorded (N= 11). Jottings were also taken during the interviews, with an attempt made to capture key verbatim quotes that serve to highlight the comments of participants. These were written up into field notes following each interview; contextual information relating to non-verbal responses to the issues raised in the interviews were also included in the field notes.   The complexities of translation in qualitative research have been widely discussed, particularly in relation to the challenges pertaining to the transfer of meaning across different cultural contexts through interlingual translation (van Nes, Abma, Jonsson, & Deeg, 2010). Notably, qualitative researchers who conduct their studies in one language are faced with the task of understanding and interpreting participant experiences as the respondents wish them to be understood (Santos, Black, & Sandelowski, 2015). For researchers undertaking cross-language studies, Santos and colleagues (2015) note that there is the added complication of transforming the participants’ words – “life-as-told from a source to a target language, to a life-as-told-as-translated and then to a life-as-interpreted-from-translation (p.135).” The difficulty with translation relates not only to language itself but also to the cultural variation between the source and target language (H. Y. Chen & Boore, 2010). In up-front transcription and standard translation, which generally strive for word equivalence, these procedures assume that the same meaning in the source language is also present in all target languages (Larkin, Dierckx de Casterlé, & Schotsmans, 2007; Twinn, 1997). However, Larkin notes that the assumption of true equivalence in such approaches to translation disregards the capacity of each language in creating its own meaning, thereby resulting in the loss of meaning through the interpretation and representation process of the participants’ narratives. Translation dilemmas thus include the efficient use of available time and financial resources, timing of translation spanning across research process, language and cultural competency of translators, as well as the context and analytical role of the translator as they relate to improving the cross-language  54  trustworthiness of the studies (Larkin et al., 2007; Santos et al., 2015; Squires, 2009; Temple & Young, 2004; Twinn, 1997; van Nes et al., 2010). In studies that involve multiple researchers with differing language competencies, the challenges of cross-languages qualitative research become even more prevalent.  In light of my immersion in the culture and language of the participants, I undertook the tasks of data collection and analysis in the study, working with the source languages (Cantonese and Mandarin) and target language (English). Further, my role as the researcher and translator alerted me to problems of cross-cultural interpretation and meaning equivalence. At the same time, I am aware that the translated words of the participants in the chapters that follow are the presentations of my interpretation and there will not be a single correct translation of the text.  The coding scheme in my study was developed through an iterative process involving repeated viewings of the interview notes and audio recordings. Although it is common practice to transcribe and translate (and often back-translate) interviews in qualitative health research, because of my linguistic skills and understanding of the culture as a Chinese person, my analysis was instead developed out of a repeated movement between my interview notes and the audio recordings themselves. This enabled me to continually engage with the interview data as spoken and in context. This approach to data analysis is in line with approaches to interview data common in disciplines such as anthropology (Emerson, Fretz, & Shaw, 1995). Adapting Emerson et al.’s (1995) approach to analyzing field notes, the initial stage of the data analysis involved open coding of the 15 sets of interview notes, to identify and formulate the range of themes in the data. As the analysis progressed, focused coding was employed, with notes coded on the basis of emerging topics that were of particular interest. NVIVO software was used to facilitate the coding of the interview notes. In the initial phase of coding, the data were first categorized according to  55  the three main sections of the interview script: experience with cancer, patient-family caregiver relationship and cultural issues for Chinese cancer patients. Sub-codes were created under each of the main sections, as common themes emerged from the interviews, and further re-categorized, integrating overlapping themes to form broader concepts that reflected the interviewees’ narratives. When coding was completed, each set of interview notes was reviewed in its entirety and the audio recordings were listened to, in full, to contextualize the codes. Where particular subject matters were repeatedly mentioned or that the interviewees had spoken in depth on the issues raised, the codes associated with these topics were highlighted to reflect the participants’ perspective of the emerging themes.  Aligned with the intent to compare patient and family caregiver responses to issues of distress and quality of life following a cancer diagnosis, the interview data were also examined within the patient-family caregiver dyad. Using a parallel approach for analyzing the interview data of the dyads, each set of interview notes was reviewed and insights were developed, with key information extracted in relation to what the patient and the family caregiver member of the dyad said about the respective issue that emerged from the earlier coding of the individual interviews. The audio recordings of the dyads were again listened to for additional information that highlighted the corresponding perspective of the dyadic member towards each of the emerging issues identified in the coding. In each dyad, the similarities and differences of the patient and the family caregiver views on the identified issues were mapped out. The information exacted from the mapping of the dyadic responses was compared with that from the data coded at the individual level; where there was new information, insights drawn from the analysis of the dyadic data were integrated with the findings that emerged from the data at the individual level.  56  3.6 Ethical considerations  The study protocol was reviewed and approved by the University of British Columbia Behavioural Research Ethics Board and the Vancouver Coastal Health (VCH) Research Institute for recruiting study participants in hospitals of VCH, prior to the commencement of data collection. I had explained the requirements and processes of the study to participants over the phone, each of whom were provided a copy of the form available in English or Chinese. As a certified counsellor with the BC Association of Clinical Counsellors, I have used my professional knowledge and skills to attune to participants’ emotional status and pace the survey and semi-structured interviews, so that participants’ welfare was safeguarded during data collection. Information on counselling services at the BC Cancer Agency was also provided to the patients and family caregivers, to ensure that resources were available for participants requiring counselling, to address issues that may have arose from participation in the study. Summary This study used a cross-sectional survey of Chinese-speaking (study population) and Anglophone (comparison group) patients (N = 55) and their family caregivers (N = 40) to examine patient QOL and correlates. Multiple imputation of missing data for incomplete dyads resulted in data for 29 Chinese-speaking and 28 Anglophone dyads. Multiple linear regression and mediation analyses were conducted to examine the predictors of QOL, and its domains, and the mediation effects of patient distress and family caregiver distress.  A subset of the Chinese-speaking survey participants comprising ten patients and six family caregivers, including five patient-family caregiver dyads, also participated in semi-structured interviews. The analysis of culturally-embedded experiences of patients and family caregivers regarding their distress and QOL was conducted, to provide further context and explanation  57  for the findings from the quantitative analysis. The study findings are presented in the following chapter, reporting first on the results from quantitative analyses, followed by results from the qualitative analysis.    58  Chapter 4: Results Results from the quantitative and qualitative analyses are presented in this chapter. Findings from the quantitative analyses are reported, followed by qualitative results. For the quantitative results, a description of the sample of patients and family caregivers (n = 57 dyads) from both the Chinese-speaking (n = 29) and Anglophone (n = 28) groups is first presented. Second, results are described pertaining to the direct effects of patient factors, illness-related factors, social/family factors and family caregiver factors, including the corresponding bivariate and multivariate associations. Third, results relating to the mediation effects of patient and family caregiver distress on patient quality of life (QOL) are provided. Finally, the results of the post-hoc analyses based on the sample of 38 dyads (prior to multiple imputation) are reported to assess the potential implications of missing data. The findings from the qualitative analysis are based on semi-structured interviews with 16 Chinese-speaking survey participants. Insights are offered into the cancer experience of the patients and their family caregivers, in relation to their psychological distress and to patient QOL in the context of relationships in the Chinese family. Particularly, results regarding the interrelatedness between patient distress and family caregiver distress on patient QOL are reported. The background of each interviewee (10 patients and 6 family caregivers) is described, to provide contextual information to the interview responses. Cross-cutting themes relating to patient distress, family caregiver distress and patient QOL that emerged from the interviews are then highlighted and illustrated with quotes from the interviewees.  4.1 Quantitative analysis 4.1.1 Sample description The sample consisted of 29 Chinese-speaking and 28 Anglophone dyads (Table 2. See Appendix H for descriptive summary of 19 dyads with incomplete information). The average age of patients was similar in the Chinese-speaking and Anglophone groups, 57 and 56, respectively. Family caregivers in  59  the Chinese-speaking group were much younger (mean age 39 years) than those in the Anglophone group (mean age 62 years). The duration of caregiving averaged 10 months and was also approximately equal across the two groups. The mean scores of patient distress were lower in the Chinese-speaking group than in the Anglophone group, whereas the mean scores of family caregiver distress were approximately equal in both groups. For patients, the mean QOL scores were higher in the Chinese-speaking group than in the Anglophone group, where Chinese-speaking patients reported higher physical, emotional and functional well-being than patients in the Anglophone group.  Statistically significant differences between groups were found in the number of years that patients (t = 10.07, p < .001) and family caregivers (t = 8.40, p < .001) had lived in Canada, family caregiver age (t = 3.34, p = .003), and patient (χ2 = 32.31, p < .001) and family caregivers (χ2 = 29.49, p = .04) English fluency. Not surprisingly, the number of years that patients and family caregivers lived in Canada was much smaller in the Chinese-speaking sample than in the Anglophone sample. Family caregivers in the Chinese-speaking sample were also younger, whereas the average age of patients was similar in both groups. As expected, the percentage of patients and family caregivers who reported advance English fluency was smaller in the Chinese-speaking group. The groups were similar with respect to all other variables.     60  Table 2: Descriptive summary of patients and family caregivers in combined sample, Chinese-speaking and Anglophone group  Combined Group (N = 57) Chinese Group (n = 29) Anglophone Group (n = 28)  Patient Caregiver Patient Caregiver Patient Caregiver Variables Pooled Mean (SD) QOL total 83.4 (17.6) -- 85.9 (17.2) -- 80.9 (18.0) -- Physical well-being  20.9 (6.2) -- 21.7 (5.3) -- 20.2 (7.0) -- Social well-being  24.0 (4.41) -- 23.7 (4.7) -- 24.3 (4.1) -- Emotional well-being 18.9 (5.4) -- 20.3 (4.6) -- 17.5 (6.0) -- Functional well-being 19.6 (5.9) -- 20.3 (5.6) -- 18.9 (6.3) -- Distress total score   5.2 (4.7)    4.6 (4.6)   5.0 (4.3)   4.6 (4.6)   5.4 (5.1)   4.6 (4.7) Agea 58.6 (11.8) 50.7 (22.2) 57.5 (10.6) 39.2 (19.5) 56.2 (13.1) 62.5 (18.1) Years lived in Canadaa 38.8 (21.4) 37.9 (26.8) 21.9 (9.7) 16.1 (11.8) 56.4 (15.3) 60.5 (17.3) Months of caregiving -- 10.5 (6.3) -- 10.5 (7.2) -- 10.6 (5.2)  Counts (%) Sex:         Femalec 45 (78.8) 25 (44.6) 23 (77.8) 15 (51.6) 22 (79.9) 10 (37.3)   Male  12 (21.2) 32 (55.4)   6 (22.2) 14 (48.4)   6 (20.1) 18 (62.7) Education:           Post-secondaryc  19 (33.5) 18 (32.8)   6 (22.6)   5 (19.2) 12 (44.7) 13 (47.1)   High school  23 (39.4) 11 (18.5) 14 (47.1)   4 (13.2)   9 (31.4)   7 (23.9)   University or advanced degree 15 (27.1) 28 (48.7)   9 (30.3) 20 (67.6)   7 (23.9)   8 (29.0) English fluencyb:           Beginner or intermediatec  21 (37.1) 12 (21.4) 21 (73.3) 12 (39.9)   0 (0)   1 (2.3)   Advanceda 36 (62.9) 45 (78.6)   8 (26.7) 17 (60.1) 28 (100) 27 (97.7) Patient-family caregiver relationship type:         Extended family/relatives or close friendsc  --   9 (16.1) --   4 (15.3) --   5 (16.9)   Spouse/partner -- 32 (56.1) -- 11 (36.8) -- 21 (76.1)   Children -- 16 (27.8) -- 14 (47.9) --   2 (7.0) Cancer site:         Esophagus, lung, lymphoma or nasopharyngealc    7 (12.3) --   3 (10.3) --   4 (14.2) --   Breast  28 (49.1) -- 16 (55.2) -- 12 (42.9) --   Bladder or colorectal 13 (22.8) --   6 (20.7) --   7 (25.0) --   Ovarian, prostate or uterine   9 (15.8) --   4 (13.8) --   5 (17.9) -- Cancer stage:         Stage I  14 (24.6) --   6 (20.7) --   8 (28.6) --   Stage IIc 18 (31.5) --   8 (27.6) -- 10 (35.7) --   Stage III  25 (43.9) -- 15 (51.7) -- 10 (35.7) -- Treatment type:         Chemotherapy, drugs, radiation or surgeryc  14 (25.4) --   6 (20.6) --   9 (30.4) --   More than 1 treatment type 43 (74.6) -- 23 (79.4) -- 19 (69.6) -- Note. a = t-test of group difference between Chinese-speaking and Anglophone samples significant at 5% level for the following variables: caregiver age, years patient lived in Canada, years caregiver lived in Canada. b = Chi-square of independence between Chinese-speaking and Anglophone samples significant at 5% level for the following variables: patient with advanced English, caregiver with advanced English. c = referent in the regression analyses. 61  4.1.2 Bivariate analyses  The significant bivariate associations with the QOL variables in the combined sample and for the Chinese-speaking and Anglophone groups are presented in Tables 3, 4 and 5 (see Appendices I through N for detailed results of all bivariate analyses). In addition to patient and family caregiver distress, only those variables that had significant associations with any of the QOL variables in the combined or in the separate groups are considered. In the combined sample and Anglophone group, statistically significant bivariate correlations (with QOL total) were observed only for patient distress and patient age. Patient distress was negatively related to QOL total, and patient age was positively related to QOL total. Patient age was not significantly related to QOL total in the Chinese-speaking group. None of the other variables had statistically significant associations with QOL total in the Anglophone and Chinese-speaking groups.  Patient distress was negatively correlated with all QOL subscales in both Chinese-speaking and Anglophone groups, as well as in the combined sample. In both groups, the largest correlation pertained to the association between distress and emotional well-being. In the combined sample, patient age was positively correlated with social well-being and emotional well-being, but not with physical well-being or with functional well-being. Patient age was not significantly correlated with any of the QOL subscales in the Chinese-speaking group, although a negative association was found between Stage III cancer (versus Stage II cancer) and physical well-being in this group. Having a high school education (versus having post-secondary education) was also found to be negatively associated with social well-being in the Chinese-speaking group and in the combined sample. In the Anglophone group, both patient age and years lived in Canada were positively correlated with emotional well-being, but not with the other QOL subscales.   62  Table 3: Summary of bivariate associations with QOL total and QOL subscales for combined sample (N=57) Note. a = Pearson correlations. b = standardized regression coefficients of dummy variables for education level. *p < .05, **  Note. a = Pearson correlations. b = standardized regression coefficients of dummy variables for education level.  c = standardized regression coefficients of dummy variables for cancer stage.  *p < .05, **p < .01.       Variables Patient distress a Caregiver distress a Patient age a Patient with high school education (vs post-secondary education)b Patient with university/ advanced degree (vs post-secondary education)b QOL total   -.80** -.10         .30*             -.10 -.12 Physical well-being -.62**  .11 .11  -.13 -.18 Social well-being -.54** -.18    .32*             -.36* -.15 Emotional well-being -.76** -.06      .45**   .06 -.14 Functional well-being -.66** -.23       .14   .06             .09 Patient distress -    .09        -.34*   .07             .12 Caregiver distress    .09 -        -.14  -.01  -.17 Variables Patient distress a Caregiver distress a Patient with high school education (vs post-secondary education)b Patient with university/ advanced degree  (vs post-secondary education)b Stage I cancer (vs  Stage II cancer)c Stage III cancer (vs  Stage II cancer)c QOL total   -.75**      -.11 -.27 -.12  .05 -.23 Physical well-being -.57**      -.03 -.14 -.11  .04  -.37* Social well-being -.57**      -.13    -.53* -.24  .11 -.04 Emotional well-being -.74**      -.04 -.13 -.12 -.08 -.09 Functional well-being -.68**      -.18 -.15 .04 .09 -.18 Patient distress -       .08  .07 .13 .04   .24 Caregiver distress   .08 -             -.07   -.18 .19   .21 Table 4: Summary of bivariate associations with QOL total and QOL subscales for Chinese-speaking group (n = 29)  63  Table 5: Summary of bivariate associations with QOL total and QOL subscales for Anglophone group (n = 28)  Note. All values are Pearson correlations. *p < .05, **p < .01.   Notwithstanding the theoretical importance of family caregiver distress, none of the correlations between this variable and the dependent variables were statistically significant in the study sample. Particularly in the Chinese-speaking group, the magnitudes of the correlations for caregiver distress were relatively small, and did not provide strong evidence to support the meaningfulness or importance of the effects of family caregiver distress on patient QOL total and on the subscales. In the Anglophone sample, the largest associations were observed between caregiver distress and social well-being and functional well-being, with correlations of -.26 and -.29, respectively. Caregiver distress was, however, found to be positively associated with physical well-being in the Anglophone group and in the combined sample.Variables Patient  distress Caregiver distress Patient age Years patient lived in Canada Caregiver age QOL total        -.86**       -.10 .43*           .34     .25 Physical well-being -.66**           .22          .28 .34        .20 Social well-being -.54** -.26          .13           .03     .12 Emotional well-being -.79** -.09   .59**           .58**     .39 Functional well-being -.64** -.29          .16           .04     .05 Patient distress -            .10         -.48**          -.34    -.43* Caregiver distress         .10 -         -.32          -.08    -.17  64  4.1.3 Multiple regression analyses 4.1.3.1 Direct effects of patient, illness-related, social/family and family caregiver factors on patient QOL The first hypothesis was evaluated by examining the direct effects of patient, illness-related, social/family and family caregiver factors on patient QOL, after controlling for the distress variables.  Main hypothesis 1: Patient factors (age, sex, education, migration duration, English fluency), illness-related factors (cancer site/stage, treatment type), social/family factors (caregiving duration, patient-family caregiver relationship), and family caregiver factors (age, sex, education, migration duration, English fluency) are directly associated with patient QOL. Patient distress and family caregiver distress were identified as key independent variables in relation to patient QOL, as theorized in the framework. Patient age was included as an independent variable in the regression on QOL total, being the only covariate that was significantly related with QOL total, as informed by the bivariate analyses. To assess the direct effects of the factors across the QOL subscales, regressions of each QOL subscale on the covariates that were significantly related to the respective QOL dependent variable (as informed by the bivariate analyses), as well as the distress variables, were performed. The covariates of the QOL subscales included cancer stage, patient education, years lived in Canada. Patient age, which is correlated with social well-being and emotional well-being, was also entered in these QOL subscales regressions. The regression models and their associated hypotheses tested under Main Hypothesis 1 are as follows:     65  Main hypothesis 1: Regression models and associated hypotheses  Regression Models  Hypotheses  Model 1a:  QOL total = b0 + b1*Patient distress + b2*Caregiver distress +                       b3*Patient age + e   Model 1b:  Physical well-being = b0 + b1*Patient distress +                                      b2*Caregiver distress + b3*Stage I cancer +                                     b4*Stage III cancer + e                                                                                   Model 1c:  Social well-being = b0 + b1*Patient distress + b2*Caregiver distress +                                   b3*Patient age + b4*Patient with high school                                  education + b5*Patient with university or                                  advanced degree + e                     Model 1d:  Emotional well-being = b0 + b1*Patient distress +                                         b2*Caregiver distress + b3*Patient age +                                        b4*Years patient lived in Canada + e                  H0 : b3 = 0 H1 : b3 ≠ 0   H0 : b3 = 0; H0 : b4 = 0 H1 : b3  ≠ 0; H1 : b4 ≠ 0    H0 : b3 = 0; H0 : b4 = 0;  H0 : b5 = 0  H1 : b3  ≠ 0; H1 : b4 ≠ 0; H0 : b5  ≠ 0   H0 : b3 = 0; H0 : b4 = 0  H1 : b3  ≠ 0; H1 : b4 ≠ 0 Model 1e:  Functional well-being = b0 + b1*Patient distress +                                          b2*Caregiver distress + e                                                                                                          No hypotheses associated withpatient, illness-related, social/family and family caregivers factors were tested, as no covariates were significantly related with functional well-being in the bivariate analyses   Note. b0 = constant; bi = unstandardized regression coefficients; e = residual    66  Assumptions for linear regression were met in all regression models. None of the interaction terms used in testing for group differences in the effects of the independent variables on QOL (and subscales) in the multiple regression models were statistically significant. Pooled unstandardized and standardized regression coefficients across the imputed datasets were reported; statistics provided in parentheses indicates range over the imputed datasets. The regression results pertaining to Main Hypothesis 1 are presented in Table 6.  