Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An evaluation of patients’ preferences and health-related quality of life in asthma McTaggart-Cowan, Helen Ming 2006

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

Item Metadata

Download

Media
831-ubc_2006-0578.pdf [ 7.07MB ]
Metadata
JSON: 831-1.0092726.json
JSON-LD: 831-1.0092726-ld.json
RDF/XML (Pretty): 831-1.0092726-rdf.xml
RDF/JSON: 831-1.0092726-rdf.json
Turtle: 831-1.0092726-turtle.txt
N-Triples: 831-1.0092726-rdf-ntriples.txt
Original Record: 831-1.0092726-source.json
Full Text
831-1.0092726-fulltext.txt
Citation
831-1.0092726.ris

Full Text

A N E V A L U A T I O N O F PATIENTS' P R E F E R E N C E S AND H E A L T H - R E L A T E D Q U A L I T Y O F L I F E IN A S T H M A by H E L E N M I N G M C T A G G A R T - C O W A N B . S c , The University of Victoria, 1999 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R OF S C I E N C E in . ' T H E F A C U L T Y OF G R A D U A T E S T U D I E S (Health Care and Epidemiology) T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A September 2006 © Helen Ming McTaggart-Cowan, 2006 A B S T R A C T Objectives: The objectives of this thesis were i) to quantify patients' preferences for asthma treatments using a discrete choice experiment (DCE), ii) to evaluate the socioeconomic status (SES) impacts on these preferences, and iii) to assess the construct validity of three preference-based instruments (Health Utility Index Mark 3, EuroQol, and Short Form 6D). Methods: One hundred fifty-seven asthma patients between 19-49 years of age residing in metropolitan Vancouver, British Columbia participated in this cross-sectional study. The patients responded to three preference-based instruments, two disease-specific instruments (standardized version of the Asthma Quality of Life Questionnaire and Asthma Control Questionnaire (ACQ)), and a DCE. The DCE was designed to measure preferences for treatment benefit (symptom-free days (SFDs)), potential risks (oral thrush and tremor/heart palpitation), ease of use (frequency of daily administration and number of inhalers required), and cost. Information regarding the patient's SES, pulmonary function, asthma medication use, and self-reported asthma control were also obtained. Results: Relationships between the relative preferences and all treatment attributes were generally in the hypothesized directions. Specifically, the patients were willing to pay an additional $14 per month to receive one extra SFD. Patients were willing to pay $26, $79, and $112 to avoid one, two, and three episodes of oral thrush, respectively, and were willing to forego 1.8, 5.5, 7.8 monthly SFDs to avoid one, two, and three episodes of oral thrush, respectively. Annual income and education level affected treatment preferences. Furthermore, the preference-based instruments were able to discriminate across levels of asthma control using the ACQ; however, there was a lack of discrimination between HRQL and asthma control using subjective measures, such as magnitude of short-acting p-agonist use and self-reported control status. Conclusions: The DCE results revealed that patients preferred treatments with more SFDs but they were willing to forego symptom relief to avoid greater frequencies of adverse events. The results demonstrated the construct validity of the preference-based instruments such that they were able to discriminate across the A C Q scores, providing evidence that preference-based instruments could detect minimal changes in asthma states. T A B L E O F C O N T E N T S A B S T R A C T ii T A B L E O F C O N T E N T S iii LIST O F T A B L E S vi LIST OF FIGURES vii LIST O F N O M E N C L A T U R E viii A C K N O W L E D G E M E N T S ix CO-AUTHORSHIP S T A T E M E N T x C H A P T E R 1: INTRODUCTION 1 1.1 Asthma 1 1.1.1 Asthma Prevalence 2 1.1.2 Economics of Asthma 2 1.2 Asthma Management 3 1.2.1 Outcomes of Inadequate Asthma Management 4 1.2.2 Etiologies of Inadequate Asthma Management 5 /. 2.3 Valuation of Health States in Asthma 6 1.3 Research Needs and Justification 7 1.4 Thesis Objectives and Organization 7 1.5 Summary 9 1.6 References 10 C H A P T E R 2: B A C K G R O U N D 14 2.1 Health Economics 14 2.2 Discrete Choice Experiment 15 2.2.1 Theoretical Background 16 2.2.2 Experimental Design 17 2.2.3 Methodological Challenges 22 2.2.4 An Evaluation of Patients' Preferences in Asthma 24 2.3 Measures of Utility 25 2.3.1 Quality-Adjusted Life Year 25 2.3.2 Direct Methods 26 2.3.3 Indirect Methods 27 2.3.4 An Evaluation of Health-Related Quality of Life in Asthma 29 2.4 Summary 29 2.5 References 31 - iii -C H A P T E R 3: PATIENTS' P R E F E R E N C E S IN T H E T R E A T M E N T O F A S T H M A : A R E V I E W O F C U R R E N T E V I D E N C E 35 3.1 Introduction 35 3.2 Methods 36 3.3 Results 36 3.3.1 Treatment Preferences 