67  Table 6: Main Hypothesis 1: Summary of multiple linear regression results  Model: Dependent variable Independent variables Combined Sample   (n = 57) Chinese Group (n = 29) Anglophone Group     (n = 28)   B   β   B   β   B   β Model 1a: QOL total Patient distress -2.99**-  -.79** -2.98** -.79** -2.98** -.79** Caregiver distress -0.12 -.03 -0.26 -.07 -0.07 -.02 Patient age  0.04  .03  0.13  .09  0.02  .01 R2 Adjusted R2 (.64, .67)** (.62, .69)** (.52, .64)** (.46, .60)** (.74, .77)** (.70, .74)** Model 1b: Physical well-being   Patient distress -0.85** -.64** -0.61* -.46* -1.01** -.76** Caregiver distress  0.25  .19  0.04  .03  0.43  .32 Stage I cancer (vs Stage II cancer) -1.15 -.08  0.63  .05 -2.79 -.20 Stage III cancer (vs Stage II cancer) -1.86 -.15 -3.19 -.26 -1.40 -.12 R2 Adjusted R2 (.36, .51)** (.31, .48)** (.36, .56)* (.25, .48)* (.48, .59)** (.39, .51)** Model 1c: Social well-being  Patient distress -0.46** -.49** -0.59* -.63** -0.42* -.45* Caregiver distress -0.11 -.12 -0.1 -.12 -0.18 -.19 Patient age  0.3  .09  0.08  .20 -0.01 -.04 Patient with high school education    (vs post-secondary) -2.72* -.31* -4.03* -.46* -1.25 -.14 Patient with university/advanced degree    (vs post-secondary) -0.86 -.09 -1.51 -.16 -0.92 -.10 R2 Adjusted R2 (.38, .43)**       (.32, .37) (.44, .58)* (.32, .49)* (.34, .44) (.19, .32) Model 1d: Emotional well-being Patient distress -0.79** -.68** -0.76** -.66** -0.76** -.65** Caregiver distress  0.02  .02 -0.07 -.06 -0.02  .02 Patient age  0.14**  .3**  0.13  .28  0.03  .07 Years patient lived in Canada -0.05 -.19 -0.08 -.30  0.12  .47 R2 Adjusted R2 (.59, .69)** (.56, .67)** (.50, .70)** (.41, .66)** (.72, .77)** (.67, .74)** Model 1e: Functional well-being Patient distress -0.81** -.64** -0.88** -.69** -0.75** -.59** Caregiver distress -0.23 -.18 -0.16 -.12 -0.3 -.23 R2 Adjusted R2 (.40, .54)** (.38, .53)** (.37, .62)** (.32, .59)** (.39, .57)** (.34, .54)** Note. A range in R-square values were generated as a result of multiple imputation. B = unstandardized regression coefficients. β = standardized regression coefficients *p < .05, **p < .01.  68  Model 1a. In Model 1a, the variables hypothesized to predict QOL total were patient distress, family caregiver distress and patient age. Model 1a had the greatest explained variance in the Anglophone group (R2 between .74 and .77, adjusted R2 between .70 and .74, p < .001). From the statistical results, there was insufficient evidence to conclude that patient age was directly associated with patient QOL total across all groups. Only patient distress was significant in the combined sample and in the two groups, where patient distress was negatively related to QOL total. The effect of patient distress on QOL total was similar across the Chinese-speaking and Anglophone groups (β = -.79, p < .001).  Model 1b. In model 1b, the variables hypothesized to predict patient physical well-being included patient distress, family caregiver distress and cancer stage. The explained variance was greatest in the Anglophone group (R2 between .48 and .59, adjusted R2 between .39 and .51, p < .001). There was insufficient evidence to conclude that having Stage I or Stage III cancer, versus having Stage II cancer, was directly associated with patient physical well-being across all groups. Patient distress was negatively related to physical well-being, though the effect was smaller in Chinese-speaking group than in Anglophone group (β = -.46 and -.76, p = .01 and < .001, respectively).  Model 1c. In Model 1c, the hypothesized predictors of patient social well-being were patient distress, family caregiver distress, patient age and patient education. The explained variance was greatest in the Chinese-speaking group (R2 between .44 and .58, adjusted R2 between .32 and .49, p < .001).  There was insufficient evidence to conclude that patient age was directly associated with patient social well-being. In the Chinese-speaking group, patient high school education was found to be negatively associated with patient social well-being, where social well-being of patients was lower among those with high school education, compared with those having post-secondary education (β = -.46, p = .03).  69  The effect of patient distress on social well-being was observed to be larger in the Chinese-speaking group than in the Anglophone group (β = -.63 and -.45, p = .001 and .02, respectively).  Model 1d. In Model 1d, the hypothesized predictors of patient emotional well-being were patient distress, family caregiver distress, patient age and years patient has lived in Canada. Model 1d had the greatest explained variance in the Anglophone group (R2 between .72 and .77, adjusted R2 between .67 and .74, p < .001). Patient age was positively related with emotional well-being in the combined sample (β = .30, p = .008). However, there was insufficient evidence to conclude that the variable “years patient has lived in Canada” was associated with patient emotional well-being in the study sample. Patient distress was negatively related to emotional well-being and the magnitudes of the effects were similar for both the Chinese-speaking and Anglophone group (β = -.66 and -.65, respectively, p < .001).  Model 1e. In Model 1e, patient distress and family caregiver distress were hypothesized to predict patient functional well-being. Model 1e had the greatest explained variance in the Chinese-speaking group     (R2 between .37 and .62, adjusted R2 between .32 and .59, p < .01). Consistent with the results of all other regression models, patient distress remained a significant predictor of patient QOL. The effect of patient distress on functional well-being was found to be larger in Chinese-speaking group than in the Anglophone group (β = -.69 and -.59, respectively, p < 0.001).  Summary of multiple regression analyses of direct effects: In summary, the greatest explained variance was achieved for Model 1a, where QOL total was regressed on patient distress, caregiver distress and patient age. Among the covariates identified in the study, direct association was found between patients with high school education and patient social well-being in the Chinese-speaking group; the social well-being of patients was lower among those with high school education, compared with those having post-secondary education. Patient age was also found to be directly associated with patient emotional well-being in the combined sample, where older patients  70  were found to have higher emotional well-being. However, patient age was not found to be significant in the Chinese-speaking nor the Anglophone group in the multiple regression analyses. There was also insufficient evidence to conclude that other covariates (illness-related, social/family and family caregiver factors) were directly associated with QOL total and with the subscales of QOL.  From the regression results, variation in the negative effects of patient distress on QOL subscales was observed across the models. In the Chinese-speaking group, the negative effect of patient distress on physical well-being was the smallest relative to other QOL subscales. On the other hand, the effect of patient distress on functional well-being was the largest; as such, highlighting the role of patient distress in the functional well-being of the Chinese-speaking participants. In the Anglophone group, the negative effect of patient distress on social well-being was the smallest among the QOL subscales. Conversely, the effect of patient distress on physical well-being was noted to be the largest, compared to other QOL subscales, hence, highlighting the role of patient distress in patient physical well-being of the Anglophone participants. 71  4.1.3.2 Indirect effects of patient, illness-related, social/family and family caregiver factors on patient QOL through patient and family caregiver distress Hypotheses 2 and 3 were tested by examining the indirect effects of factors of patient, illness-related, social/family and family caregiver on patient QOL, as mediated by patient and family caregiver distress. Main hypothesis 2: Patient factors (age, sex, education, migration duration, English fluency), illness-related factors (cancer site/stage, treatment type), social/family factors (caregiving duration, patient-family caregiver relationship), family caregiver factors (age, sex, education, migration duration, English fluency) are also directly associated with patient distress and with family caregiver distress. Based on the bivariate analyses, only patient age and family caregiver age were found to be significantly related to patient distress; none of the patient, illness-related, social/family and family caregiver variables were significantly related with family caregiver distress. To assess the direct effects of patient age and family caregiver age on patient distress, multiple linear regression was performed in the combined sample (Chinese-speaking and Anglophone groups), the Chinese-speaking, and the Anglophone groups. The regression model and associated hypotheses tested under Main Hypothesis 2 are:  Model 2:  Patient distress = b0 + b1Patient age + b2Caregiver age + e H0 : b1 = 0; H0 : b2 = 0 H1 : b1  ≠ 0; H1 : b2 ≠ 0  72  Regression results pertaining to Main Hypothesis 2 are presented in Table 7. From the regression results of patient distress on patient age and family caregiver age in the combined sample, patient age was significantly related to patient distress (β = -.3, p = .05), where older patients reported lower levels of distress. However, family caregiver age was not significantly related with patient distress. When the hypotheses were tested in the Chinese-speaking group and again in the Anglophone group, no statistically significant results were observed. Table 7: Main Hypothesis 2: Summary of multiple linear regression results  Model: Dependent variable Independent variables   Combined Sample (n = 57) Chinese Group (n = 29) Anglophone Group (n = 28)   B β B β B β Model: Patient distress  Patient age  -0.12+ -.30* -0.05 -.13 -0.13 -.34 Caregiver age  -0.02 -.11 -0.03 -.14 -0.06 -.30 Note. B = unstandardized regression coefficients. β = standardized regression coefficients. *p < .05, +p =.05   Main Hypothesis 3: Patient distress and family caregiver distress are directly related to patient QOL; and thus, mediate the associations between patient QOL and the factors of patient, illness-related, social/family and family caregiver.  From the earlier analyses on the direct association between the factors, including the distress variables, and patient QOL, patient distress was found to be directly related to patient QOL total and the subscales of QOL. Further, patient age was also directly associated with patient distress. Given these results, it is possible that the effect of patient age on patient QOL is mediated by patient distress. From the bivariate analyses, patient age was significantly related only to QOL total, social well-being and emotional well-being; patient age was not found to be related with physical well-being or with functional well-being. Family caregiver distress was not found to be associated with any of the dependent variables in the study.     73   Given the conceptual significance of family caregiver distress in the study framework, the caregiver distress variable was entered in the regressions tested under Main Hypothesis 3, when assessing the mediation effect of patient distress on patient QOL total, social well-being and emotional well-being. The selections of covariates in the respective model tested under Main Hypothesis 3 were congruent with the regression models identified under Main Hypothesis 1. Main Hypothesis 3: Regression models and associated hypotheses    Regression Models  Mediation hypotheses Model 3a i) Patient distress = b0 + bi1Patient age + bi2Caregiver distress + e   ii) QOL total = b0 + bii1Patient distress + bii2Patient age +                       bii3Caregiver distress + e                               Model 3b:  i) Patient distress = b0 + bi1Patient age + bi2patient with high school                             education + bi3Patient with university/advanced                             degree + bi4Caregiver distress + e  ii) Social well-being = b0 + bii1Patient distress + bii2Patient age +                                   bii3Patient with high school education +                                   bii4Patient with university/advanced degree +                                  bii5Caregiver distress + e  Model 3c:  i) Patient distress = b0 + bi1Patient age + bi2Years patient lived in                                         Canada + bi3Caregiver distress + e  ii) Emotional well-being = b0 + bii1Patient distress + bii2Patient age +                                                     bii3Years patient lived in Canada +                                                     bii4Caregiver distress + e  H0 : bi1 = 0  H1 : bi1 ≠ 0  H0 : bii1 = 0  H1 : bii1 ≠ 0   H0 : bi1 = 0     H1 : bi1 ≠ 0    H0 : bii1= 0 H1 : bii1 ≠ 0     H0 : bi1 = 0 H1 : bi1 ≠ 0  H0 : bii1= 0     H1 : bii1 ≠ 0   Note. b0 = constant; b = unstandardized regression coefficients; e = residual. Subscript “i” referred to the model where patient distress was the dependent variable. Subscript “ii” referred to the model where QOL (and subscales) was the dependent variable.     74  Model 3a: Indirect effects of patient age on QOL total through patient distress: From the simple regression that assessed the effect of patient age on QOL total, patient age was positively related to QOL total in the combined sample and the Anglophone group (β = .3 and .43, p = .03 and .02, respectively). However, the effect of patient age was not significant in the Chinese-speaking group. When patient distress was regressed on patient age and caregiver distress, patient age was found to be negatively related to patient distress in both the combined sample and Anglophone group (β = -.34 and -.50, p = .02 and .007, respectively). In the multiple linear regression of QOL total, patient age was no longer significant in the combined sample and in the Anglophone group. On the other hand, patient distress remained negatively related to QOL total in all groups (β = -.79, p < .001).   The indirect effects for the combined sample and for the Anglophone group were .27 (-.34 x -.79) and .40 (-.50 x -.79), respectively. Joint test of significance (Fritz & MacKinnon, 2007) suggested the indirect effects were significant (p < .05), in the combined sample and Anglophone group. The percentages of mediation for the combined sample and Anglophone group were 90% (.27/.30) and 93% (.40/.43), respectively. Using Kenny’s recommendation on assessing indirect effects, with over 80% of mediation found in the combined sample and Anglophone group, it may be said that patient distress has completely mediated the positive effect of patient age on QOL total in the Anglophone group and the combined sample, where the effect of patient age on patient QOL total was through patient distress. While patient age was found to be a significant factor influencing patient QOL in the study, its effect may be completely explained by the presence of patient distress. Results of the mediation analyses relating to the indirect effects of patient age on QOL total through patient distress (Model 3a) are presented in the path diagrams as follows.       75  Figure 3: Model 3a - Indirect effects of patient age on QOL total through patient distress, combined sample           *p < .05, **p < .01  Figure 4: Model 3a - Indirect effects of patient age on QOL total through patient distress, Anglophone group         *p < .05, **p < .01  Model 3b: Indirect effects of patient age on social well-being through patient distress: In the simple regression of social well-being on patient age, patient age was found to be significantly related to social well-being in the combined sample (β = .32, p = .02), where patient age was positively related to patient social well-being. The variables were not found to be significantly related in the Chinese-speaking and Anglophone groups. When patient distress was regressed on patient age, caregiver distress and the covariate associated with social well-being in the combined sample, patient age was found to be Patient age QOL total Patient  distress -0.34* -.79** .03 Caregiver  distress .04 -.03 Patient age QOL total Patient  distress -.50* -.79** .01 Caregiver  distress -.07 -.02  76  negatively related to patient distress (β = -.33, p = .03). In the multiple regression of social well-being, patient distress was negatively related to social well-being in the combined group (β = -.49, p < .001). Patient age was, however, no longer significantly related to social well-being.   The indirect effect was .16 (-.33 x -.49); joint test of significance (Fritz & MacKinnon, 2007) suggested the indirect effect was significant. With 50% (.16/.32) of total effects explained by mediated effects, we may say that patient distress partially mediated the positive effect of patient age on social well-being in the combined sample. Results of the mediation analyses relating to the indirect effects of patient age on social well-being through patient distress (Model 3b) are presented in the path diagrams as follows.  Figure 5: Model 3b - Indirect effects of patient age on social well-being through patient distress, combined sample              *p < .05, **p < .01  Model 3c: Indirect effects of patient age on emotional well-being through patient distress: Based on the simple regression, patient age was found to be positively related to patient emotional well-being in the combined sample and in the Anglophone group (β = .45 and .59, respectively, p = .001). However, no significant results were obtained in the Chinese-speaking group. When patient distress was regressed -.33* .09 Social   well-being Patient distress Patient age Caregiver distress Patient with high school education Patient with university or advanced degree -.49** -.09 .05 .01 .05 -.12 -.31*  77  on patient age, caregiver distress and the covariate associated with emotional well-being in the combined sample, patient age was found to be negatively related to patient distress (β = -.36, p = .04). Patient age was, however, not significantly related with patient distress in the Anglophone group. In the multiple regression of emotional well-being, both patient distress and patient age were significantly related with emotional well-being in the combined sample (β = -.68 and .30, p < .001 and .007, respectively). Patient age was not significantly related to emotional well-being in the Anglophone group.  The indirect effect in the combined sample and Anglophone group were .24 (-.36 x -.68) and .33 (-.51 x -.65), respectively. Joint test of significance suggested the indirect effect in the combined sample was statistically significant (p < .05). Approximately 53% (.24/.45) of the total effect of patient age was mediated by patient distress in the combined group, where it may be said that patient distress partially mediated the positive effect of patient age on patient emotional well-being. In addition to the direct effect of patient age on patient emotional well-being, the effect of patient age may also be explained by the presence of patient distress, where part of the effects of patient age on emotional well-being can be attributed to the patient’s level of distress. Results of the mediation analyses relating to the indirect effects of patient age on emotional well-being through patient distress (Model 3c) are presented in the path diagrams as follows in Figures 6 and 7.               78  Figure 6: Model 3c - Indirect effects of patient age on emotional well-being through patient distress, combined sample            *p < .05, **p < .01   Figure 7: Model 3c - Indirect effects of patient age on emotional well-being through patient distress, Anglophone group           *p < .05, **p < .01  Summary of mediation analyses:  The role of patient distress in explaining patient QOL was revealed in the assessment of indirect effects of patient age on patient QOL total and on the QOL subscales in the combined sample and Anglophone group. The effects of patient age on patient QOL total appeared to be completely mediated by patient distress. Examining the mediation effects in the QOL subscales, the effects of patient age on patient social well-being and on patient emotional well-being were partially mediated by patient distress -.36* .30** Emotional well-being Patient age Caregiver distress Years patient lived in Canada -.68** .04 .04 .02 -.19 Patient distress -.51 .07 Emotional well-being Patient distress Patient age Caregiver distress Years patient lived in Canada -.65** .02 -.06 .02 .47  79  in the combined sample. No mediated effects of patient age on the QOL subscales were detected in the sub-group analyses of Chinese-speaking and Anglophone patients.   4.1.4 Post hoc analyses The analyses were repeated using the sample of 38 dyads with complete information. These results were compared with those from the full study sample (38 dyads, 17 patients without family caregiver participation, and two family caregivers without patient participation) where multiple imputation (MI) was conducted, in order to determine the extent to which conclusions may have differed when different approaches in handling missing data were used. A summary of the bivariate results is presented in Table 8.  Consistent with the results from the analyses of the full sample involving MI, patient distress was significantly related to patient quality of life total scores and all the QOL subscales (physical, social, emotional, and functional well-being). Results of bivariate analyses on family caregiver distress were comparable to those obtained when using the full sample. Results from the sample of 38 dyads revealed that patient age was significantly related to QOL total, as well as to social and emotional well-being. Similar to the bivariate analyses results of the full sample, patient age was also significantly related to patient distress. There was a significant association between patient level of education and social well-being, where social well-being was lower among those patients with high school education, compared to those with post-secondary education. From the smaller sample of 38 dyads, there were also other significant associations not found in the full sample. Years the patient had lived in Canada was negatively related to family caregiver distress. The factor relating to family caregiver with advanced English variable was negatively related to patient physical well-being. Similar to the results of the full sample, the bivariate associations relating to family caregivers were not significant.  80  Table 8: Summary of bivariate associations (correlations and regression coefficients) with QOL total and QOL subscales (n = 38)  Variables Patient distress a Caregiver distress a Patient age a Patient with high school education (vs post secondary education)b Patient with university or advanced degree  (vs post secondary education)b Caregiver with advanced English  (vs beginner or intermediate English)c Years patient lived in Canada a QOL total -.84** -.13  .39* -.24 -.14 -.22  .04 Physical    well-being -.71**  .01  .24 -.30 -.12 -.34* -.10 Social     well-being -.60** -.15  .32* -.42* -.21  .03  .24 Emotional    well-being -.76** -.10  .51** -.03 -.20 -.28  .01 Functional    well-being -.72** -.19  .24 -.12  .05 -.11  .02 Patient distress    -  .14 -.39*  .26  .22  .12 -.14 Caregiver    distress  .14    - -.23  .09 -.02 -.02 -.34* Note. a = Pearson correlations. b = standardized regression coefficients of dummy variables for education level.  c = standardized regression coefficient of binary variable for English fluency. * p < .05, **p < .01.    Multiple linear regression was performed on the sample of 38 dyads, to test Main Hypothesis 1 investigating the association between patient, illness-related, social/family and family caregiver factors, and patient QOL. Additional regressions were run for regression models related to physical well-being and emotional well-being, as informed by the bivariate analyses results of n = 38. Additional regressions conducted were: i)  Physical well-being = b0 + b1Patient distress + b2Caregier distress +    b3Caregiver with advanced English + e ii) Emotional well-being = b0 + b1Patient distress + b2Caregiver distress + b3Patient age + e  Results of the regressions on Main Hypothesis 1 when using the sample of 38 dyads were consistent with the findings on the full sample, except that the factor regarding educational level was no longer found related to patient social well-being in the smaller sample. From the results of the additional  81  regression for physical well-being, the factor relating to family caregiver with advanced English remained significant in predicting patient physical well-being. The results of the multiple linear regression are presented in Table 9, which follows a brief discussion.  As noted above on the bivariate results using the sample of 38 dyads, only patient age was significantly related with patient distress. Although the factor relating to years patient has lived in Canada was significantly associated with family caregiver distress, the caregiver distress variable was not associated with the dependent variables (QOL total and subscales). It followed that results of the testing of Main Hypothesis 2, regarding the associations for each of the patient/illness/social/family caregiver factors in relation to the distress variables, was essentially the bivariate results of the direct effects of patient age on patient distress.   To test Main Hypothesis 3 on the indirect effects of the factors of patient, illness-related, social/family and family caregivers on QOL, three regression models were run. The selection of variables in the models was informed by the results of Hypothesis 1, where the associations among patient age, patient distress and patient QOL were assessed. Since family caregiver distress was not found to be associated with QOL total and its subscales, family caregiver distress was not tested for its mediation role in the study. The three regression models conducted involved the dependent variables (QOL total, social well-being and emotional well-being) and their respective associations with patient age, controlling for the distress variables as well as the factors found to be associated with the dependent variables. The conclusions on the mediated effects of patient age through patient distress on QOL in the post hoc analyses were similar to those from the full sample. The effects of patient age on QOL total were completely mediated by patient distress, while the effects of patient age on patient social well-being and emotional well-being were partially mediated by patient distress.  82  Table 9: Main Hypothesis 1: Summary of multiple linear regression results (n=38) Models: Dependent variable  Combined Sample   (n=38) Independent variables B  β  Model 1a: QOL total   Patient distress -2.99** -0.81** Caregiver distress  0.02 0.004 Patient age  0.11 0.08 R2 Adjusted R2 (.71)** (.69)** Model 1b:Physical well-being    Patient distress -0.81** -0.64** Caregiver distress  0.16  0.09 Stage I cancer   1.57  0.12 Stage III cancer R2 Adjusted R2  1.09 (.53)** (.47)**  0.10  2nd Physical well-being regression: Physical well-being                                      Patient distress -0.796** -0.696** Caregiver distress  0.165  0.098 Caregiver with advanced English R2 Adjusted R2 -3.522* (.58)** (.54)** -0.253* Model 1c: Social well-being                   Patient distress -0.47** -0.51** Caregiver distress -0.04 -0.03 Patient age  0.04  0.10 Patient with high school education -2.57 -0.27 Patient with university or advanced degree R2 Adjusted R2 -0.73 (.43)** (.34)** -0.07 Model 1d: Emotional well-being Patient distress -0.76** -0.65** Caregiver distress  -0.03 -0.01 Patient age  0.18**  0.37** Years patient lived in Canada R2 Adjusted R2 -0.07* (.67)** (.63)** -0.26* 2nd Emotional well-being regression: Emotional well-being Patient distress -0.78** -0.661** Caregiver distress  0.088 0.051 Patient age R2 Adjusted R2 0.126* (.62)** (.59)** 0.258* Model 1e: Functional well-being  Patient distress -0.83** -0.71** Caregiver distress R2 Adjusted R2 -0.16 (.53)** (.50)** -0.09 * p < .05, **p < .01 83  4.1.5 Summary of quantitative analyses Results of post-hoc analyses using the sample of 38 dyads were consistent with the key findings from the full sample of n = 57, where multiple imputation was conducted. The results highlighted the positive association between patient age and QOL total, as well as social well-being and emotional well-being of patients. Further, patient age was also negatively related to patient distress, where older patients reported lower distress. The family caregiver distress variable remained statistically insignificant, in relation to QOL total and subscales, and to patient distress. While the effects of patient age on QOL total were completely mediated by patient distress, the effects of patient age on social well-being and emotional well-being were partially mediated by patient distress. Patient and family caregiver distress In the current study, patient distress has emerged as the key factor that influences patient QOL, where greater distress reduces patient QOL. Family caregiver distress was not significantly related to patient QOL total nor to any the subscales of QOL. Notably, patient distress impacts QOL differently across the domains of QOL: physical, social emotional, and functional well-being. Comparing across the four subscales of QOL, patient distress appeared to have the greatest impact on Chinese-speaking patient functional well-being. In the Anglophone group, the effect of patient distress on patient physical well-being was the largest among the QOL subscales. In both the Chinese-speaking and Anglophone groups, patient distress seemed to have the least impact on patient social well-being.  Patients, illness-related, social/family, family caregiver factors When examining the direct effects of the factors on QOL total and QOL subscales, results from the multiple regression analyses suggested that patient age was a significant predictor of patient emotional well-being in the combined sample, such that older patients reported higher emotional well-being. Patient age was also a significant predictor of patient social well-being in the combined sample,  84  although the association was not significant in the individual Chinese-speaking and the Anglophone groups. Similarly, direct effect of patient level of education on patient social well-being was found, with the social well-being of participants in the Chinese-speaking group lower among those with high school education, compared with those having post-secondary education. However, education did not appear to predict social well-being in Anglophone patients. The difference observed across groups highlighted the role of patient education among the Chinese-speaking participants, as it related to their social well-being. The effects of patient age on patient QOL were mediated by patient distress, where the magnitude of this mediation differed across subscales of QOL. The effects of patient age on QOL total could be completely explained by the presence of patient distress, while the effects of patient age on patient social well-being and on patient emotional well-being were only partially explained by patient distress. No mediation effects of patient age on the QOL subscales were detected in the sub-group analyses. The mediated effects, found in the combined sample, highlighted the role of patient distress in predicting patient QOL, and more specifically, the social and emotional well-being of the patients.  85  4.2 Qualitative analysis 4.2.1 Description of interview sample Of the 46 Chinese-speaking survey participants, ten patients and six family caregivers were interviewed, including five patient-family caregiver dyads. While there were more female than male patients (10 vs. 6) among the interviewees, there was an equal number of female and male family caregiver interviewees. The patients were largely immigrants from Hong Kong (7) who were 50 years or older. Similarly, most family caregivers were also born in Hong Kong (5), although they were younger than 50 years old. Seven of the patients were diagnosed with Stage III cancer. Three patients were diagnosed with breast cancer; other cancer sites represented in the interview sample included: colon/colorectal (2), lung (1), nasopharyngeal (1), ovarian (2) and prostate (1). The family caregivers interviewed included spouses, children and relatives of the patient; four had provided care to the patient for less than one year at the time of the interview. The demographic characteristics of the interviewees are summarized in Table 10 (which follows in the midst of this discussion). The corresponding number associated with each patient and family caregiver ID represented dyads in the interviews. For most of the patients interviewed, their cancer diagnosis was not an acute incident that constituted a medical crisis. Rather, the diagnosis of cancer spanned from a few months to one year, involving a series of tests and consultations with family doctors and specialists. With one exception, the patients interviewed in the study were diagnosed with cancer for the first time. In most instances, the news of the cancer diagnosis was given by a family doctor, many of whom were Chinese-speaking. The patients were fully aware that they had cancer, having received the news of the diagnosis from their doctors. The family caregivers were also informed of the diagnosis, and some were present with the patients during the medical consultation.  86  Cross-cutting themes relating to patient distress, family caregiver distress and patient quality of life that emerged from the interviews are presented in the following section. Excerpts of the interviews and key phrases are used to illustrate the themes relevant to the study focus, namely, patient and family caregiver relationship and quality of life. Perspectives of the patients and their family caregivers are presented. Dyadic responses in relation to the emerging themes are provided, in cases where both the patient and his/her family caregiver participated in the interviews.  Table 10: Characteristics of interview participants  ID Chinese Patients (CP)  ID Chinese Caregivers (CC) CP1 Female in 50s; migrated from Hong Kong 18 years ago; has university degree; diagnosed with Stage III nasopharyngeal cancer; 1 adult child   CC1 Male in 20s; migrated from Hong Kong 18 years ago; has post-secondary education; provided care to patient (mother) for 11 months   CP2 Female in 60s; migrated from Hong Kong 25 years ago; has high school education; diagnosed with Stage III ovarian cancer;  3 adult children  CC2 Female in 60s; migrated from Hong Kong 25 years ago; has master’s degree; provided care to patient  (sister-in-law) since the first diagnosis of breast cancer 7 years ago  CP4 Male in 50s; migrated from Hong Kong 19 years ago; has high school education; diagnosed with Stage III prostate cancer; no children   CC4 Female in 40s; migrated from Hong Kong 16 years ago; has post-secondary education; provided care for patient (husband) for 12 months.   CP5 Female in 60s; migrated from China 41 years ago; has master’s degree; diagnosed with Stage III ovarian cancer; 2 adult children    -- CP6 Male in 60s; migrated from Hong Kong 40 years ago; has university degree; diagnosed with Stage III lung cancer;  4 adult children    -- CP7 Female in 60s; migrated from China 19 years ago; has university degree; diagnosed with Stage I or II breast cancer; 1 adult child.   --  87  ID Chinese Patients (CP)  ID Chinese Caregivers (CC)  CP9 Male in 60s; migrated from Hong Kong 15 years ago; has high school education; diagnosed with stage III colorectal cancer; 1 adult child    --  --  CC10 Male in 30s; migrated from Hong Kong 19 years ago; has university degree; provided care for patient (mother) for 9 months.   CP15  Female in 40s; migrated from Hong Kong  8 years ago;  has high school education; diagnosed with Stage I or II breast cancer;  2 young  children      CC15  Male in 40s; migrated from Hong Kong 8 years ago; has master’s degree; provided care for patient (wife) for 7 months. CP18 Female in 50s; migrated from Hong Kong 39 years ago; has post-secondary education; diagnosed with Stage III colon cancer; 3 adult children   CC18 Female in 20s; born and raised in Canada; has university degree; provided care for patient (mother) for 10 months. CP21 Female in 50s; migrated from Taiwan 18 years ago; has university degree; diagnosed with Stage I breast cancer;  2 adult children.     --  88 4.2.2 Cultural understanding of cancer and disclosure of diagnosis Patients and family caregivers observed that there were misconceptions about cancer held among Chinese people. The lack of understanding of the disease also contributed to tensions in disclosing the diagnosis in the family and in the community. At the same time, patients and family caregivers recognized the informational and emotional support gained, especially from cancer survivors, when the cancer diagnosis was disclosed to selected individuals.  4.2.2.1 Misconceptions about cancer In response to a question on the understanding of cancer in the Chinese community and in the broader Canadian society, some patients and family caregivers juxtaposed ‘Chinese’ and ‘Canadian’ experiences of cancer, although a few family caregivers said they had not observed differences between the Chinese community and Canadian society more broadly in their understanding of cancer. One female patient (CP2) who had volunteered in the Cancer Centre and participated in an international conference on cancer, described Canadian society as “more open” and “not discriminatory” towards people with cancer or with special needs. She further commented that Chinese people tended to be less willing to be involved in affairs that were deemed to have little impact on their personal lives: “I feel Chinese are less concerned with other Chinese, not just other Chinese but other matters in general…maybe to avoid trouble (maa faan).” She added that the beliefs of Chinese immigrants changed as they were integrated into Canadian society. Elaborating on the openness in talking about cancer in the Canadian society, her family caregiver (CC2) commented: “They (Canadians) would deal with it as a matter of fact.” The family caregiver continued that the misconception of cancer often arose from overwhelming hearsay circulated in the Chinese community about the disease, although she believed that the level of education influenced one’s understanding of cancer. Commenting on the challenges that Chinese patients and families encountered, a male patient (CP6) concurred that Chinese  89 were generally “afraid of trouble” (paa si), and he reported a Chinese cancer survivor friend’s passive interaction with the healthcare providers while receiving care. He reasoned that the hesitance to engage and converse, coupled with the language barrier, inhibited the ability to cope with the illness. Repeatedly, patients and family caregivers cited language barriers as the key challenge in communicating with the healthcare providers. They further said that the lack of language skills to understand the medical team during consultation prohibited patients and families from asking appropriate questions regarding treatment. One female family caregiver (CC4), who had used interpreters during the medical consultation about her husband’s treatment for prostate cancer, said that the lack of English skill was particularly challenging when seeking to understand medical terminology. During their respective interviews, the caregiver and her husband (both of whom assessed themselves to have intermediate English fluency) reported an incident when they found the interpreter to be providing inaccurate translation. The patient and family caregiver concurred that the incident caused much emotional distress to the patient, and they found that the attitude of the translator and the quality of the translation worried them. One female patient (CP15) observed that there was limited cancer-related information in the Chinese language that was specific to the Canadian healthcare context, and that the Chinese-speaking community had to rely on information that originated from other Asian countries, such as Hong Kong. Patients further emphasized that limitations in English skill inhibited their access to knowledge about cancer; fewer language appropriate resources on cancer care in turn heightened their fear and anxiety surrounding a cancer diagnosis. Participants were largely in agreement that there has been increasing understanding and acceptance of cancer among the Chinese population, even though the notion of cancer as a death sentence was still held by some in the community. Patients and family caregivers noted that the association of cancer with being terminal and with death in the Chinese community often led to  90 significant fear and anxiety following a cancer diagnosis. Speaking about understandings of cancer in the Chinese community, one male patient (CP4) commented: “Cancer is very horrific, a lot of people feel that you have contracted terminal illness…ever since I have the illness, I have been in contact with support group and have more information about it. Actually, it can be cured. Then, I start feeling less fearful.” In comparison, patients and family caregivers observed that there appeared to be more understanding about cancer as a chronic illness in the broader Canadian society. Speaking of her initial fear and subsequent understanding of cancer, a female patient (CP15) said: “With breast cancer, to the Chinese, you do not want to let people know. Like, it’s very, hmm, it’s not so good. To the Chinese, there will be this feeling. Like me, earlier on, I also have this feeling. But I have overcome it. I will tell people, if it’s breast cancer, then it’s breast cancer (so be it).” She further said that since reading more about cancer, she had realized that the disease is regarded as a chronic illness, adding that this was very important and supportive information to her. In a separate interview with her husband (CC15), he concurred with the view that fear and the assumption that a diagnosis is a death sentence were common reactions in the Chinese community. He added that most Chinese people do not generally look up information to learn more about cancer, and are therefore “overly nervous” because of uncertainty and their lack of knowledge about the disease. A female patient (CP5) noted that older Chinese-speaking patients with less education may be more isolated, particularly when they were less connected with their English-speaking children who were raised in Canada. Some of the patients and family caregivers also observed that Chinese people were more reluctant to talk about the disease, and that Canadians may have less extreme reactions to the disease than the Chinese patients and families who viewed cancer as terminal. One female patient (CP21) said her friends had advised her not to tell others that she has breast cancer, for fear that she might be “looked down on” (belittled) and labelled as “damaged goods.” Another female patient (CP7)  91 commented that some of her Chinese friends referred to cancer as “that thing,” to avoid talking about the disease. She recalled a Chinese friend advising her “not to share food” in case of infecting others; yet another friend said to the patient it was “embarrassing” to talk about her diagnosis of breast cancer. The patient said that she was “surprised” by her friend’s comments, adding that she felt “awkward” and was hesitant to disclose her cancer site, especially among males in her social network. A male patient (CP6) also said that he had to explain to his friends that his illness was not contagious, and to calm their anxiety when they learned of his cancer diagnosis, although he noted that some friends visited him less frequently since he had disclosed his diagnosis. This patient further said that “there is no point talking” about his cancer, if the talking did not help him in his situation. 4.2.2.2 Tensions in disclosure Patients also spoke of disclosing information about the illness selectively and with discretion, in order to manage the illness and to regain a sense of control following the cancer diagnosis. They explained that the energy spent in attending to undue questioning about their diagnosis from others aggravated their distress. Further, several patients highlighted the importance of disclosing only when they had more clarity about their illness, so as to avoid creating anxiety for the more vulnerable members in the family. These family members included elderly parents, loved ones with pre-existing health concerns and younger children in the family. A female patient (CP5) said she did not want to disclose her cancer diagnosis to her elderly mother (in her 90s) who was living with her, until she had more information about her treatment plan, adding that she noted others were more reactionary to the news of her cancer diagnosis than herself. Similarly, a male patient (CP9) said that his siblings agreed with his decision not to tell their elderly mother about the cancer diagnosis. The patient also said he was careful not to disclose his illness to his friends, for fear his mother would learn of his diagnosis through his friends.    92 Male and female patients alike said they were concerned about the adverse impact of the news on their loved ones, which, in turn, added to their distress. As a female patient (CP7) recalled the events around disclosure of her diagnosis to her son and his family over the telephone, she broke down. When she regained her composure, she said: “I don’t want to upset them,” by telling the family her diagnosis of cancer. She further said that she knew that her son was “trying to stay cool,” while his wife was “crying deeply, the moment she picked up the phone.” The patient reiterated: “That’s why I don’t want to tell them.” In response to her family’s distress, another female patient (CP21) said she tried to respond calmly when she was with her daughter, explaining that she did not want her daughter to be worried, acknowledging that her daughter’s distress would in turn cause anxiety for her.   Family caregivers also spoke of disclosing only to selected individuals, to shield the patients from further anxiety arising from overwhelming queries from concerned relatives and friends. At the same time, some family caregivers described experiencing stress from the repeated queries of concerned individuals. One female family caregiver (CC18) said that she did not reach out to tell other people about her mother’s diagnosis of colon cancer, as she had observed the anxiety and helplessness others felt at the news of the cancer diagnosis, and that their repeated questioning of her mother’s condition had created much stress for the family. The family caregiver said that she got progressively more tired and annoyed with people who gave unsolicited “invasive and insensitive advice on remedies for cancer,” which she believed has made her mother more anxious, adding “I am annoyed because I want to protect my mom.” 4.2.2.3 Support through disclosure  While there were tensions around disclosure of cancer diagnosis, patients and family caregivers did selectively choose to disclose the diagnosis to non-family members for the purpose of rallying support for the family in coping with the illness. Disclosure of the diagnosis to friends was largely at the  93 discretion of the patients, with patients and their families choosing to disclose to individuals who could provide support to them. Where patients sought to reach out to their friends for support through disclosure, they noted that conversations with their friends helped them regain stability in their moods. One male family caregiver (CC1) said that he left it to his mother to tell their relatives of her cancer diagnosis, as per her preference. He, however, took the initiative to share the news with his friends who were concerned for him. The family caregiver added that he felt supported, having heard the experience of his friends whose parents were also diagnosed with cancer. Another male family caregiver (CC15) said that he was selective about the friends he told about his wife’s diagnosis of breast cancer, where he felt these individuals would be of support to his family. He added that his friends introduced him and his wife to other cancer survivors, which he found to be most helpful. A male caregiver (CC10) also said that he realised the prevalence of breast cancer and was assured that the disease is treatable, as he talked to his friends about his mother’s illness. The caregiver in turn comforted his mother that “breast cancer is not a big thing” and that it had “touched many lives,” while updating her of the relatives and friends who were also living with cancer. Some patients said that peer support from cancer survivors was particularly helpful, providing information to prepare them psychologically and emotionally for cancer treatment. One female patient (CP15) said that in the earlier period of her diagnosis, she constantly sought information over the Internet about treatment and diet for cancer patients. She added that the information was overwhelming and “too distressing.” She decided to stop seeking information online but rather consulted with friends who were cancer survivors. The support from cancer survivors was regarded as beneficial by both male and female patients, as well as by family caregivers. Some of the patients and family caregivers highlighted that the connection with cancer survivors for mutual support helped to alleviate patients’ fear and sense of isolation. Speaking of the strong support she received from friends who were cancer  94 survivors, a female patient (CP15) said that these friends helped to prepare her for what she could expect throughout her illness trajectory, adding, “I’ll be lost without them.” Her husband (CC15) also emphasized that support from cancer survivors was particularly helpful to his wife, commenting that “they (cancer survivors) have experienced it and know it.”  Another male patient (CP4) said that hearing the positive outcomes of other cancer patients in the support group he attended reduced his fear of cancer, as he gained greater confidence that his disease could be cured. A female patient (CP2), who had attended English and Chinese-speaking cancer support groups, said that it would be important for Chinese cancer patients to connect with other cancer patients, learning “to walk the journey of cancer survivors and assume the new normal.” However, her family caregiver (CC2) expressed concern that the sharing in the support groups might impact her sister-in-law (patient) negatively, when the patient learned that her cancer may recur. Despite her concern, the caregiver added that she believed her sister-in-law would want to know the nature of her illness and could face the fact of possible cancer recurrence. 4.2.3 Distress and emotional regulation in the family Patients and family caregivers spoke of their distress following the diagnosis of cancer, where worry about one another was cited by both the ill person and care provider in the family. While coping with their distress, patients and family caregivers also sought to regulate their emotion, with members of the family seeking to downplay their distress in order to protect the patient from distress. At the same time, patients also sought to conceal their distress from their family, particularly their children, to shield them from distress.   