37 3.3.2 Asthma Management 40 3.3.3 Decision-making 41 3.3.4 Symptom Preferences 42 3.4 Discussion 42 3.5 Tables 46 3.6 References 50 C H A P T E R 4: E V A L U A T I N G PATIENTS' P R E F E R E N C E S IN A S T H M A T H E R A P Y USING A D I S C R E T E C H O I C E E X P E R I M E N T 54 4.1 Introduction 54 4.2 Methods 55 4.2.1 Study Population 55 4.2.2 Discrete Choice Experiment 55 4.2.3 Patient Information 56 4.2.4 Statistical Analysis 57 4.3 Results 59 4.3.1 Sample Characteristics 59 4.3.2 Unadjusted Model 60 4.3.3 Sub-population Preferences 61 4.4 Discussion 61 4.5 Tables 65 4.6 Figures 72 4.7 References 73 C H A P T E R 5: A C O M P A R I S O N O F H E A L T H - R E L A T E D Q U A L I T Y O F L I F E M E A S U R E S IN A S T H M A 76 5.1 Introduction 76 5.2 Methods 78 5.2.1 Study Population 78 5.2.2 Data Collection 78 5.2.3 Analysis 80 5.3 Results 81 5.3.1 Sample 81 5.3.2 Description of Global Utilities 82 5.3.3 Construct Validity 82 5.4 Discussion 83 5.5 Tables 86 5.6 Figures 89 5.7 References 92 - iv -C H A P T E R 6: S U M M A R Y , CONTRIBUTIONS, AND R E C O M M E N D A T I O N S 95 6.1 Summary of Key Research Findings 95 6.2 Unique Contributions, Impact, and Implications 98 6.3 Study Limitations 99 6.4 Recommendations 100 6.5 Conclusions 102 6.5 References 103 APPENDICES 104 Appendix I Copies of Approved Ethics Certificate 104 Appendix II Consent Form 109 Appendix III Discrete Choice Experiment 116 Appendix I V Asthma Assessment Questionnaire 133 Appendix V Standardized Version of the Asthma Quality of Life Questionnaire 151 Appendix V I Asthma Control Questionnaire 157 LIST O F T A B L E S Table 3.1: Characteristics of treatment preference studies in asthma 46 Table 3.2: Characteristics of decision-making, delivery of care, and symptom preference studies in asthma 47 Table 3.3: Significant findings from treatment preferences studies 48 Table 3.4: Significant findings from delivery of asthma care studies 49 Table 3.5: Significant findings from decision-making studies 49 Table 4.1: Attributes and levels used in the study 65 Table 4.2: Demographic and asthma-specific information related to the participants in the survey 66 Table 4.3: Patient characteristics affecting the preference to not have their asthma treated.... 67 Table 4.4: Relative preferences and marginal rates of substitution for the nested logit model 68 Table 4.5: Willingness-to-pay estimates for the stratified income model 69 Table 4.6: Marginal rates of substitution for the stratified education model 70 Table 4.7: Willingness-to-pay estimates for the stratified short-acting P-agonist model 71 Table 5.1: Characteristics of study population 86 Table 5.2: Relationship between asthma control and the H R Q L instruments 87 Table 5.3: Correlations (Spearman's rho) for the instruments 88 - v i -LIST OF FIGURES Figure 4.1: Example of a choice set 72 Figure 5.1: Distribution of global utility values across the preference-based instruments ....89 Figure 5.2: Frequency of global utility values across the preference-based instruments 90 Figure 5.3: Relationship between A C Q and preference-based instruments 91 - v i i -LIST O F N O M E N C L A T U R E A C Q Asthma Control Questionnaire A N O V A analysis of variance A P I Autonomy Preference Index A Q L - 5 D preference-based Asthma Quality of Life Questionnaire A Q L Q Asthma Quality of Life Questionnaire A Q L Q ( S ) standardized version of the Asthma Quality of Life Questionnaire B C British Columbia CI confidence interval CPS Control Preference Scale D C E discrete choice experiment DPI dry powder inhaler EQ-5D EuroQol® F E V , forced expired volume in one second H R Q L health-related quality of life H U I Health Util i ty Index® ICS inhaled corticosteroid IIA independence of irrelevant alternatives M D I metered dose inhaler M e S H medical subject heading M N L • multinominal logit M R S marginal rate of substitution N L M nested logit model OR odds ratio Q A L Y quality-adjusted life years Q O L quality of life PEF peak expiratory flow S A short-acting S B L separate binary logit SES socioeconomic status SF-36 Short Form 36 SF-6D Short Form 6D SFD symptom-free day S G standard gamble T T O time trade-off V A S visual analogue scale W T P willingness-to-pay - Vll l -A C K N O W L E D G E M E N T S I am extremely grateful to my graduate advisors, Dr. Larry Lynd and Dr. Aslam Anis for their invaluable mentorship. They continuously provided guidance and great opportunities to allow me to succeed in all my research endeavours. I would like to especially thank Larry for taking me on as his inaugural graduate student and his careful review of all my work. I would also like to thank my committee members, Dr. Mark FitzGerald and Dr. Jacek Kopec, for all their insightful comments and their genuine interest in my work. Thank you to the Health