4.2.3.1 Sources of patient distress Among the patients interviewed, common reactions to cancer diagnosis were “disbelief,” “confusion,” “shock,” “fear,” “depression” and “worry about (their) families,” particularly their children  95 and elderly parents. Some patients said they were shocked with their diagnosis, adding that they had been conscientious in maintaining a healthy lifestyle, and did not anticipate that they had cancer. Despite the initial shock, patients spoke of their subsequent acceptance of their illness. In describing their distress relating to cancer treatment, patients reiterated that the major side effects of cancer treatment they encountered were: fatigue, pain, numbness, loss of taste and appetite (enjoyment of food). Some female patients particularly emphasized the pain of cancer and side effects of treatment as “suffering” (tung fu) that extended beyond physical distress, while other patients reported relatively less disruption arising from the effects of cancer treatment. On adjusting to cancer treatment, patients spoke of restricted mobility coupled with memory loss and reduced energy, as major changes they underwent.  In addition to living with the treatment side effects, patients spoke of their reduced independence and increased reliance on others while undergoing cancer treatment. Male and female patients alike mentioned their desire to be self-reliant and “not wanting to burden friends and family members” in caring for them. As a female patient (CP18) spoke of her reliance on family and friends for rides to medical appointments and help in dressing her wounds after surgery, she expressed her concern that she had “troubled “ (imposed on) others. Her daughter, who provided care for her, also noted the tension the patient experienced as a result of the ways her independence had been compromised following her cancer treatment. The family caregiver (CC18) reported occasions when she found her mother crying when she was unable to complete simple day-to-day tasks on her own, adding that her mother felt “frustrated and overwhelmed.” The family caregiver further said:  “It’s very hard on Mom. She’s always very independent. She doesn’t want to bother other people and felt bad asking for help.”  Several female patients also commented on their constraints in caring for their children while coping with their illness. Throughout the illness trajectory, patients expressed concern about the impact of their illness on their children, including adult children. One female patient (CP2) said that “the cancer  96 is not just about me,” as she spoke of feeling remorseful and worried about her daughter’s diagnosis of ovarian cancer, and concern for her granddaughters, who were found to also carry the genotype related to the patient’s cancer. She further explained that her concern for her family differed from the distress of having cancer, even though she felt her children were very understanding and she did not feel blamed. Her sister-in-law caring for her (CC2) also noted the patient’s distress about her family, commenting that the patient felt responsible for her daughter’s cancer, but was comforted that her daughter’s cancer was in remission.  4.2.3.2 Sources of family caregiver distress In addition to shock and worry, female and male family caregivers alike reported strong emotions of “fear,” “numbness,” “anger,” “sadness,” and “helplessness.” Among the family caregivers interviewed, some commented that family members were more worried than the patients following the cancer diagnosis. In response to the news of the cancer diagnosis, one female family caregiver (CC4) caring for her husband diagnosed with prostate cancer said this of herself: “Very unhappy. I am more unhappy than him (patient).” Later in the interview, she described her distress when a post-surgery acute incident prompted the patient to be admitted to the emergency unit: “It was the most difficult…I felt isolated (gu duk) and helpless (mou zo)…I have to face things alone,” as she spoke about being away from her family in Hong Kong and that she and her husband did not have family support in Vancouver. Another female family caregiver (CC18) observed that her father appeared to be more distressed than her mother, who was diagnosed with colon cancer: “Dad was not eating…he looks more sick than Mom.” She continued to describe her mother as “optimistic and strong,” and that the patient would console and encourage friends and relatives who were distressed by her cancer diagnosis.  While the family sought to provide comfort to the patient, family caregivers said they recognized their limitations in relieving the patient's distress from cancer treatment. In describing her concern for  97 her sister-in-law’s diagnosis with ovarian cancer, one female family caregiver (CC2) spoke of her helplessness in relieving the patient’s distress from chemotherapy, however much she wanted to help. The family caregiver spoke of the side effects of the treatment on the patient, noting the impact to be intense while describing her stance as “willing but unable to help” (oi mok nang zo). This sense of helplessness in the family was also observed and highlighted by the patient herself (CP2), who commented that she knew her family would not be able to bear her suffering, and she would need to “clear her own hurdle” (gwo zi gei yi jat gwaan).   In addition to their sense of helplessness, most family interviewees – including men and women – spoke of the physical and emotional weight of caregiving. One male caregiver (CC10) caring for his mother said he had to coordinate the family’s schedule and spent time commuting between his home and his parents’ home over the past year of caregiving; this resulted in little time with his wife and his newborn child. Another male family caregiver (CC1), providing care to his mother diagnosed with nasopharyngeal cancer, described his caregiving experience as “pretty tough, emotionally,” adding that his mother needed much emotional support. Being the only child and the caregiver in the family, he spoke of feeling lethargic and stressed over the year of caregiving. He also said that he has “cut off (his) social life completely” and has “lost some friends,” while taking time off from work to care for his mother. At the same time, the caregiver said he was caring for his ill father for close to half a year; in her interview, his mother (CP1) disclosed that her former husband was hospitalized for relapse of bipolar depression at the time of her diagnosis. The family caregiver said that the caregiving had had major impact on him at a critical period of his life, and that he was “still recovering” from the loss of business networking contacts and his productivity. The patient (CP1) also observed the distress in her son, adding that she worried whether he could “get through the crisis” of coping with illness in the family. A female family caregiver (CC18), caring for her mother diagnosed with colon cancer, concurred with the  98 challenges in being the primary caregiver; she sought to balance caregiving and her career, staying home to care for her mother while her father was away on business trips. Reflecting on the impact of caregiving, the caregiver said: “It is hard, putting life on hold…but I feel guilty to even think that way.” She added she had lost weight and had difficulty with her sleep, and others had noted her weariness. Family caregivers’ anxiety for the patient’s recovery was also highlighted in the interviews. In caring for the family members diagnosed with cancer, some family caregivers spoke of the distress of the patient and the family caregiver that stemmed from their conflicting expectations regarding treatment and self-care. A male family caregiver (CC1) caring for his mother diagnosed with nasopharyngeal cancer characterised his mother as “pessimistic,” and also said, “She believes she would die and did not believe in the treatment.” He added that she “has not taken care of herself” when she had not followed his advice on managing her symptoms after chemotherapy, which subsequently resulted in her admission to the emergency unit. When interviewed, the patient (CP1) said that her son (CC1) had perceived her to be “negative” when she chose to do what was most comfortable for her and she had not followed all his advice during treatment. 4.2.3.3 Emotional regulation between patient and family caregiver Even as family caregivers in the study commented on their distress with their loved one’s diagnosis of cancer, some sought to conceal their emotions from the patients. In response to her mother’s diagnosis of colon cancer, a female family caregiver (CC18) said: “I almost didn’t know how to respond…I would show my concern and anxiety to everyone else but her (mother).” The caregiver further described herself to be “playing it up” and “trying to act” when she was with her mother. At the same time, the family caregiver also described the times she had “broke down in uncontrollable crying” when she was with her close friends, adding that she did not realise the “emotions that were bottling inside” her. Speaking of the family’s reaction to her cancer diagnosis, the family caregiver’s mother  99 (CP18), noted her daughter’s and other family members’ attempt to downplay their distress, in their interaction with her. The patient observed that her family (husband, daughter, daughter’s fiancé) looked startled when the doctor disclosed the news of her cancer diagnosis, but they did not show much emotion in her presence, following the initial disclosure at the doctor’s office. The patient further commented that she was aware that family members were attempting to put up a strong front in her presence, when she observed that they were “overly calm” in response to her illness.  While family members sought to conceal their emotions from the patients, some patients also spoke of hiding their distress when they were with family members. One male patient (CP6) said that when his daughters visited him at the hospital, he “pretended” he had no pain, to assure and “cheer them up.” He further said he did not reveal much of his negative emotions, so as not to affect others, adding “I do not want others’ emotion to affect me,” and that “there is no point when everyone could not be happy.” Rather, he said that he wanted to be treated “as normal a person as possible.” To alleviate anxiety in the family, a female patient (CP15) said that she did not use the word “cancer” when speaking with her 10-year old daughter, having observed her daughter’s distressing response to the side effects of treatment on the patient. The patient also said that she always smiled in her daughter’s presence, seeking to “console and cheer her up.”  4.2.4 Approaches to family caregiving and quality of life  Patients reiterated the pivotal role that their family caregivers played in giving them informational, practical and emotional support. Notwithstanding the distress that patients and family caregivers encountered following a cancer diagnosis, they agreed that there were positives of cancer in the family that contributed to their quality of life. Patients spoke of personal growth while coping with their illness, and family caregivers also noted similar growth in themselves through the caregiving experience.  100 4.2.4.1 The “right” family support  All patients interviewed spoke of the concern their families displayed towards their diagnosis and the support they rendered to uphold the patient’s quality of life. Female patients, in particular, reiterated the pivotal role that their family played in supporting them throughout the cancer trajectory. The support that patients described included informational, practical and emotional help from their family caregivers. One female patient (CP18) spoke of the informational support from her daughter (family caregiver), who was a dietitian: “She (daughter) understood the patient’s needs, and did not overwhelm me with too much care.” The patient added that her daughter’s knowledge of drugs and ability to answer her questions had been especially helpful, providing her with appropriate care. In describing the practical and emotional support she received from her sister-in-law caring for her during her illness, a female patient (CP7) said: “I was taken by surprise with her (sister-in-law) openness to help without thinking twice…her sincerity and willingness to make herself available to help gave me much comfort.” The patient said her sister-in-law had given her the “right support and help,” not overwhelming her with questions but referring her to other cancer survivors to provide advice on managing the disease. The patient said she was particularly appreciative of her sister-in-law's attitude of “not digging into too much (information)” but respecting her and giving her space. Often, family members accompanied patients to medical appointments. These family members included spouses, children and sisters-in-laws, most of whom were deemed proficient in English and could facilitate patient communication with healthcare professionals. In addition to accompanying patients to appointments, the practical help that patients highlighted also involved preparing soup/food according to their dietary needs, running errands, and attending to patients’ personal care (like dressing, especially after surgery). Although patients often cited practical help when describing the support they received following their cancer diagnosis, some female patients also highlighted the role that emotional  101 support played in their coping. In describing her sister-in-law's caregiving, one female patient (CP2) commented, “she (sister-in-law) understands,” and explained the importance of being understood without being burdened with excessive questions, particularly in relation to her pain after cancer treatment. The patient said that her sister-in-law's approach to caregiving had in turn strengthened her ability to cope with the distress of the side effects of treatment. Her family caregiver (CC2) also described the intensity of the side effects from chemotherapy that she witnessed in the patient, adding that she felt helpless and sought to provide practical help to relieve the patient’s distress. The family caregiver observed that when the family was overly protective of the patient, the family caregivers imposed their expectation on the patient to comply with the family’s wishes rather than respecting the patient's autonomy in managing the illness. Further, she emphasized that although the family’s demands stemmed from “good intentions” (hoe yi), it was important for the family to also accept the cancer diagnosis, as much as the patient had accepted the illness. 4.2.4.2 Positives of cancer in the family Patients and family caregivers agreed that the cancer diagnosis had shifted the dynamics of the social networks of the patients and family caregivers. The majority of the patients and family caregivers said that the family had become closer; they had also drawn closer to friends who provided support at the time of the cancer diagnosis, particularly among friends who were cancer survivors. A male patient (CP9) said communication with his daughter and siblings had increased since his cancer diagnosis, and that his daughter had shown more concern for him, and contributed more to household chores. While speaking of the couple relationship, one female patient (CP15) broke down as she commented, “we were unable to swallow when eating together,” reflecting the emotion and burden the couple shared. As she regained composure, she said that her husband had been her greatest source of help, and that she would not have been able to cope without his support. Her husband (CC15) also commented on the couple’s  102 coping with cancer, noting that he paid particular attention in caring for the patient, while taking care of their two children and working full-time.    Several patients mentioned an increase in their appreciation for their family and friends who cared for them, which in turn expanded their vision to support other cancer patients. One female patient (CP2) said she had entrusted her life to God and saw pain and suffering as a gift that she could use to help other people. She further said that life is very “lovable” (amazing), as she experienced sincere care from people, which had enlarged her capacity to better appreciate her family’s and friends’ love for her. Another female patient (CP21) commented that the “good thing” from her cancer diagnosis was her increased ability to express herself to others, even though she experienced tension between the freedom to speak up and her fear of facing cancer. She added that her illness was a “privilege,” allowing her to understand other cancer patients and to help them “get through the cancer journey.” At the same time, some family caregivers described the positive impact of caregiving on them, in the same way that female patients spoke of the positive of cancer they experienced.  Among the male and female family caregivers who spoke of the positive change from their experience of caring for their ill family members, one male family caregiver (CC1) said he had become more inquisitive about cancer since his mother’s diagnosis with nasopharyngeal cancer, and took the initiative to talk to his friends about the disease. In caring for his wife diagnosed with breast cancer, a male family caregiver (CC15) said: “I have learned how to care (for the patient) with patience. The most important is to be patient, and also patiently take care of the children, so that they know the family is still complete and beautiful (yuan mei).” Another female family caregiver (CC2) caring for her sister-in-law diagnosed with ovarian cancer noted: “I take upon myself to help. I do not think it is a sacrifice…I feel that it is logical and natural (lei saw dong yin) (to help the patient). She has this difficulty, and I can share a bit.” The caregiver further said, “My attitude has changed, not to be distraught, but to do what needs to be done. My concentration  103 is not on how sad it is, but on what I should do to cope with this change.” The family caregiver continued that her caregiving experience has helped her to understand her responses better, increasing her awareness of her behaviour and thoughts. Summary Summarizing the study findings, patient distress emerged as the key factor in explaining patient QOL. The effects of patient age on patient QOL, and particularly in patient social and emotional well-being, were mediated by patient distress, such that lower distress in older patients explained better social and emotional functioning. A key theme that surfaced from the qualitative data analysis was the emotional regulation occurring in the family when coping with cancer. The negative impact of family caregiver distress on patient emotional well-being was acknowledged by both the Chinese-speaking patients and the family caregivers interviewed, although statistical significance was not found between family caregiver distress and any of the variables related to the patient in the quantitative analyses. In the interviews, younger and older patients alike indicated anxiety and worries about the impact of their cancer diagnosis on their families. The following chapter provides an interpretation of the study findings, integrating the results from the quantitative and the qualitative analyses, discussing the key findings in light of the prior knowledge of the research topic gleaned from the literature reviewed in the dissertation.   104 Chapter 5: Discussion The diagnosis of cancer evokes multiple responses, some of which are shared experiences among patients and families, regardless of their ethnic and linguistic background. Other responses stem from the socio-cultural context of the cancer experience, where cultural values and norms shape the Chinese-speaking patients’ and families’ perceptions and expressions of their illness experience. For the patients and family caregivers interviewed in this study, their integration in the broader Canadian society may have shaped their cancer experience, as reflected in their responses to the issues they encountered following a cancer diagnosis in the family.  Revisiting the conceptual framework used in the study to explore factors influencing patient QOL, the study provided evidence to support the mediation role of patient distress hypothesized in the framework. Patient distress emerges as the key factor in explaining patient QOL. While the effects of patient age on patient QOL total were found to be mediated by patient distress in the combined sample and in the Anglophone group, mediation effects in the QOL subscales were only observed in the combined sample. The variation in the impact of distress across the QOL subscales found in the study highlighted the importance to delineate patient QOL according to the domains of physical, social, emotional and functional, within the conceptual framework. Contrary to the hypotheses proposed in the study framework, family caregiver distress was not found to be significantly related to patient distress nor with patient QOL in the study. Despite the lack of statistical evidence, the conceptual importance of family caregiver distress as it relates to patient distress and QOL supports the inclusion of the family distress variable presented in the study framework. From the interviews, the Chinese-speaking patients and their family caregivers clarified the sources of their distress and the impact on patient QOL. Age was not articulated as a factor influencing these patients’ and their family caregivers’ distress and QOL.  105 Younger and older patient interviewees alike indicated anxiety and worries about the impact of their cancer diagnosis on the family.     A key theme that surfaces from the qualitative data analysis is the emotional regulation occurring in the family when coping with the cancer diagnosis. The negative impact of family caregiver distress on patient emotional well-being was acknowledged by both the patients and the family caregivers interviewed, highlighting the role of family caregiver distress in patient QOL. In addition to the findings drawn from both the quantitative and qualitative analyses pertaining to the associations among the variables depicted in the conceptual framework, the interviews also provide insights to what the Chinese-speaking patients and family caregivers considered to be the right family support. The positives of cancer as experienced by the patients as well as the family caregivers were also articulated in the interviews, providing further understanding to the factors that influenced patient QOL.                   5.1 Impact of patient distress on patient QOL The diagnosis of cancer evokes an array of emotions and concerns across the domains of day-to-day living. From the quantitative analyses, a higher distress level in the patients was found to be associated with lower QOL regardless of the language group of the patient participants. This finding is congruent with literature that examines the adverse impact of psychological distress on health-related outcomes and QOL (Kim et al., 2008; Mehnert & Koch, 2008). Further, no group difference was found regarding the impact of distress on the patient QOL. Regardless of the cultural and linguistic background of patients, cancer may be understood as a serious illness that challenges the order of the individual’s world, spanning the domains of the person’s QOL. To better understand the disruption that cancer brings to a person’s life and the implications of living with cancer, it is important to clarify the nature of the patient’s distress and the process through which it affects the patient’s QOL.    106 5.1.1 Sources of patient distress Responses to the items in the Kessler distress scale quantify their “feelings during the past 30 days.” To understand the nature of the distress, the interviews with the Chinese-speaking patients provide clarification regarding the sources of their distress. In several accounts, patients described their distress as stemming from the side effects of their cancer treatment. Such distress relates to feelings of fatigue, pain, numbness, loss of taste and appetite, loss of mobility and memory loss. Family caregivers interviewed also cited similar sources of distress observed in their ill family members. Regardless of gender, patients spoke of the distress following cancer treatment that compromised their physical and functional well-being.  The impact of cancer on the well-being of patients has been well documented in studies examining the distress following cancer diagnosis and cancer treatment, where distress is said to be multidimensional and includes the physical, emotional/ psychological, social, existential and spiritual domains of quality of life (Meyerowitz & Oh, 2009; Miller, 2010). While side effects of treatment are noted to physically affect the patients and their daily functioning, it is conceivable that the experiences of such distress will also have adverse effects on the emotional and social well-being of the ill individuals (Siegel et al., 2012). Particularly, fatigue and pain have been found to be associated with depressed mood and anxiety (Brem & Kumar, 2011; Byar, Berger, Bakken, & Cetak, 2006). The frustration of losing their capacity to manage their day-to-day activity and being dependent on others while in treatment were reported to be the “most difficult” aspects in patients’ experience of the illness. In a previous study on the cancer experience of Chinese-speaking patients, the practice of self-reliance was highlighted when the patients spoke about issues relating to coping and recovery (Bell et al., 2010). With the diagnosis of cancer and the onset of distress associated with the illness, these patients may be  107 facing additional tensions when seeking to maintain the intricate balance of being independent within a collective community, further undermining the individual’s emotional well-being.     The concern patients have about the impact of cancer on their children was highlighted by the patient interviewees. Several of the female patients broke down when they spoke of the effects and implications of the cancer diagnosis on their loved ones, including adult children and grandchildren. The worry about their children also included working adult children, who were themselves parents. Distress relating to family among the Chinese-speaking cancer populations has been examined in other studies which investigated the impact of cancer on family relationships (J. Lee & Bell, 2011; Loke et al., 2003). In families that adhere to the cultural norm where parenting responsibilities remain throughout the family life span, the burden of coping with a serious illness may be further exacerbated by the patient's distress due to the loss of capacity to continue the role as caregiver of the family.    5.1.2 Variation in the negative effects of patient distress across QOL domains Variation in the negative effects of patient distress on patient QOL was observed across the domains of the FACT-G scale. While QOL reflects the overall well-being of the patient, it is important to note that QOL is a multi-dimensional construct; the breadth in QOL definition includes subjective evaluations (such as satisfaction), perceived health status, or both evaluation and health status (Ferrans, 2005; Ferrell et al., 1995). The FACT-G scale used in this study includes the subscales tapping four major dimensions of health-related QOL: physical, functional, emotional and social (Cella & Tulsky, 1993). Understandably, each domain of QOL may vary in importance and priority in an individual’s life, according to individual context. Further, an individual’s weighting of each domain may also vary across a person’s life span, where one domain of QOL may be of greater or lesser concern in comparison to another domain over time.   108 The patterns in which patient distress was related to each of the QOL subscales also differed across the language groups of patient participants, although the group difference was not statistically significant. Conceivably, the cancer patient’s QOL may also differ according to cultural considerations, including views of cancer etiology and family structure, even as QOL may be defined differently in different cultures (Gotay, 2004). The lack of statistical significance in difference between the Chinese-speaking and Anglophone groups may be attributed to the small sample sizes, or that the individual responses between the two groups of cancer patients in the study are indeed congruent. Notwithstanding the lack of statistical evidence, it would be prudent to also examine how distress is described in relation to the patient’s domains of QOL among the interviewees in the study. This line of query would be relevant and important in a culturally diverse cancer population in Canada. The multiple regression results indicate that patient distress has the greatest impact on the Chinese-speaking patient’s functional well-being, after controlling for the associated covariates. The distress of no longer being able to “rely on oneself” (kaau zi gei 靠自己) but having to “trouble” (maa faan 麻煩) others following cancer treatment may explain the strong association between patient distress and patient functional well-being in the Chinese-speaking group, when the ill persons could not maintain their day-to-day functioning without the help of others. When speaking of patient distress following cancer treatment, the patient's sense of helplessness and frustration over reduced ability to be self-sufficient in managing day-to-day life was articulated by Chinese-speaking patients as well as family caregivers. The value of self-sufficiency, held by the Chinese-speaking cancer patients, may illuminate the importance ascribed to deriving satisfaction from one’s ability to function and manage daily activities. This help-seeking behaviour of individuals may parallel the observations made of families with cultural values where the admission of having a problem is questioned and coping with illness is a private matter (Chan et al., 2004; Rolland, 2005). For Chinese-speaking patients, the need to seek help  109 from others in managing their daily life may be undermining the fulfilment they could derive when help was rendered, when the tension between needing help and wanting to be self-sufficient arises. It follows that functional well-being would be a critical domain when considering the quality of life of the Chinese-speaking cancer patients, where compromised self-sufficiency may lower their functional wellness.  5.1.3 Understanding the relation between cancer and social well-being   Chinese-speaking patients with high school education have reported lower social well-being, compared to those with more education in the current study. The role of socio-economic status in the context of cancer has been well described in the oncological literature, where the systematic variations in health outcomes across the range of income, ethnicity and education contribute to the adverse consequences of cancer (Ashing-Giwa & Lim, 2011; Bowen et al., 2007). Related to education is proficiency in English among the Chinese-speaking cancer population seeking to navigate the Canadian health care system in a foreign language. While the association between English fluency and patient QOL in the Chinese-speaking group was not found to be significant in the quantitative analyses, the patients and family caregivers interviewed in the study reiterated the importance of language in their coping with the illness. Specifically, the lack of English language skills was said to be a barrier in communication with the healthcare professionals by patients and family caregivers, along with related limited access to language-appropriate resources. To help them manage the disease, patients and family caregivers consulted with Chinese language resources from the cancer care network based in Asia. The use of resources from the non-Canadian healthcare context was also seen as potentially problematic by many of the patients and family caregivers interviewed, as the information and approach to cancer care is contextualized to the environment where such resources originated. An example of this is the use of traditional Chinese medicine and food practice that is upheld as complementary to the Western bio- 110 medical approach to cancer care in Asia. However, such forms of medicine may not be supported in the North American system of care and by the practice of the healthcare professionals, resulting in tension for the Chinese-speaking patients and their families living in a Western society, seeking to integrate their cultural practice of healing with the mainstream prescription of cancer treatment. While the lack of English skills among patients with lower education may inhibit their quest for knowledge about cancer, the constraints in communicating with healthcare professionals and accessing appropriate resources could also limit the understanding of cancer among the Chinese-speaking patients and family caregivers. The patients and family caregivers interviewed cited several misconceptions said to be commonly held in the Chinese community, including identifying cancer as a "death penalty" and "contagious." This finding concurs with Wong-Kim and colleagues’ study on assessing the cancer beliefs among the Chinese immigrant community in the US, where the belief of cancer as incurable and contagious was reported (Wong-Kim, Sun, & DeMattos, 2003). The misconceptions circulated in the Chinese community further heightened fear and social stigma towards a cancer diagnosis, where patients and families were discouraged from talking about the illness and the distress they encountered in the family and in their social network (Wong-Kim, Sun, Merighi, & Chow, 2005). In addition to the patients’ and family caregivers’ own reservations about disclosing the cancer diagnosis to non-family members, some patients also received advice from friends to "not tell" others about the illness, so as to avoid being evaluated. For those patients who feared being negatively regarded following their cancer diagnosis, the potential risk of being stigmatized would further hinder them from reaching out to others for help in coping with their illness. As Cheng and colleagues observed, the stigmatization effects could lead to social isolation, reducing the patient’s social network for drawing support (H. L. Cheng et al., 2013). Conceivably, the tendency to withdraw instead of reaching out to converse about the concerns around cancer will continue to isolate these individuals in their experience as cancer patients. For the  111 Chinese-speaking patients who felt disconnected with their social environment, such isolation would further hinder understanding of their illness; the misconception and social stigmas associated with cancer diagnosis could continue to distance them from their social network, compromising the patient quality of life. 5.1.4 Connection of disclosure of cancer diagnosis and social support  While the social stigma attached to cancer may have prevented patients and their family caregivers from discussing the illness outside of the family, findings from the interviews provide further insights into the issue of disclosing the cancer diagnosis in the Chinese community. Several of the interviewed patients and family caregivers spoke of their reservations in disclosing the cancer diagnosis within the family and to their social network, in order to protect each other and themselves from the negative reaction of others towards cancer. The protective stance that these patients and the family caregivers adopted is instrumental in their intent to strengthen the social support within the family, where each party assumes the role of gate-keeper to shield each other from distress and to safe-guard one another’s well-being. Among the patients who indicated reservations in disclosure during the interview, they also emphasized the importance of having some clarity regarding their condition prior to talking about their illness with others, so as to regulate the anxiety stemming from the uncertainty of their cancer diagnosis. For the patients and family caregivers who highlighted the importance of exercising discretion in disclosing the cancer diagnosis, sharing the news of the cancer diagnosis only to selected individuals is congruent with their goal to solicit support to best cope with the illness. Notwithstanding the reservation about disclosure, some patients and family caregivers spoke of the practical help and encouragement they received when they reached out to talk about the cancer diagnosis among family members and friends. Others sought to connect with the larger community of cancer survivors, who provided the patients and families with informational and emotional support while  112 undergoing cancer treatment. Particularly, the patients who spoke of their connection with cancer survivors noted the decrease in their fear of the disease and increase in their confidence in coping with the illness through such support. The impact of peer support hinges on the ability of patients to identify with cancer survivors, as patients spoke of “being understood” by those who have experienced the distress of living with cancer. At the same time, the confidence that patients spoke of in coping with their illness may also be conditioned by the positive outcomes of the cancer survivors whom they sought for encouragement. As one family caregiver said of peer support, she was concerned that her loved one might be distressed when she heard of cancer recurrence among participants in support groups. Where peer cancer survivors provide the unique benefits of emotional and informational support, the patients gain from the mutual sharing of experience; they speak of being understood and finding positive role models for coping with the illness (Bell et al., 2010; H. L. Cheng et al., 2013; Ussher, Kirsten, Butow, & Sandoval, 2006). For those who found peer support to be important, the identification with other cancer patients helped to normalize the patients’ distress, reconnecting the patients to their social environment while living with cancer. In the words of a female patient who spoke positively of peer support, Chinese patients who connected with peers could learn “to walk the journey of cancer survivors” and assume the “new normal,” adapting to life after a cancer diagnosis. 5.2 Mediating role of distress in patient QOL Within the conceptual framework of this study, several factors pertaining to the patient’s and family caregiver’s background, illness characteristics and social/family relationships were identified as having direct effects on the patient’s QOL. However, only patient age and patient education were found to be directly associated with patient QOL, although these findings varied across participant groups and across domains of QOL. The positive association between patient age and patient emotional well-being concurs with past studies where older patients reported lower emotional distress and better quality of life  113 in the mental health domain following a cancer diagnosis (Mehnert & Koch, 2008; Thomas et al., 2011). Further, Nishigaki and colleagues assert that life stages, as they relate to changes in the individual’s social and personal roles, should be considered when examining the psychological adjustment of cancer patients (Nishigaki et al., 2007).  When examining the narratives of older women diagnosed with breast or gynaecological cancer, the findings from Sinding and Wiernikowski suggest that a long life of hardship may provide a context for the assessment of cancer as non-disruptive (Sinding & Wiernikowski, 2008). The authors observe that the experience of this latest chronic illness may be rendered normal by those who had accumulated adaptation skills for significant health and social problems throughout a lifetime leading to old age. While the life experiences of older patients may have enhanced their capacity to adapt in times of distress, the roles and responsibilities of younger patients in their respective life stages may also add to their burden in managing the disease. These responsibilities included providing care to dependents and maintaining employment; the worry about their children and the anxiety of returning to work after cancer treatment were cited as concerns to the younger cancer patients interviewed. Conceivably, the impact of a life-threatening illness permeating the life domains of the younger patients, particularly younger women, may be more destabilizing to their efforts to cope with the distress of cancer (Wong-Kim & Bloom, 2005). In response to literature suggesting older people already living with illness may experience cancer as less disruptive, Sinding and Wiernikowski recommend that researchers be cognizant of the influence of demographic characteristics (including age, class, gender) in people’s construction of their chronic illness experience, where a long life characterized by sufficiency may also be associated with a less distressing assessment of cancer (Sinding & Wiernikowski, 2008). The higher social well-being reported among older patients in the study reflects their satisfaction with their social relationships, where the ill persons felt supported and connected with their family and  114 friends. While literature has suggested that larger social network size and adequate emotional support are associated with higher QOL, Cheng and colleagues’ study of Chinese cancer survivors highlights that it is patient satisfaction with social support that influences QOL, rather than the network size per se (H. L. Cheng et al., 2013). For younger patients, the roles and responsibilities (such as career and parenting) in their respective life stage may compromise their capacity to cultivate their social network. In addition to facing time constraints in building up social relationships outside of the family, the demands from work and family life may leave these individuals with less mental space to reflect on and strengthen the quality of their social relationships. It follows that in times of serious illness, gaps in social support may emerge and these younger patients may feel less satisfied with their communication and connections in their social network. Where an older patient's life experience or life stage may serve to buffer some of the negative impact of distress during a serious illness, the satisfaction with social relationships may also lessen the impact of psychological distress at times when social support is critical to well-being. Such positive evaluation of the social network may thereby protect the older patient from further anxiety following a cancer diagnosis. 5.3 Impact of family caregiver distress on patient QOL   Notwithstanding the conceptual importance of the role of the family caregiver in the patient’s coping with serious illness, the quantitative findings did not provide evidence to support the hypothesized negative association between family caregiver distress and patient QOL; nor with any of the QOL subscales. Further, no significant association was found between family caregiver distress and patient distress. There may be several reasons for the lack of evidence in the current study to support the role of family caregivers in patient QOL. Family caregivers in both the Chinese-speaking and Anglophone groups have reported lower distress, compared to the patients. This sample of less distressed family caregivers may reflect a selection bias; family caregivers who volunteered to  115 participate in the study may have in fact experienced less distress at the time of study enrolment or they may have been less inclined to report of the adverse of impact on themselves. The family caregiver sample is also much smaller than that of the patients, resulting in less power to detect an effect.    Another reason for the lack of significant findings may also be attributed to the choice of instrument used in measuring family caregiver level of distress, where the measurement of patient distress may have differed from the measurement of family caregiver distress. The Kessler distress scale (K6) is a short screening scale consisting of “a heterogeneous set of questions that define behavioral, emotional, cognitive and psychophysiological manifestations of psychological distress” (Kessler et al., 2003, p. 185) in terms of frequency over a period of “past 30 days.” In Fletcher and colleagues’ conceptual model of cancer family caregiving, the distress appraisal also involves caregiver appraisal of burden, ability, needs and future outlook (Fletcher, Miaskowski, Given, & Schumacher, 2012). Northouse and colleagues propose that when assessing family caregiver distress, it is important to consider questions that seek insight of the family caregivers’ access to resources to provide optimal care, ability to manage multiple responsibilities, engagement in self-care and capacity to keep their spirits up (Northouse et al., 2012). It follows that the assessment of family caregiver distress may require items centred on concerns of the family caregivers, to better reflect the family distress in caring for loved ones diagnosed with cancer. Notwithstanding the inclusive results regarding the role of family caregiver distress in the study, it would be prudent to have understanding of the nature of the distress that family caregivers experienced, as it relates to caregiving and the patient’s well-being.  5.3.1 Sources of family caregiver distress   In caring for loved ones undergoing cancer treatment, family caregivers articulated a sense of helplessness when the patient was suffering from the side effects of the treatment. The effects of a patient’s distress on the family caregiver has been examined in family caregiving literature. One meta- 116 analysis found a tendency for family caregivers to report more distress during the treatment phase when the demands of caregiving were high. In another study on gender differences in caregiving distress, husband caregivers were found to report higher caregiver’s esteem and lower caregiving distress, compared to wife caregivers (Kim, Loscalzo, Wellisch, & Spillers, 2006).  In cases where husbands provided care to their wife with poorer psychosocial functioning, Kim and colleagues found that husband caregivers reported greater caregiving distress, highlighting the role of caregiver gender and patient psychosocial adjustment in explaining distress in family caregiving.  While the comment of family caregivers being “more distressed” than the patients was raised by some female family members interviewed in the study, such remarks may be pointing to the family caregivers’ feelings of helplessness while seeking to comfort the patients. In addition to witnessing the distress of the patients, the family caregivers are concurrently wrestling with limitations in relieving the suffering of their loved ones, which may further exacerbate the family caregiver’s distress. Conceivably, the interrelatedness between the patient and family caregiver distress is a significant source of tension for both patient and caregiver throughout the illness trajectory, where the burden of caregiving adversely impacts the quality of care rendered (Hagedoorn et al., 2008; Milbury, Badr, Fossella, Pisters, & Carmack, 2013; Northouse et al., 2012; Segrin & Badger, 2014). For family caregivers who seek to provide care by protecting their ill loved ones from the physical and mental pain of cancer, the tension arising from feeling helpless may be more vexing, adding to the burden of caregiving (J. Lee & Bell, 2011; Liu et al., 2005).   Family caregivers also spoke of feeling isolated when assuming the role of primary care provider for the ill loved one in the family. This sense of aloneness was identified as particularly distressing for those where migration has restricted support from extended family and close friends still living in their country of origin (Heidenreich, Koo, & White, 2014). Notably, the impact of migration and the nature of  117 the distress of migrant families are key considerations in understanding an individual's illness experience and help-seeking behaviour (R. Lee et al., 2001). For immigrant families, the separation from other family members and friends, who had been a key source of support in times of critical need, may also create a sense of abandonment in patients and families seeking to navigate a health care system that requires competency in language and in self-advocacy.  Notably, the lack of English skills may be a barrier to communication with healthcare providers and access to disease information. Moreover, for families from a culture where healthcare providers are esteemed as the experts to give directives in patient care, such beliefs may be at tension with values of patient autonomy and self-advocacy promoted in the Canadian healthcare system (H. L. Cheng et al., 2013; So et al., 2013). Conceivably, the distinct health beliefs and value systems, as well as the linguistic background of immigrant care recipients that differentiate them from the host society, may put them at odds with the Canadian health system, adding to the distress in their illness experience (Marshall, Wong, Haggerty, & Levesque, 2010; Z. Wu, Penning, & Schimmele, 2005). The incongruence between the family caregivers’ beliefs and the values of the mainstream culture about care provision may result in the families feeling lost in their quest to meet the care needs of their loved ones diagnosed with a life-threatening disease. Coupled with the absence of a network of social support, the family caregivers’ feeling of isolation may well be an impediment to their well-being over the course of their family caregiving experience.  In addition to the physical and emotional demands of caregiving, a major source of tension the family caregiver interviewees identified was the ongoing concern about balancing the competing priorities between providing care to their ill loved ones and attending to their own needs. Studies on family caregiving have reported the demands of caregiving may impact various aspects of the family caregivers’ lives, including family and social relationships, career, financial, physical and emotional  118 health (Golant & Haskins, 2008; Milbury et al., 2013; Stenberg et al., 2010). Caregiving may be particularly stressful when caregivers, who are employed full-time, are also entrusted with the tasks of caring for the patient and other members of the family. The struggle that younger family caregivers encountered also reflects the tension in balancing demands of their caregiving role with career demands, especially when cancer care persists over time (Kim et al., 2010). In work environments where the family caregivers do not feel supported in their caregiving role, the demands of caregiving and employment can heighten their distress, even as the time required to provide care will impact the family caregivers’ personal routine/schedule and social functioning (Kurtz et al., 2004). With reduced opportunity to build up social support network and time for respite, the distress of the family caregivers may increase.   The family caregivers of the Chinese-speaking patients interviewed included spouses, children and family-in-law, all of whom were said to have assumed a key role in enhancing patient coping with cancer. Of particular interest is the type of relationship between the patients and the family caregivers observed in the Chinese-speaking sample in the study, where the persons providing care may not be members of the immediate family. This phenomenon has been cited in earlier work on the functioning of Chinese cancer support groups, where the family-in-law of the patient would attend the group meetings together with the patient, or attend the meetings without the patient (Bell et al., 2010). Understandably, the role of the family caregiver in the family unit shapes the experience of caring for their ill loved ones. It follows that the sources of distress may vary according to the nature of the relationship between the patient and the family caregiver. 5.3.2 Distress and emotional regulation in the family    The Chinese-speaking patients that were interviewed spoke about the interrelatedness between the ill person's and the family caregiver's distress, despite the lack of significant findings in the  119 association between patient distress and family caregiver distress in the quantitative analysis. In the interviews, both patients and family caregivers described situations where they sought to maintain a positive front by concealing their negative emotions from each other, in order to protect each party from further distress. Family members sought to downplay their distress in order to protect the patients, while patients noted their family’s attempt to regulate their emotions. The findings on the concealment of negative emotions in the family, following the diagnosis of a life-threatening illness, is consistent with other literature on the family caregiver's experience of cancer (J. Lee & Bell, 2011; Liu et al., 2005; Zahlis & Lewis, 2010). The practice of protective truthfulness in response to distressing news involves family caregivers choosing not to disclose the news of a cancer diagnosis to the patients; and also, patients deciding against disclosing their cancer diagnosis to more vulnerable members in the family. Such practices of non-disclosure encompass emotional concealment, with the goal to shield one another from the emotional pain that is regarded as disruptive to an already destabilized state in the family following the diagnosis of a life-threatening illness (Heidenreich et al., 2014). In this study, patients revealed that they also sought to protect their young children and elderly parents by not telling them of the news of the cancer diagnosis. Notably, members in the family were particularly sensitive and concerned with the impact of the illness on one another. To regulate the level of distress in the family following a cancer diagnosis, patients and family caregivers both assumed the responsibility to safeguard each other’s mental health, upholding non-disclosure in the interest of the other.   Much as the safeguard of an individual’s well-being is of priority in coping with distress in the family, maintaining harmony in the family unit may also be fundamental to the beliefs and values in some Chinese families. The quest for harmony is congruent with the notion of preserving stability, and for some Chinese people, it may also be deemed a virtue to be upheld in an individual’s life and in the family (Ching, Martinson, & Wong, 2012). In times of a life-threatening illness, the family caregivers  120 assumed the personal responsibility to regain stability in the family, by protecting their ill loved ones from further distress through emotional regulation. However, during the interviews, family caregivers (regardless of gender) spoke of their weariness and the uncertainty of coping with cancer in the family, especially where attempts to conceal distress from the patients had added further demands to the caregiving role. While family caregivers faced the ongoing concern to protect the patients, they also felt torn between acknowledging their own negative emotions and putting the patients’ interest ahead of their own, which also heightened the stress of caregiving. Female and male patients alike reported the distress they observed in their children, and at the same time, recognized the impact of the distress of their loved ones on their own emotional well-being. Particularly when distress relates to the fear of recurrence, the distress of the patient and the family caregiver will mutually influence one another (Mellon et al., 2007). Cognizant of the adverse impact that the distress of their loved ones could have on the patients themselves, the ill persons also sought to shield themselves from the distress of their family caregivers by regulating their emotions. This highlights the reciprocity of patient–caregiver relationships, “each party affect[ing] the other emotionally and behaviourally” (Chan et al., 2001, p. 389), where emotional regulation serves to protect other and oneself from the distress of cancer in the family.  5.4 Benefits of family caregiving and the positives of cancer Notwithstanding the burden of caregiving on family caregivers, the patients interviewed were articulate about the positive impact of family caregiving on their well-being. Notably, family caregivers played a pivotal role in strengthening patients to cope with cancer and enhance their QOL. The help that patients received was not confined to the practical tasks family caregivers performed in meeting the immediate needs of patients. In addition to practical help, the care that family caregivers rendered also encompassed informational, emotional and social support to the patients. To better prepare the patients  121 in managing the disease, family caregivers sought to update the patients with information on the nature of their illness and the recommended cancer care. Family caregivers also attempted to strengthen the patient's network of social support by seeking the help of friends to connect with other cancer survivors. Of particular importance to the patients were the family caregivers’ efforts to make themselves available and taking cues from the patients on how they wish to be helped. The capacity of the family caregivers to be present and to comply with the patients’ wishes on the "right" care, was deemed to be comforting to the patients and respectful of patient autonomy. This finding concurs with literature that highlights the importance of providing care and support that matches situation-specific patient needs (Chan et al., 2004; H. L. Cheng et al., 2013; So et al., 2013). Where there is incongruence between the patient and family caregiver expectations of care needs, the resultant tension will add to the distress of both the patient and the family caregiver.  Patients and family caregivers also spoke of the positive aspects of cancer when the cancer situation was seen to improve their quality of life. The benefits from the cancer experience may be classified according to three domains of life, namely: enhanced social resources, enhanced personal resources, and improved coping skills (Jim & Jacobsen, 2010). Despite the distress of the cancer diagnosis, patients and family caregivers reported positive changes that were experienced individually, and as a family unit, while living with the illness. Notwithstanding the challenges encountered, these cancer patients spoke of the experience of a new emotional depth and enhanced relationships with family that characterized life after the cancer diagnosis (H. L. Cheng, Sit, & Cheng, 2015; Miller, 2010). Patient encounters of personal growth were included as reports of increased capacity to find meaning and purpose in their suffering, also motivating them to seek ways to support other cancer patients. In addition, the patients’ narratives of enlarged vision also included increased appreciation for family members and friends, thereby strengthening relationships in the social network of the patients and  122 further enhancing patient QOL. However, such teachable moments of cancer may not be the experience of all cancer patients. Indeed, the illness experiences among patients are diverse, and may include both positive and negative transformations in the ill person’s perspective and identity following a cancer diagnosis (Kahana et al., 2011). It follows that it would be important to neither insist nor reject benefit-findings in the patients’ experience of cancer, where either stance would be disregarding of the patients’ interpretation of their own illness experience, and possibly leading to increased isolation of the individuals (Jim & Jacobsen, 2010).  In caring for their ill loved ones, some family caregivers similarly reported positive impacts on themselves and on family relationships. The findings of positive experiences of family caregivers concur with results of previous studies on the family caregivers' experience in caring for terminally ill cancer patients in Hong Kong, where family caregivers associated the care they provided to a relational commitment that showed their love and support for the patient (Mok et al., 2003; Wong & Chan, 2007). For these family caregivers, they recognized an increased self-awareness, acknowledged their limitations in relieving the patient's suffering, and, at the same time, aspired to enhance the quality of the care they provide to their ill loved one and other members in the family. As in the patient's experience of changes in their social relationships, several family caregivers described closer relationships with the patient, and being more sensitive to patient needs. While some family caregivers interviewed reported that they have always shared close a relationship with their ill loved one, the role of caregiving has heightened their awareness of the patient's struggle with illness. This insight allowed family caregivers to share the pain of the patient; patients felt understood and validated and the dyadic relationship was strengthened.  While caregiving for their ill loved ones, the family caregivers observed increased openness to talk about cancer with friends and other individuals outside of the family, and, as such, enhanced the  123 support for their caregiving role. The willingness to reach out for support by disclosing to others the cancer diagnosis in the family reflected the family’s transition from their initial fear of the disease to motivation for managing the illness in the family. By proactively seeking to gain understanding of cancer, family caregivers are empowered in their role as care providers, and the family’s adaptation to the illness is further strengthened.  Summary Regardless of the cultural and linguistic background of patients and family caregivers, cancer may be understood as a serious illness that challenges the order of the individual’s world. The study findings on patients’ and family caregivers’ experience of distress and the impact on their QOL may be common to all diagnosed with cancer, irrespectively of their cultural and linguistic background. At the same time, the differences in the impact of patient distress across QOL domains between the Chinese-speaking and Anglophone groups in the study would suggest potential cultural variation in the responses to cancer. The insights to the nature of the patient’s and the family caregiver’s distress, and the process through which it affects the patient’s QOL, will enhance our understanding of the disruption that cancer brings to a person’s life and the implications of cancer for the family. Where older patients are found to report higher QOL, as mediated by lower distress, the changes in an individual’s social and personal roles are important considerations in clarifying the psychological adjustment of the ill person. In families where members are particularly adverse to the psychological distress arising from a cancer diagnosis, the responsibility to safeguard the mental health of members in the family is upheld by both the patient and the family caregiver, as each seeks to regulate the level of distress in the family. Notwithstanding the challenges of coping with cancer, the positive aspects of cancer are also reported by some patients and family caregivers, highlighting the complexity of living with the illness, individually  124 and as a family. The following chapter concludes the dissertation, with recommendations for future research to advance knowledge on the study topic.   125 Chapter 6: Conclusions and Recommendations  6.1 Summary of key findings In conclusion, the following key findings are summarized in this study, which focused on how the QOL of a cancer patient could be impacted by the level of distress of both the patient and the family caregiver, particularly in a Chinese-speaking population. First, the negative influence of patient distress on patient QOL was found to vary across the subscales of the QOL measure. This emphasizes the importance of examining the domains of QOL when investigating the impact of cancer on a patient’s well-being. Second, the study sheds light on the mediation effects of patient distress on patient emotional well-being, where lower distress in older patients explained better emotional functioning Third, the association among family caregiver distress, patient distress and patient QOL may be better understood in light of the nature of distress in family caregiving, as spoken about in the interviews. Although quantitative findings on the associations involving family caregiver distress are inconclusive, this study raises important questions about the sources of distress of the family caregiver, when examining how the role of family members can influence a patient as they cope with cancer.  6.2 Strengths and limitations of dissertation This study examined an important topic pertinent to the care of an understudied population of Chinese-speaking cancer patients and family caregivers, and encompassed a number of strengths. First, this study enhances our understanding of the impact of cancer on families of a minority ethnic group in Canada, a country with an increasingly diverse population. Even in the broader Canadian context, studies that examine perspectives on the impact of cancer care on patients and families are limited. Second, the inclusion of patient and family caregiver dyads in the study provides insights into an individual perspective as well as a dyadic perspective of the interrelatedness of patient and family caregiver experiences of cancer, increasing knowledge about the impact of cancer on the family. Third,  126 the study is enhanced by the diversity of cancer sites, as well as the perspectives of male and female patients and family caregivers represented in this research. Fourth, as a Chinese-speaking immigrant, I share many of the cultural beliefs and values of those whom I interviewed; the commonality allowed me to remain close to the data while I sought to make meaning of the narratives of the patients and family caregivers. Fifth, my familiarity with the culture enabled me to understand the subtext of the statements made in the interviews, thereby creating a space for the patients and family caregivers in which they could speak freely, minimizing the pressure on them to produce socially desirable responses. These factors may have increased the trustworthiness of the data. Lastly, the mixed-methods approach used in this study contributes to the richness and relevance of the data, where the voice of the health care recipients provides quantitative and qualitative evidence for the development of patient and family-centred care.  Notwithstanding the strengths of the dissertation, several limitations in the study need mentioning. The cross-sectional survey design used in the study sought to investigate the association among the variables of primary interest, namely patient and family caregiver distress and patient QOL. However, it did not allow for temporal relationships to be established. Particularly, the causal relations between the distress and QOL variables hypothesized in the conceptual framework may not be imputed without a longitudinal design. Nonetheless, the findings provide a means to assess compatibility with past studies concerning the association of the study variables. The small sample size of 57 patient and family caregiver pairs of observations limited the power to detect statistical significance in subgroup analysis. In this study, language is used as the key descriptor of the study populations. As such, patients and family caregivers of Chinese descent who do not speak Chinese are not represented in this research, and the cancer experiences investigated are limited to a segment of the broader Canadian Chinese population. The participants of the study were patients and their family caregivers who volunteered to be  127 in the study. Where voluntary choice is the means by which participants are accrued in a study, self-selection bias may occur. Among the self-selection of Chinese-speaking participants, these participants demonstrated knowledge about cancer and were receptive to talking about the illness, even as these individuals may also have felt strongly about the impact of cancer on the family. Further, these patient participants have sought to support other cancer survivors in efforts to improve cancer care, and hence, may have had less distress than other Chinese-speaking patients and families in the larger cancer community. However, it is not logistically feasible to conduct a true random survey of the cancer population. Where the small sample size of cancer patients and family caregivers reflects the challenges in primary data collection in my study setting, the difficulty in recruitment may be addressed if the routine collection of psychosocial data, such as QOL and distress scales, is considered as part of clinical cancer care. Despite the limitations, the study contributes to our understanding of ways that cultural and linguistic factors might explain differences in the experiences of cancer by patients and their family caregivers, while providing possibilities for future studies about cancer in culturally diverse settings. 6.3 Implications and knowledge translation This study provides insights into the cancer experience of a group of Chinese-speaking patients and family caregivers, expanding our knowledge of cross-cultural cancer care. The intended outcome is congruent with the Canadian national agenda of patient and family-centred care approach, where practitioners and system managers are further informed regarding the delivery of culturally-appropriate care that responds to the needs, values and beliefs of the ill persons and their family members (Accreditation Canada, 2016). In this regard, it is important to consider how cultural background shapes service needs, and in turn service delivery, to cancer patients and their families. Further, early intervention with patients and their family caregivers is important and necessary to alleviate distress in both members as a consequence of the disease and caregiving. Where age may be a factor affecting  128 distress and a patient’s quality of life, it is recommended to tailor interventions with patients and family members in younger and older age groups, to address the specific concerns that align with their experiences of cancer.  Study findings suggest that the patient-family caregiver dyad be considered as the unit of care; because of emotional reactions to cancer within a patient and their family, there is a recognized benefit to offer supportive cancer care from a systems approach. Some family caregivers encounter difficulties in finding adequate help or services to support their caregiving role. This deficit may persist when the family caregiver must put the needs of their ill loved ones ahead of their own. At the same time, the needs of the caregivers themselves could be neglected by healthcare professionals or their own friends when the focus is foremost on the needs of the patient. To help enhance the quality of life of patients, interventions also need to include language appropriate resources, to support the caregiving role of family members. Providing family caregivers with information about the disease, and the management of the physical and emotional effects of the disease, could help strengthen coping with the multiple demands on family associated with cancer. By providing information and mobilizing social resources to support cancer patients and family caregivers, understanding of the illness would increase and social isolation in cancer experiences could be prevented.  As part of my ongoing efforts to bridge research and practice, with the goal to contribute to our understanding of issues in cross-cultural research and care needs of cancer patients and families, insights about the research process and preliminary findings have been presented at national and international conferences, including the Canadian Association of Psychosocial Oncology Conference, the World Congress of Psycho-Oncology and the International Society of Quality of Life; researchers, clinicians and policy-makers were all represented in the audiences. Contributions to the body of literature on psycho-oncology will continue through journal publications, conference presentations, webinars, and  129 article postings focusing on patient and family caregivers research and advocacy networks. Following from presentations made to healthcare professionals during the data collection phase of the study, I will continue to disseminate information from this study to cancer care professionals and decision-makers at the BC Cancer Agency annual conference, professional rounds at cancer centres and presentations to cancer care organizations in the community. Other planned important knowledge translation activities include talks at cancer support groups, in order to engage cancer patients and their families in conversations on care that aligns with their goals. The feedback from patients and family caregivers will be conveyed to clinicians through the activities outlined above in order to engage healthcare professionals active in their practice of patient and family-centered care.   6.4 Future research  From this study, we have observed the complexities of investigating the impact of cancer on the family, and particularly in the efforts undertaken to deepen our understanding of the relations between psychological distress and QOL among cancer patients and family caregivers from a culturally diverse population in Canada. It follows that more studies, quantitative and qualitative, are needed to strengthen the evidence of this emerging topic in supportive cancer care and family caregiving research. Future studies could build on this work by adopting a longitudinal study design. This design would enable tracking changes in patient quality of life over time and clarifying causal associations among the key variables examined in the study. For future investigations of the role of family caregiver distress, it would be prudent to consider using a different measure for this variable. Particularly, the family caregiver distress scale could include measurement of family caregiver preparedness and mastery in providing care, acknowledging family caregiving burden. In recognition of the role of the family in patient coping, more research is needed to evaluate the impact of interventions that empower the family as part of integrative patient care. In light of the greater burden and higher emotional distress reported by  130 women family caregivers, it will be important that more studies examine the experience of female caregivers and identify strategies to help them manage the demands associated with providing care. While the conceptual framework underpinning this study centred on the quality of life of patients, it would also be important to examine the impact of distress on family caregiver QOL. Considering the burden of caregiving and the interrelatedness of patient and family caregiver responses to cancer highlighted in the study findings, further investigation on the family caregiver’s QOL as an outcome variable would help us better understand the impact of cancer on the family. In view of the cultural context in which individuals make meaning of cancer, future studies that include qualitative interviews with other ethnic groups would be important to better understand the impact of the illness on family relationships. While the interview questions in this study were relatively prescriptive, further qualitative research is needed, to better understand how an inductive approach in posing the questions may elicit different responses. Granted the limitation of using language as the only indicator of culture in this study, future research to examine patient and family caregiver dyads of Chinese ancestry where they do not speak Chinese would be relevant, to gain insights into the role of acculturation on family adaptation to cancer. At the same time, the knowledge of the persistence and change of cultural values and practices over time in such a cancer population will enhance the understanding of Canadian patients and family caregivers living between cultures. With the average annual number of new cancer cases estimated to increase 79% in the next two decades, a substantial number of Canadians are anticipated to be living with a cancer diagnosis (Canadian Cancer Society, 2015). At the same time, the improving survival rate reported for most types of cancer, also means that people are living longer after their diagnosis, and are anticipated to have unique and complex healthcare needs in the course of their cancer journey. Where cancer is a family disease, the illness journey also includes the family caregivers, being the “invisible pillars of the  131 healthcare system” (The Family Caregivers’ Network Society, 2010, p. 7). The existing limitations of  healthcare budgets will mean that families will have to carry even more increased burdens of caregiving. Particularly for families where cultural values and beliefs define the role of the family caregivers, understanding of the cultural and social context of ill persons and family members is foundational to patient- and family-centred care. Future planning of health resources, towards a sustainable system of cancer care in Canada, must involve partnership with the family and provide support for family caregivers.  132 References Accreditation Canada. (2016).  Retrieved June 21, 2016 from  https://accreditation.ca/client-and-family-centred-care. Almutairi, A. F., Dahinten, V. S., & Rodney, P. (2015). Almutairi's Critical Cultural Competence model for a multicultural healthcare environment. Nursing Inquiry, 22(4), 317-325. doi:10.1111/nin.12099 Ashing-Giwa, K., & Lim, J. W. (2011, 2011). Examining emotional outcomes among a multiethnic cohort of breast cancer survivors. Oncology nursing forum, 38, 279-288. Ashing-Giwa, K., Padilla, G., Tejero, J., Kraemer, J., Wright, K., Coscarelli, A., . . . Hills, D. (2004). Understanding the breast cancer experience of women: A qualitative study of African American, Asian American, Latina and Caucasian cancer survivors. Psycho-oncology, 13(6), 408-428.  Bell, K., Lee, J., Foran, S., Kwong, S., & Christopherson, J. (2010). Is there an "ideal cancer" support group? Key findings from a qualitative study of three groups. Journal of Psychosocial Oncology, 28(4), 432-449. doi:10.1080/07347332.2010.488140 Bevan, J. L., & Pecchioni, L. L. (2008). Understanding the impact of family caregiver cancer literacy on patient health outcomes. Patient education and counseling, 71(3), 356-364.  Bloom, J. R. (2000). The role of family support in cancer control. In L. Baider, C. L. Cooper, & A. Kaplan De-Nour (Eds.), Cancer and the family (pp. 55-71). New York: John Wiley & Sons. Boehmer, S., Luszczynska, A., & Schwarzer, R. (2007). Coping and quality of life after tumor surgery: Personal and social resources promote different domains of quality of life. Anxiety, Stress, & Coping, 20(1), 61-75. doi:10.1080/10615800701195439 Bowen, D. J., Alfano, C. M., McGregor, B. A., Kuniyuki, A., Bernstein, L., Meeske, K., . . . Barbash, R. B. (2007). Possible socioeconomic and ethnic disparities in quality of life in a cohort of breast cancer survivors. Breast Cancer Research and Treatment, 106(1), 85-95. doi:10.1007/s10549-006-9479-2 Bower, J. E., Meyerowitz, B. E., Desmond, K. A., Bernaards, C. A., Rowland, J. H., & Ganz, P. A. (2005). Perceptions of positive meaning and vulnerability following breast cancer: predictors and outcomes among long-term breast cancer survivors. Ann Behav Med, 29(3), 236-245. doi:10.1207/s15324796abm2903_10 Brem, S., & Kumar, N. B. (2011, 2011/02//). Management of treatment-related symptoms in patients with breast cancer: current strategies and future directions. Clinical journal of oncology nursing, 15, 63+. Broderick, C. B. (1993). Understanding family processes: Basics of family systems theory. Newbury Park, CA: Sage. Byar, K. L., Berger, A. M., Bakken, S. L., & Cetak, M. A. (2006). Impact of Adjuvant Breast Cancer Chemotherapy on Fatigue, Other Symptoms, and Quality of Life. Oncology nursing forum, 33(1), E18-E26. doi:10.1188/06.ONF.E18-E26 Canadian Cancer Society. (2015). Canadian Cancer Statistics.  Retrieved July 1, 2015 from  http://www.cancer.ca/~/media/cancer.ca/CW/publications/Canadian%20Cancer%20Statistics/Canadian-Cancer-Statistics-2015-EN.pdf. Cella, D. F., & Tulsky, D. S. (1993). Quality of Life in Cancer: Definition, Purpose, and Method of Measurement. Cancer Investigation, 11(3), 327-336.  Cella, D. F., Tulsky, D. S., Gray, G., Sarafian, B., Linn, E., Bonomi, A., . . . Brannon, J. (1993). The Functional Assessment of Cancer Therapy scale: development and validation of the general measure. Journal of Clinical Oncology, 11(3), 570-579.   133 Chan, C. W. H., Hon, H. C., Chien, W. T., & Lopez, V. (2004). Social support and coping in Chinese patients undergoing cancer surgery. Cancer nursing, 27(3), 230-236.  Chan, C. W. H., Molassiotis, A., Yam, B. M. C., Chan, S. J., & Lam, C. S. W. (2001). Traveling through the cancer trajectory: Social support perceived by women with gynecologic cancer in Hong Kong. Cancer nursing, 24(5), 387-394. doi:Doi 10.1097/00002820-200110000-00011 Chen, H. Y., & Boore, J. (2010). Translation and back-translation in qualitative nursing research: methodological review. Journal of Clinical Nursing, 19(1-2), 234-239. doi:10.1111/j.1365-2702.2009.02896.x Chen, M. L., Chu, L., & Chen, H. C. (2004). Impact of cancer patients' quality of life on that of spouse caregivers. Supportive Care in Cancer, 12(7), 469-475. doi:DOI 10.1007/s00520-004-0636-z Cheng, H. L., Sit, J. W. H., Chan, C. W. H., So, W. K. W., Choi, K. C., & Cheng, K. K. F. (2013). Social support and quality of life among Chinese breast cancer survivors: Findings from a mixed methods study. European Journal of Oncology Nursing, 17(6), 788-796.  Cheng, H. L., Sit, J. W. H., & Cheng, K. F. (2015). Negative and positive life changes following treatment completion: Chinese breast cancer survivors’ perspectives. Supportive Care in Cancer, 24(2), 791-798. doi:10.1007/s00520-015-2845-z Cheng, Y. H., Chi, I., Boey, K. W., & Ko, L. S. F. (1996). Community care and the elderly in Hong Kong: health-related quality of life and service utilisation patterns. Hong Kong Journal of Gerontology, 10(Suppl.), 355-360.  Ching, S. S. Y., Martinson, I. M., & Wong, T. K. S. (2012). Meaning Making: Psychological Adjustment to Breast Cancer by Chinese Women. Qualitative health research, 22(2), 250-262. doi:10.1177/1049732311421679 Chui, T., Tran, K., & Flanders, J. (2005). Chinese Canadians: Enriching the cultural mosaic. Canadian Social Trends(76), 24.  Cohen, C., Colantonio, A., & Vernich, L. (2002). Positive aspects of caregiving: rounding out the caregiver experience. International Journal of Geriatric Psychiatry, 17(2), 184-188. doi:Doi 10.1002/Gps.561 Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155-159.  Cordova, M. J., Cunningham, L. L. C., Carlson, C. R., & Andrykowski, M. A. (2001). Social constraints, cognitive processing, and adjustment to breast cancer. Journal of Consulting and Clinical Psychology, 69(4), 706-711. doi:Doi 10.1037//0022-006x.69.4.706 Creswell, J., & Plano Clark, V. (2011). Designing and conducting mixed methods research. Sage Publications: Thousand Oaks, California.  de Leeuw, E. (2005). To Mix or Not to Mix Data Collection Modes in Surveys. Journal of Official Statistics, 21(2), 233–255.  Dobbins, J. F. (2007). Connections of Care: Relationships and Family Caregiver Narratives. In R. Josselson, A. Lieblich, & D. P. McAdams (Eds.), (pp. 189-211). Washington, DC, US: American Psychological Association. Dumont, S., Turgeon, J., Allard, P., Gagnon, P., Charbonneau, C., & Vezina, L. (2006). Caring for a loved one with advanced cancer: Determinants of psychological distress in family caregivers. Journal of Palliative Medicine, 9(4), 912-921. doi:DOI 10.1089/jpm.2006.9.912 Dunning, H., Allison, W., Abonyi, S., & Crooks, V. (2008). A Mixed Method Approach to Quality of Life Research: A Case Study Approach. Social Indicators Research, 85(1), 145-158. doi:10.2307/27734571  134 Edwards, B., & Clarke, V. (2004). The psychological impact of a cancer diagnosis on families: The influence of family functioning and patients' illness characteristics on depression and anxiety. Psycho-oncology, 13(8), 562-576.  Emerson, R. M., Fretz, R. I., & Shaw, L. L. (1995). Writing Ethnographic Fieldnotes. Chicago: University of Chicago Press.  Eton, D. T., Lepore, S. J., & Helgeson, V. S. (2005). Psychological distress in spouses of men treated for early-stage prostate carcinoma. Cancer, 103(11), 2412-2418.  Feld, S., Dunkle, R. E., Schroepfer, T., & Shen, H. W. (2006). Expansion of elderly couples' IADL caregiver networks beyond the marital dyad. International Journal of Aging & Human Development, 63(2), 95-113. doi:Doi 10.2190/Cw8g-Pb6b-Ncgh-Ht1m Ferrans, C. (2005). The meaning of health-related quality of life. Oncology nursing forum, 32(1), 157-158.  Ferrans, C., Zerwic, J. J., Wilbur, J. E., & Larson, J. L. (2005). Conceptual model of health-related quality of life. Journal of Nursing Scholarship, 37(4), 336-342. doi:DOI 10.1111/j.1547-5069.2005.00058.x Ferrell, B. R., Hassey Dow, K., & Grant, M. (1995). Measurement of the quality of life in cancer survivors (Vol. 4, pp. 523-531): Springer Netherlands. Field, A. (2005). Discovering statistics using SPSS. Sage publications: London.  Fletcher, B. S., Miaskowski, C., Given, B., & Schumacher, K. (2012). The cancer family caregiving experience: An updated and expanded conceptual model. European Journal of Oncology Nursing, 16(4), 387-398.  Furukawa, T. A., Kessler, R. C., Slade, T., & Andrews, G. (2003). The performance of the K6 and K10 screening scales for psychological distress in the Australian National Survey of Mental Health and Well-Being. Psychological Medicine, 33, 3357-3362.  Gaugler, J. E., Hanna, N., Linder, J., Given, C. W., Tolbert, V., Kataria, R., & Regine, W. F. (2005). Cancer caregiving and subjective stress: a multi-site, multi-dimensional analysis. Psycho-oncology, 14(9), 771-785. doi:10.1002/pon.916 Given, B., Given, C., & Kozachik, S. (2001). Family support in advanced cancer (pp. - 213): - John Wiley & Sons, Ltd. Given, B., Sherwood, P., & Given, C. (2009). Family care during cancer care. In S. M. Miller, D. J. Bowen, R. T. Croyle, & J. H. Rowland (Eds.), Handbook of Cancer Control and Behavioural Science (pp. 391-408). Washington DC: American Psychological Association  Given, B., Sherwood, P., & Given, C. (2011). Support for Caregivers of Cancer Patients: Transition After Active Treatment. Cancer Epidemiology Biomarkers & Prevention, 20(10), 2015-2021.  Given, B., & Sherwood, P. R. (2006). Family Care for the Older Person With Cancer. Seminars in oncology nursing, 22(1), 43-50. doi:10.1016/j.soncn.2005.10.006 Glajchen, M. (2004). The emerging role and needs of family caregivers in cancer care. J Support Oncol, 2(2), 145.  Golant, M., & Haskins, N. (2010). Family and caregivers of cancer survivors: Being a strengthened ally-A community perspective. In K. Miller (ed) Medical and psychosocial care of the cancer survivor, 163-179.  Golant, M., & Haskins, N. V. (2008). "Other Cancer Survivors": The Impact on Family and Caregivers. Cancer Journal, 14(6), 420-424.  Golant, M., & Taylor-Ford, M. (2010). Post-traumatic stress in cancer survivors. In K. Miller (ed) Medical and psychosocial care of the cancer survivor, 31-42.   135 Gotay, C. (2000). Culture, cancer, and the family. In L. Baider, C. L. Cooper & A. Kaplan De-Nour (Eds.). Cancer and the family, . New York: John Wiley & Sons., 95-110.  Gotay, C. (2004). Quality of life in culturally diverse cancer patients In R.J. Moore and D. Spiegel (eds). Cancer, culture, and communication. Plenum Publishers: New York. (pp. 55-75). New York: Kluwer Acaedmic/Plenum Publisher. Graham, J. (2012). Missing data: Analysis and design. New York: Springer. Graham, J., & Schafer, J. (1999). On the performance  of multiple imputation for multivariate data with small sample size. In R. Hoyle (Ed.), Statistical Strategies for Small Sample Research, 1-29, Thousand Oaks, CA: Sage.  Gyimah, S. O.-M. (2001). Missing Data in Quantitative Social Research. Population Studies Centre Discussion Papers Series, 15(14).  Hagedoorn, M., Sanderman, R., Bolks, H. N., Tuinstra, J., & Coyne, J. C. (2008). Distress in couples coping with cancer: A meta-analysis and critical review of role and gender effects. Psychological Bulletin, 134(1), 1-30. doi:10.1037/0033-2909.134.1.1 Heidenreich, M. T., Koo, F. K., & White, K. (2014). The experience of Chinese immigrant women in caring for a terminally ill family member in Australia. Collegian, 21(4), 275-285. doi:http://dx.doi.org/10.1016/j.colegn.2013.06.002 Helgeson, V. S. (2003). Social Support and Quality of Life. Quality of Life Research, 12, 25-31.  Helgeson, V. S., Novak, S. A., Lepore, S. J., & Eton, D. T. (2004). Spouse social control efforts: Relations to health behavior and well-being among men with prostate cancer. Journal of Social and Personal Relationships, 21(1), 53-68. doi:Doi 10.1177/0265407504039840 Ho, S. M. Y., Saltel, P., Machavoine, J., Rapoport-Hubschman, N., & Spiegel, D. (2004). Cross-cultural aspects of cancer care. In R. J. Moore & D. Spiegel (Eds.), Cancer, culture, and communication. (pp. 157-183). New York: Plenum Publishers. Hodges, L. J., Humphris, G. M., & Macfarlane, G. (2005). A meta-analytic investigation of the relationship between the psychological distress of cancer patients and their carers. Social science & medicine, 60(1), 1-12. doi:10.1016/j.socscimed.2004.04.018 Hodgkinson, K., Butow, P., Fuchs, A., Hunt, G. E., Stenlake, A., Hobbs, K. M., . . . Wain, G. (2007). Long-term survival from gynecologic cancer: Psychosocial outcomes, supportive care needs and positive outcomes. Gynecologic Oncology, 104(2), 381-389.  Hodgkinson, K., Butow, P., Hunt, G. E., Wyse, R., Hobbs, K. M., & Wain, G. (2007). Life after cancer: couples' and partners' psychological adjustment and supportive care needs. Supportive Care in Cancer, 15(4), 405-415.  Honea, N. J., Brintnall, R., Given, B., Sherwood, P., Colao, D. B., Somers, S. C., & Northouse, L. L. (2008). Putting evidence into practice®: nursing assessment and interventions to reduce family caregiver strain and burden. Clinical journal of oncology nursing, 12(3), 507-516. doi:10.1188/08.CJON.507-516 Hsu, J. (1985). Chinese family: relations, problems, and therapy. Chinese culture and mental health.  Huang, X., Butow, P., Meiser, B., & Goldstein, D. (1999). Attitudes and information needs of Chinese migrant cancer patients and their relatives. Australian and New Zealand Journal of Medicine, 29(2), 207-213. doi:DOI 10.1111/j.1445-5994.1999.tb00685.x Jim, H., & Jacobsen, P. (2010). Finding benefits in the cancer experience: Post-traumatic growth. In K. Miller (Ed.), Medical and psychosocial care of the cancer survivor. (pp. 43-54). London: Jones and Bartlett Publishers.  136 Kagawa-Singer, M., & Nguyen, T. (2000). A cross-cultural comparison of social support among Asian-American and Euro-American women following breast cancer. In L. Baider, C. L. Cooper, & A. Kaplan De-Nour (Eds.), Cancer and the family (pp. 73-94). New York: John Wiley & Sons. Kahana, B., Kahana, E., Deimling, G., Sterns, S., & VanGunten, M. (2011). Determinants of altered life perspectives among older-adult long-term cancer survivors. Cancer Nurs, 34(3), 209-218. doi:10.1097/NCC.0b013e3181fa56b0 Kessler, R. C., Barker, P. R., Colpe, L. J., Epstein, J. F., Gfroerer, J. C., Hiripi, E., . . . Zaslavsky, A. M. (2003). Screening for serious mental illness in the general population. Archieves of General Psychiatry, 60, 184-189.  Kim, Y., Baker, F., Spillers, R., & Wellisch, D. (2006). Psychological adjustment of cancer caregivers with multiple roles. Psycho-oncology, 15(9), 795-804. doi:10.1002/pon.1013 Kim, Y., & Given, B. (2008). Quality of life of family caregivers of cancer survivors - Across the trajectory of the illness. Cancer, 112(11), 2556-2568. doi:Doi 10.1002/Cncr.23449 Kim, Y., Kashy, D., Wellisch, D., Spillers, R., Kaw, C. K., & Smith, T. (2008). Quality of Life of Couples Dealing with Cancer: Dyadic and Individual Adjustment among Breast and Prostate Cancer Survivors and Their Spousal Caregivers. Annals of Behavioral Medicine, 35(2), 230-238. doi:10.1007/s12160-008-9026-y Kim, Y., Kashy, D. A., Spillers, R. L., & Evans, T. V. (2010). Needs assessment of family caregivers of cancer survivors: three cohorts comparison. Psycho-oncology, 19(6), 573-582.  Kim, Y., Loscalzo, M. J., Wellisch, D. K., & Spillers, R. L. (2006). Gender differences in caregiving stress among caregivers of cancer survivors. Psycho-oncology, 15(12), 1086-1092. doi:Doi 10.1002/Pon.1049 Kim, Y., Schulz, R., & Carver, C. S. (2007). Benefit finding in the cancer caregiving experience. Psychosomatic medicine, 69(3), 283-291.  Kotkamp-Mothes, N., Slawinsky, D., Hindermann, S., & Strauss, B. (2005). Coping and psychological well being in families of elderly cancer patients. Critical reviews in oncology/hematology, 55(3), 213-229. doi:10.1016/j.critrevonc.2005.03.006 Kurtz, M. E., Kurtz, J. C., Given, C. W., & Given, B. A. (2004). Depression and physical health among family caregivers of geriatric patients with cancer - a longitudinal view. Medical Science Monitor, 10(8), Cr447-Cr456.  Larkin, P. J., Dierckx de Casterlé, B., & Schotsmans, P. (2007). Multilingual Translation Issues in Qualitative Research: Reflections on a Metaphorical Process. Qualitative health research, 17(4), 468-476. doi:10.1177/1049732307299258 Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.  Lee, J., & Bell, K. (2011). The Impact of Cancer on Family Relationships Among Chinese Patients. Journal of Transcultural Nursing, 22(3), 225-234.  Lee, M. D. (2007). Correlates of consequences of intergenerational caregiving in Taiwan. Journal of advanced nursing, 59(1), 47-56. doi:10.1111/j.1365-2648.2007.04274.x Lee, R., Rodin, G., Devins, G., & Weiss, M. G. (2001). Illness experience, meaning and help-seeking among Chinese immigrants in Canada with chronic fatigue and weakness. Anthropology & Medicine, 8(1), 89-107. doi:10.1080/13648470120063915 Leeman, J., Chang, Y., Voils, C. I., Crandell, J. L., & Sandelowski, M. (2011). A mixed-methods approach to synthesizing evidence on mediators of intervention effects. West J Nurs Res, 33(7), 870-900. doi:10.1177/0193945911402365 Lewis, F. M. (1986). The impact of cancer on the family: A critical analysis of the research literature. Patient education and counseling, 8(3), 269-289. doi:10.1016/0738-3991(86)90005-4  137 Li, H. (2004). Barriers to and Unmet Needs for Supportive Services: Experiences of Asian-American Caregivers (Vol. 19, pp. 241-260): Springer Netherlands. Lingler, J. H., Sherwood, P. R., Crighton, M. H., Song, M., & Happ, M. B. (2008). Conceptual Challenges in the Study of Caregiver-Care Recipient Relationships. Nursing research, 57(5), 367-372. doi:10.1097/01.NNR.0000313499.99851.0c Lissandrello, G., Respini, D., & Tralongo, P. (2006). Latent troubles of a cancer patient's caregiver. Supportive & Palliative Cancer Care, 2(3), 123-126.  Liu, J. E., Mok, E., & Wong, T. (2005). Perceptions of supportive communication in Chinese patients with cancer: experiences and expectations. J Adv Nurs, 52(3), 262-270. doi:10.1111/j.1365-2648.2005.03583.x Loke, A. Y., Liu, C. F. F., & Szeto, Y. (2003). The difficulties faced by informal caregivers of patients with terminal cancer in Hong Kong and the available social support. Cancer nursing, 26(4), 276-283. doi:Doi 10.1097/00002820-200308000-00004 Lui, C.-W., Ip, D., & Chui, W. H. (2009). Ethnic Experience of Cancer: A Qualitative Study of Chinese–Australians in Brisbane, Queensland. Soc Work Health Care, 48(1), 14-37. doi:10.1080/00981380802440403 MacKinnon, D. P. (2008). Introduction to statistical mediation analysis. Taylor & Francis Group: New York.  Marshall, E. G., Wong, S. T., Haggerty, J. L., & Levesque, J.-F. (2010). Perceptions of unmet healthcare needs: what do Punjabi and Chinese-speaking immigrants think? A qualitative study. Bmc Health Services Research, 10(1), 1-8. doi:10.1186/1472-6963-10-46 Mehnert, A., & Koch, U. (2008). Psychological comorbidity and health-related quality of life and its association with awareness, utilization, and need for psychosocial support in a cancer register-based sample of long-term breast cancer survivors. Journal of psychosomatic research, 64(4), 383-391. doi:http://dx.doi.org/10.1016/j.jpsychores.2007.12.005 Mellon, S. (2002). Comparisons between cancer survivors and family members on meaning of the illness and family quality of life. Oncol Nurs Forum, 29(7), 1117-1125. doi:10.1188/02.ONF.1117-1125 Mellon, S., Kershaw, T., Northouse, L., & Freeman-Gibb, L. (2007). A family-based model to predict fear of recurrence for cancer survivors and their caregivers. Psycho-oncology, 16(3), 214-223. doi:10.1002/pon.1074 Meyerowitz, B., & Oh, S. (2009). Psychosocial response to cancer diagnosis and treatment. In S. Miller, D. Bowen, R. Croyle, & J. Rowland (Eds.), Cancer Control and Behavioural Science. (pp. 361-377). Washington, D.C.: American Psychological Association. Milbury, K., Badr, H., Fossella, F., Pisters, K. M., & Carmack, C. L. (2013). Longitudinal associations between caregiver burden and patient and spouse distress in couples coping with lung cancer. Supportive Care in Cancer, 21(9), 2371-2379.  Miller, K. (2010). Cancer survivorship today. In K. Miller (Ed.), Medical and psychosocial care of the cancer survivor. (pp. 9-20). London: Jones and Bartlett Publishers. Mok, E., Chan, F., Chan, V., & Yeung, E. (2003). Family experience caring for terminally ill patients with cancer in Hong Kong. Cancer nursing, 26(4), 267-275.  Moore, R., & Butow, P. (2004). Culture and Oncology: Impact of context effects In Moore & Spiegel (eds). Cancer, culture and communication (pp. 15-54). New York: Kluwer Academic/Plenum Publisher. Morse, J. M. (1991). Approaches to Qualitative-Quantitative Methodological Triangulation. Nursing research, 40(2), 120-123. doi:10.1097/00006199-199103000-00014  138 Ngo-Metzger, Q., McCarthy, E. P., Burns, R. B., Davis, R. B., Li, F. P., & Phillips, R. S. (2003). Older asian americans and pacific islanders dying of cancer use hospice less frequently than older white patients. The American Journal of Medicine, 115(1), 47-53. doi:10.1016/S0002-9343(03)00258-4 Nijboer, C., Triemstra, M., Tempelaar, R., Mulder, M., Sanderman, R., & van den Bos, G. A. (2000). Patterns of caregiver experiences among partners of cancer patients. The Gerontologist, 40(6), 738-746.  Nishigaki, M., Oya, M., Ueno, M., Arai, M., Yamaguchi, T., Muto, T., & Kazuma, K. (2007). The Influence of Life Stage on Psychosocial Adjustment in Colorectal Cancer Patients. Journal of Psychosocial Oncology, 25(4), 71-87. doi:10.1300/J077v25n04_05 Northouse, L. (2005). Helping families of patients with cancer. Oncology nursing forum, 32(4), 743-750. doi:10.1188/05.ONF.743-750 Northouse, L., Katapodi, M., Schafenacker, A., & Weiss, D. (2012). The Impact of Caregiving on the Psychological Well-Being of Family Caregivers and Cancer Patients. Seminars in oncology nursing, 28(4), 236-245. doi:http://dx.doi.org/10.1016/j.soncn.2012.09.006 Northouse, L., Kershaw, T., Mood, D., & Schafenacker, A. (2005). Effects of a family intervention on the quality of life of women with recurrent breast cancer and their family caregivers. Psychooncology, 14(6), 478-491. doi:10.1002/pon.871 Northouse, L., Mellon, S., Harden, J., & Schafenacker, A. (2009). Long-term effects of cancer on families of adult cancer survivors. In S.M. Miller, D.J. Bowen, R.T. Croyle & J.H. Rowland (eds). Handbook of Cancer Control and Behavioural Science. APA.  Northouse, L., Mood, D., Kershaw, T., Schafenacker, A., Mellon, S., Walker, J., . . . Decker, V. (2002). Quality of life of women with recurrent breast cancer and their family members. Journal of Clinical Oncology, 20(19), 4050-4064.  Northouse, L., Mood, D., Templin, T., Mellon, S., & George, T. (2000). Couples' patterns of adjustment to colon cancer. Social science & medicine, 50(2), 271-284. doi:10.1016/S0277-9536(99)00281-6 Pecchioni, L., Thompson, L., & Anderson, J. (2006). Interrelations between family communication and health communication. In Turner LH, West R, editors. The family communication sourcebook. Thousand Oaks, CA:Sage., 447–468.  Pierce, G., Sarason, B., Sarason, I., Joseph, H., & Henderson, C. (1996). Conceptualizing and assessing social support in the context of the family. Handbook of Social Support and the Family (eds). G. Pierce, B. Sarason, I. Sarason. Plenum Press, New York, 3-23.  Pinquart, M., & Sorensen, S. (2004). Associations of caregiver stressors and uplifts with subjective well-being and depressive mood: a meta-analytic comparison. Aging & Mental Health, 8(5), 438-449. doi:Doi 10.1080/13607860410001725036 Pinquart, M., & Sorensen, S. (2005). Ethnic differences in stressors, resources, and psychological outcomes of family caregiving: a meta-analysis. Gerontologist, 45(1), 90-106.  Ridner, S. H. (2004). Psychological distress: concept analysis. Journal of advanced nursing, 45(5), 536-545. doi:10.1046/j.1365-2648.2003.02938.x Rolland, J. (2005). Cancer and the Family: An Integrative Model (pp. - 2584): - Wiley Subscription Services, Inc., A Wiley Company. Rubin, D. B. (1987). Multiple Imputation for Nonresponse in Surveys. J. Wiley & Sons, New York.  Sanders, S. J., & Kittay, E. F. (2005). Shouldering the burden of care. Hastings Center Report, 35(5), 14-15. doi:DOI 10.1353/hcr.2005.0085  139 Santos, H. P. O., Black, A. M., & Sandelowski, M. (2015). Timing of Translation in Cross-Language Qualitative Research. Qualitative health research, 25(1), 134-144. doi:10.1177/1049732314549603 Schumacher, K. L., Stewart, B. J., Archbold, P. G., Caparro, M., Mutale, F., & Agrawal, S. (2008). Effects of caregiving demand, mutuality, and preparedness on family caregiver outcomes during cancer treatment. Oncology nursing forum, 35(1), 49-56.  Segrin, C., & Badger, T. A. (2014). Psychological and physical distress are interdependent in breast cancer survivors and their partners. Psychology, Health & Medicine, 19(6), 716-723. doi:10.1080/13548506.2013.871304 Sherwood, P., Given, B., Bender, C., Baum, A., Schulz, R., Lieberman, F., & Sereika, S. (2007). Mind-body interactions in neuro-oncology caregivers. Oncology nursing forum, 34(1), 232-232.  Siegel, R., DeSantis, C., Virgo, K., Stein, K., Mariotto, A., Smith, T., . . . Ward, E. (2012). Cancer treatment and survivorship statistics, 2012. CA: A Cancer Journal for Clinicians, 62(4), 220-241. doi:10.3322/caac.21149 Simpson, P. (2005). Hong Kong families and breast cancer: Beliefs and adaptation strategies. Psycho-oncology, 14(8), 671-683.  Sinding, C., & Wiernikowski, J. (2008). Disruption foreclosed: older women's cancer narratives. Health (London), 12(3), 389-411. doi:10.1177/1363459308090055 So, W. K. W., Chan, C. W. H., Choi, K. C., Wan, R. W. M., Mak, S. S. S., & Chair, S. Y. (2013). Perceived Unmet Needs and Health-Related Quality of Life of Chinese Cancer Survivors at 1 Year After Treatment. Cancer nursing, 36(3), E23-E32. doi:10.1097/NCC.0b013e318263f28e Squires, A. (2009). Methodological challenges in cross-language qualitative research: A research review. International journal of nursing studies, 46(2), 277-287. doi:http://dx.doi.org/10.1016/j.ijnurstu.2008.08.006 Stanton, A. L., Danoff-Burg, S., Cameron, C. L., Bishop, M., Collins, C. A., Kirk, S. B., . . . Twillman, R. (2000). Emotionally expressive coping predicts psychological and physical adjustment to breast cancer. J Consult Clin Psychol, 68(5), 875-882.  Statistics Canada. (2006). Census 2006.  Retrieved February 24, 2012 from  http://www.statcan.gc.ca/pub/11-402-x/2012000/chap/imm/tbl-eng.htm. Statistics Canada. (2011). Canada Year book: Foreign-born and visible minority population projections, by census metropolitan area, 2006 and 2031.:  Retrieved February 24, 2012 from  http://www.statcan.gc.ca/pub/11-402-x/2012000/chap/imm/imm-eng.htm. Stenberg, U., Ruland, C. M., & Miaskowski, C. (2010). Review of the literature on the effects of caring for a patient with cancer. Psycho-oncology, 19(10), 1013-1025. doi:10.1002/pon.1670 Tabachnick, B., & Fidell, L. (2007). Using multivariate statistics. (5th ed.) Boston, MA: Pearson Education.  Tashakkori, A., & Teddlie, C. (1998). Mixed Methodology: Combining qualitative and quantitative approaches. Thousand Oaks, CA: Sage.  Temple, B., & Young, A. (2004). Qualitative Research and Translation Dilemmas. Qualitative Research, 4(2), 161-178. doi:10.1177/1468794104044430 The Family Caregivers’ Network Society. (2010). Supporting Family Caregivers: An Action Plan for British Columbia. Retrieved from  Thomas, B. C., NandaMohan, V., Nair, M. K., & Pandey, M. (2011). Gender, age and surgery as a treatment modality leads to higher distress in patients with cancer. Supportive Care in Cancer, 19(2), 239-250. doi:10.1007/s00520-009-0810-4 Thorne, S. (1985). The family cancer experience. Cancer Nurs, 8(5), 285-291.   140 Twinn, S. (1997). An exploratory study examining the influence of translation on the validity and reliability of qualitative data in nursing research. Journal of advanced nursing, 26(2), 418-423. doi:10.1046/j.1365-2648.1997.1997026418.x Ussher, J., Kirsten, L., Butow, P., & Sandoval, M. (2006). What do cancer support groups provide which other supportive relationships do not? The experience of peer support groups for people with cancer. Social science & medicine, 62(10), 2565-2576. doi:http://dx.doi.org/10.1016/j.socscimed.2005.10.034 van Nes, F., Abma, T., Jonsson, H., & Deeg, D. (2010). Language differences in qualitative research: is meaning lost in translation? European Journal of Ageing, 7(4), 313-316. doi:10.1007/s10433-010-0168-y van Ryn, M., Sanders, S., Kahn, K., van Houtven, C., Griffin, J. M., Martin, M., . . . Rowland, J. (2011). Objective burden, resources, and other stressors among informal cancer caregivers: a hidden quality issue? Psycho-oncology, 20(1), 44-52.  Wong-Kim, E., & Bloom, J. (2005). Depression experienced by young women newly diagnosed with breast cancer. Psycho-oncology, 14(7), 564-573. doi:10.1002/pon.873 Wong-Kim, E., Sun, A., & DeMattos, M. C. (2003). Assessing cancer beliefs in a Chinese immigrant community. Cancer Control, 10(5 suppl), 22.  Wong-Kim, E., Sun, A., Merighi, J. R., & Chow, E. A. (2005). Understanding quality-of-life issues in Chinese women with breast cancer: a qualitative investigation. Cancer Control, 12 Suppl 2, 6.  Wong, M. S., & Chan, S. W. C. (2007). The experiences of Chinese family members of terminally ill patients - a qualitative study. Journal of Clinical Nursing, 16(12), 2357-2364.  Wu, D. Y. H., & Tseng, W. S. (1985). Introduction: the characteristics of Chinese culture. In W. S. Tseng & D. Wu (Eds.), Chinese culture and mental health. (pp. 3-13). London: Academic Press, Inc. Wu, Z., Penning, M. J., & Schimmele, C. M. (2005). Immigrant Status and Unmet Health Care Needs. Canadian Journal of Public Health / Revue Canadienne de Sante'e Publique, 96(5), 369-373.  Yu, C. L. M., Fielding, R., Chan, C. L. W., Tse, V. K. C., Choi, P. H. K., Lau, W. H., . . . Sham, J. S. T. (2000). Measuring quality of life of Chinese cancer patients. Cancer, 88(7), 1715-1727. doi:10.1002/(SICI)1097-0142(20000401)88:7<1715::AID-CNCR28>3.0.CO;2-K Zahlis, E. H., & Lewis, F. M. (2010). Coming to Grips with Breast Cancer: The Spouse's Experience with His Wife's First Six Months. Journal of Psychosocial Oncology, 28(1), 79-97. doi:10.1080/07347330903438974 Zhou, C., Chu, J., & Wang, T. (2008). Reliability and validity of 10-item kessler scale (K10) Chinese version in evaluation of mental health status of Chinese population. . Chinese Journal of Clinical Psychology, 16(6), 627-629.      141 Appendices 142 Appendix A:  Demographics/Background information questions I. Anglophone patient 1. What year were you born? 2. Where were you born? 3. What is your highest education level attained? 4. What type (site) of cancer do you have? 5. What stage is your cancer diagnosis?  6. When were you diagnosed with cancer? 7. What cancer treatments have you had?  II. Anglophone family caregiver 1. What year were you born? 2. Where were you born? 3. What is your highest education level attained? 4. What type (site) of cancer is the patient diagnosed with? 5. What stage is his or her cancer diagnosis?  6. What is your relationship with the patient? 7. How long (months) have you been providing care to the patient?  III. Chinese-speaking patient 1. What year were you born? 2. Where were you born? 3. When did you migrate to Canada? (years in Canada?) 4. Did family members accompany you to Canada? Number of children and ages? 5. What is your highest education level attained? 6. What is your mother tongue?  7. How do you assess your English fluency (beginner, intermediate, advanced)? 8. What type (site) of cancer do you have? 9. What stage is your cancer diagnosis?  10. When were you diagnosed with cancer? 11. What cancer treatments have you had?  IV. Family caregiver of Chinese-speaking patient 1. What year were you born? 2. Where were you born? 3. If not born in Canada, when did you migrate to Canada? (years in Canada?) 4. What is your highest education level attained? 5. What is your mother tongue?  6. How do you assess your English fluency (beginner, intermediate, advanced)? 7. What type (site) of cancer is the patient diagnosed with? 8. What stage is his or her cancer diagnosis?  9. What is your relationship with the patient? 10. How long (months) have you been providing care to the patient? 143 Appendix B: Kessler Distress Scale (K6), English version   144 Appendix B: Kessler Distress Scale (K6), English version (continued)     145 Appendix C: Kessler Distress Scale (K6), Chinese version     146 Appendix C: Kessler Distress Scale (K6), Chinese version (continued)     147 Appendix D: Functional Assessment of Cancer Therapy (FACT-G), English version 4.0  148 Appendix D: Functional Assessment of Cancer Therapy (FACT-G), English version 4.0 (continued)   149 Appendix E: Functional Assessment of Cancer Therapy (FACT-G), Chinese version 4.0   150 Appendix E: Functional Assessment of Cancer Therapy (FACT-G), Chinese version 4.0 (continued)   151 Appendix F: Semi-structured interview script for Chinese-speaking patient I. Experience with cancer 1. How did you first learn of the diagnosis? 2. How did you react when you learned of the news? 3. How did you tell others, family and non-family members? How did they (family members and non-family members) react? 4. What changes have you noticed in your mood and feelings, from the time of the cancer diagnosis? 5. How has the experience of living with cancer affected your life? - Physically - Relationally with family and friends (social/family) - Emotionally - Ability to manage day to day living  (functionally)  II. Patient-family caregiver relationship 1. Who is the primary family member who provides care for you since the cancer diagnosis?  2. How does he/she (family caregiver) react to your illness? 3. How do you respond to his/her reaction?   III. Cultural issues for Chinese cancer patients 1. What are the main challenges facing Chinese people who are diagnosed with cancer? 2. How would you characterize your interactions with medical professionals (GP, treatment centre/hospital staff) since the cancer diagnosis? - Do you use an interpreter when meeting the oncologist?  3. How is cancer understood in the Chinese community? 4. Are there any differences between how cancer is understood in the Chinese community and Canadian society more broadly?  IV. Conclusion 1. Is there anything else you would like to tell me that we have not already discussed?   152 Appendix G: Semi-structured interview script for Chinese-speaking family caregiver I. Experience with cancer 1. How did you first learn of the diagnosis? 2. How did you react when you learned of the news? 3. How did you tell others, family and non-family members? How did they (family members and non-family members) react? 4. What changes have you noted in (patient) mood and feelings, from the time of the cancer diagnosis? 5. How has the experience of living with cancer affected his/her (patient) life? - Physically  - Relationships with family and friends (social/family)  - Emotionally - Ability to manage day to day living  (functionally)   II. Patient-family caregiver relationship 1. What is your experience like in caring for your__ (patient)? 2. How do you respond to the changes you noted in your __ (patient)?  3. What changes have you noted in your mood and feelings, since the time of the cancer diagnosis? 4. How has the experience of living with cancer affected your life? - Physically - Relationships with patient, family, friends  - Emotionally - Ability to manage day to day tasks  III. Cultural issues for Chinese cancer patients 1. What are the main challenges facing Chinese people who are diagnosed with cancer? 2. How would you characterize your interactions with medical professionals (GP, treatment centre/hospital staff) since the cancer diagnosis? - Do you use an interpreter when meeting the oncologist?  3. How is cancer understood in the Chinese community? 4. Are there any differences between how cancer is understood in the Chinese community and Canadian society more broadly?  IV. Conclusion 1. Is there anything else you would like to tell me that we have not already discussed?  153 Appendix H: Descriptive summary of 19 dyads with incomplete information   Variables Combined Group (n=19) Chinese-speaking Group  (n=12) Anglophone Group (n=7)  Mean (Standard deviation) Patient distress 4.53   (3.36)  4.82   (3.12) 4.0     (4.0) Caregiver distress 15.0   (1.41) 14.0   (--) 16.0   (--) QOL total 86.08 (15.3) 87.88 (14.68) 82.78 (17.27) Physical well-being 22.53 (5.78) 22.36 (5.8) 22.83 (6.27) Social well-being 24.02 (3.48) 23.97 (4.12) 24.11 (2.17) Emotional well-being 19.53 (3.91)        20.0 (3.38) 18.67 (4.97) Functional well-being 20.0   (5.7) 21.55 (4.13) 17.17 (7.41) Patient age 56.82 (9.97) 54.91 (8.78) 60.33 (11.94) Years patient lived in Canada 34.12 (18.82 24.36 (9.88) 52.0   (18.58) Caregiver age 58.5   (10.61) 66.0 (--) 51.0   (--) Years caregiver lived in Canada 39.5   (16.26) 28.0 (--) 51.0   (--) Months of caregiving        16.0  (11.31)        8.0   (--)        24.0  (--)      Counts (percentages) Male patient Male caregiver Breast cancer Bladder or colorectal cancer   Ovarian, prostate or uterine    cancer Stage I cancer Stage III cancer Patient with high school    education Patient with university or     advanced degree Caregiver  with high school     education Caregiver with university or     advanced degree Patient with advanced English Caregiver with advanced    English  Caregiver a spouse/partner of    patient Caregiver a child of patient More than 1 treatment type  2   (11.8) 1   (50) 11 (57.9) 5   (26.3) 2   (10.5)  5   (26.3) 5   (26.3) 7   (41.2)  5   (29.4)  0   (0)  1   (50)  8   (47.1) 2   (100.0)  1   (50)  1   (50) 12 (70.6) 1   (9) 1   (100) 9   (75) 1   (8) 1   (8)  2   (17) 4   (33) 6   (55)  3   (27)  0   (0)  1   (100)  2   (8) 1   (100)  1   (100)  0   (0) 9   (75) 1 (17) 0 (0) 2 (29) 4 (57) 1 (14)  3 (43) 1 (14) 1 (17)  2 (33)  0 (0)  0 (0)  6 (100) 1 (100)  0 (0)  1 (100) 3 (43)  154 Appendix I: Pearson correlations in combined sample   *p < .05, **p < .01Variables 1  2  3  4  5   6  7 8  9 10   11 12 1.  Patient age   -               2.  Years patient has lived in Canada  .43**    -           3.  Caregiver age   .41**  .51**   -          4.  Years caregiver has lived in Canada   .37**  .80**  .76**   -         5.  Months of caregiving   .12  .21  .16  .08   -        6.  Patient distress  -.34* -.12 -.23 -.08 -.05   -       7.  Caregiver distress  -.14 -.12 -.10 -.17  .09  .09   -      8.  QOL total   .30*  .01  .11 -.07  .10 -.80** -.10   -     9.  Physical well-being   .11 -.01  .05 -.09  .15 -.62**  .11 .81**   -    10.  Social well-being   .32*  .11  .19  .14  .02 -.54** -.18 .66** .24   -   11.  Emotional well-being   .45**  .03  .12 -.12  .08 -.76** -.06 .85** .66** .41**   -  12.  Functional well-being   .14 -.06  .01 -.12  .04 -.66** -.23 .88** .59** .59** .62**   -  155 Appendix J: Pearson correlations in Chinese-speaking group  *p < .05, **p < .01 Variables 1  2  3  4  5   6  7 8  9 10   11 12 1.  Patient age   -               2.  Years patient has lived in Canada  .43   -            3.  Caregiver age    .22  -.11   -          4.  Years caregiver has lived in Canada   .42*  .41   .50   -         5.  Months of caregiving   .09  .23  .17   .11   -        6.  Patient distress  -.16 -.14 -.16 -.001  .003   -       7.  Caregiver distress   .07 -.39 -.08 -.24 -.17  .08   -      8.  QOL total   .19  .04   .20  .01  .10 -.75** -.11   -     9.  Physical well-being  -.11 -.19   .08 -.16 -.02 -.57** -.03 .83**   -    10.  Social well-being   .32  .18   .23  .16  .10 -.57** -.13 .79** .46*   -   11.  Emotional well-being   .34  .09   .26  .08  .07 -.74** -.04 .86** .66** .54**   -  12.  Functional well-being   .14  .09   .12 -.02  .18 -.68** -.18 .93** .69** .72** .75**   -  156 Appendix K: Pearson correlations in Anglophone group *p < .05, **p < .01 Variables 1  2  3  4  5   6  7 8  9 10   11 12 1.  Patient age   -               2.  Years patient has lived in Canada  .70**   -            3.  Caregiver age    .63**   .35   -          4.  Years caregiver has lived in Canada   .60**  .39  .84**   -         5.  Months of caregiving   .17  .48  .17  .14   -        6.  Patient distress  -.48* -.34 -.43* -.35 -.13   -       7.  Caregiver distress  -.32 -.08 -.17 -.33  .43 0.10   -      8.  QOL total   .43*  .34 ¤   .25  .14  .11 -.86** -.10   -     9.  Physical well-being   .28  .34  .20  .12  .36 -.66**   .22  .79**   -    10.  Social well-being   .31  .03  .12  .13 -.11 -.54**  -.26 .56* .08   -   11.  Emotional well-being   .59**  .58**  .39  .26  .13 -.79**  -.09 .85** .64**  .39   -  12.  Functional well-being   .16  .04  .05 -.06 -.14 -.64**  -.29 .83** .50* .50* .52**   -  157 Appendix L Bivariate regression coefficients in combined sample Variable QOL total Physical  well-being Social  well-being Emotional well-being Functional well-being Male patient  (vs female) a  .04   -.12   .05  .16  .06 Male family caregiver  (vs female) b -.04   -.07   .15 -.13 -.03 Breast cancer  (vs esophagus, lung, lymphoma or nasopharyngeal) c  .09 -.001  -.04 -.21  .06 Bladder or colorectal cancer  (vs Esophagus, lung, lymphoma or nasopharyngeal) c -.09   .19   .10  .12 -.13 Ovarian, prostate or uterine cancer  (vs esophagus, lung, lymphoma or nasopharyngeal) c -.06   .02   .01 -.21  .06 Stage I cancer  (vs Stage II cancer) d  .09   .07   .10  .06  .06 Stage III cancer   (vs Stage II cancer) d -.05  -.09 -.02  .08 -.13 Patient with high school education  (vs post-secondary education) e -.10  -.13 -.36*  .06  .06 Patient with university or advanced degree  (vs post-secondary education) e -.12  -.18 -.15 -.14  .09 Family caregiver with high school education  (vs post-secondary education) f  .05   .02  .18 -.04  .02 Family caregiver with university or advanced degree  (vs post-secondary education) f  .02  -.12  .08  .07  .06 Patient with advanced English fluency  (vs. beginner or intermediate) g -.08  -.08  .11 -.22 -.01 Family caregiver with advanced English fluency  (vs. beginner or intermediate) h -.21  -.25 -.01 -.23 -.16 Family caregiver being a spouse/partner of the patient  (vs. extended family/relatives or close friends) i -.03  -.01  .23 -.08 -.19 Family caregiver being a child of the patient  (vs. extended family/relatives or close friends) i -.05   .09 -.09  .004 -.17 More than 1 treatment type  (vs. chemotherapy, drugs, radiation or surgery) j  .06   .12  .03 -.04  .05 Note. a = standardized regression coefficient of binary variable for sex of patient. b = standardized regression coefficient of binary variable for sex of family caregiver. c = standardized regression coefficients of dummy variables for cancer site. d = standardized regression coefficients of dummy variables for cancer stage. e = standardized regression coefficients of dummy variables for education level of patient. f = standardized regression coefficients of dummy variables for education level of family caregiver. g = standardized regression coefficient of binary variable for English fluency of patient. h = standardized regression coefficient of binary variable for English fluency of family caregiver. i = standardized regression coefficients of dummy variables for relationship type. j = standardized regression coefficient of binary variable for treatment type. *p < .05 158 Appendix M: Bivariate regression coefficients in Chinese-speaking group Variable QOL total Physical  well-being Social  well-being Emotional well-being Functional well-being Male patient  (vs female) a -.07  -.10  -.14  .12 -.12 Male family caregiver  (vs female) b -.05  -.08  -.02  .01 -.06 Breast cancer  (vs esophagus, lung, lymphoma or nasopharyngeal) c  .06   .04   .00 -.28 -.01 Bladder or colorectal cancer  (vs Esophagus, lung, lymphoma or nasopharyngeal) c -.07   .09   .20 -.04 -.02 Ovarian, prostate or uterine cancer  (vs esophagus, lung, lymphoma or nasopharyngeal) c -.09  -.10   .07 -.14  .03 Stage I cancer  (vs Stage II cancer) d  .05   .04   .11 -.08  .09 Stage III cancer   (vs Stage II cancer) d -.23 -.37* -.04 -.09 -.18 Patient with high school education  (vs post-secondary education) e -.27 -.14 -.53* -.13 -.15 Patient with university or advanced degree  (vs post-secondary education) e -.12 -.11 -.24 -.12  .04 Family caregiver with high school education  (vs post-secondary education) f  .01  .12   .20 -.13 -.14 Family caregiver with university or advanced degree  (vs post-secondary education) f  .15  .19   .21  .90  .004 Patient with advanced English fluency  (vs. beginner or intermediate) g  .09  .03   .13 -.04  .19 Family caregiver with advanced English fluency  (vs. beginner or intermediate) h -.19 -.25  -.08 -.13 -.14 Family caregiver being a spouse/partner of the patient  (vs. extended family/relatives or close friends) i  .05  .17   .10  .05 -.15 Family caregiver being a child of the patient  (vs. extended family/relatives or close friends) i -.18 -.03  -.19 -.16 -.22 More than 1 treatment type  (vs. chemotherapy, drugs, radiation or surgery) j  .26  .29   .11  .07  .33 Note. a = standardized regression coefficient of binary variable for sex of patient. b = standardized regression coefficient of binary variable for sex of family caregiver. c = standardized regression coefficients of dummy variables for cancer site. d = standardized regression coefficients of dummy variables for cancer stage. e = standardized regression coefficients of dummy variables for education level of patient. f = standardized regression coefficients of dummy variables for education level of family caregiver. g = standardized regression coefficient of binary variable for English fluency of patient. h = standardized regression coefficient of binary variable for English fluency of family caregiver. i = standardized regression coefficients of dummy variables for relationship type. j = standardized regression coefficient of binary variable for treatment type. *p < .05 159 Appendix N: Bivariate regression coefficients in Anglophone group Variable  QOL total Physical  well-being Social  well-being Emotional well-being Functional well-being Male patient  (vs female) a  .13 -.15  .22  .18  .22 Male family caregiver  (vs female) b  .02 -.03  .30 -.19  .03 Breast cancer  (vs esophagus, lung, lymphoma or nasopharyngeal) c  .11 -.11 -.05 -.23 -.12 Bladder or colorectal cancer  (vs Esophagus, lung, lymphoma or nasopharyngeal) c -.16  .26  .02  .24 -.18 Ovarian, prostate or uterine cancer  (vs esophagus, lung, lymphoma or nasopharyngeal) c -.04  .10  .08 -.28 -.03 Stage I cancer  (vs Stage II cancer) d  .13  .09  .09  .17  .05 Stage III cancer   (vs Stage II cancer) d  .09  .18  .04  .17 -.11 Patient with high school education  (vs post-secondary education) e -.01 -.20 -.2  .13  .21 Patient with university or advanced degree  (vs post-secondary education) e -.22 -.30 -.11 -.32  .03 Family caregiver with high school education  (vs post-secondary education) f  .04 -.12  .37 -.22  .17 Family caregiver with university or advanced degree  (vs post-secondary education) f -.11 -.41  .01  .00  .09 Patient with advanced English fluency  (vs. beginner or intermediate) g     -    -     -    -     - Family caregiver with advanced English fluency  (vs. beginner or intermediate) h -.19 -.21  .09 -.19 -.25 Family caregiver being a spouse/partner of the patient  (vs. extended family/relatives or close friends) i -.06 -.16  .39 -.11 -.20 Family caregiver being a child of the patient  (vs. extended family/relatives or close friends) i  .10  .35 -.05  .17 -.20 More than 1 treatment type  (vs. chemotherapy, drugs, radiation or surgery) j -.11 -.01  .01 -.16 -.17 Note. a = standardized regression coefficient of binary variable for sex of patient. b = standardized regression coefficient of binary variable for sex of family caregiver. c = standardized regression coefficients of dummy variables for cancer site. d = standardized regression coefficients of dummy variables for cancer stage. e = standardized regression coefficients of dummy variables for education level of patient. f = standardized regression coefficients of dummy variables for education level of family caregiver. g = standardized regression coefficient of binary variable for English fluency of patient is constant. h = standardized regression coefficient of binary variable for English fluency of family caregiver. i = standardized regression coefficients of dummy variables for relationship type. j = standardized regression coefficient of binary variable for treatment type. *p < .0 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0319081/manifest

Comment

Related Items