Economics Group, both past and current members, for sharing their knowledge with me and making my graduate life very enjoyable. A special thank you is in order to Dr. Peilin Shi for providing statistical support during the entire analysis process. Many thanks to the respiratory staff at the University of British Respiratory Clinic, Pacific Lung Centre, and the Surrey Memorial Respiratory Services who facilitated in the data collection phase, especially Linda Hui, Tanja Teofilovic, Bev Beaudin, and Louella Markortoff. I am also deeply indebted to the study participants who volunteered their time. Their eagerness to participate to aid research in asthma care was an inspiration to me. Finally, I want to thank Gord for his endless encouragement, patience, and support. His constant commitment to strive for the best and his attention to detail provided me with the motivation and determination to complete this work to the highest level. To him, I dedicate this thesis. The work presented in this thesis is generously supported by the Michael Smith Foundation for Health Research and the British Columbia Lung Association. - ix -CO-AUTHORSHIP S T A T E M E N T The work presented in this thesis was conducted and disseminated by the Master's candidate. The co-authors of the manuscripts that comprise part of this thesis made contributions only as is commensurate with a thesis committee or as experts in a specific area as it pertains to the work. The co-authors provided direction and support. The co-authors reviewed each manuscript prior to submission for publication and offered critical evaluations; however, the candidate was responsible for the writing and the final content of these manuscripts. - x -C H A P T E R 1 INTRODUCTION 1.1 Asthma Chronic respiratory diseases represent a significant challenge to public health due to their rising prevalence and corresponding economic impacts.[1] Asthma, in particular, is one of the most widespread chronic conditions affecting Canadians. [2] Due to the increasing prevalence of asthma since the 1980s, there has been extensive research in this area, leading to the publication of management guidelines and the development of novel therapies. Despite advances in the understanding of asthma and the availability of effective treatments, asthma-related morbidity in Canada is still a significant concern. [3] Asthma is a chronic inflammatory pulmonary disorder characterized by a reversible obstruction of the airways. As there is no widely accepted standard definition, evaluating the worldwide burden of asthma is problematic. The Canadian Asthma Consensus Group provides the following definition [4]: "Asthma is characterized by paroxysmal or persistent symptoms such as dyspnea, chest tightness, wheezing, sputum production, and cough, associated with variable airflow limitation and a variable degree of hyperresponsiveness of airways to endogenous or exogenous stimuli. " This definition tends to be more abstract than operational in identifying asthma. Because of the similarity of the symptoms of asthma with those of other respiratory diseases, as well as the lack of a reliable clinical test, asthma is often misdiagnosed.[3] Previous asthma studies have used patient self-report of physician diagnosis, in conjunction with evidence of disease-related symptoms (i.e. wheezing, shortness of breath) as evidence of asthma; however, these methods do not assess the variability in airflow obstruction, obtained through measurements of the forced expired volume in the first second (FEV|) or peak expiratory flow (PEF), or even the hyperresponsiveness to methacholine inhalation challenges.[4] Due to the subjectivity involved with identifying this disorder, challenges in determining the true prevalence and burden of asthma are encountered. 1.1.1 Prevalence of Asthma According to the World Health Organization, over 300 mil l ion people worldwide are afflicted with asthma. [5] The global incidence of asthma is also on the rise [6]; it is predicted that an additional 100 mil l ion people wi l l suffer from asthma by 2025.[1] Asthma prevalence tends to vary with age: 3-6% in adults and 10-11% in children.[2] Currently, 2.2 million Canadians over the age of 12 years are diagnosed with asthma. [7] The Canadian prevalence of asthma among adults has increased significantly from 2.3% in 1979 to 7.5% in 1998. [2] In 2004, 8.4% of the Canadian population was diagnosed with asthma,' one of the highest incidence rates in the world.[8] The rise in disease prevalence may be due to the 'hygiene hypothesis', which theorizes that the lack of infections and unhygienic contact may result in the development of allergic illnesses. [9] Specifically, the increased exposures to outdoor air pollution and indoor airborne allergens, and the increased contact with dust mites and animal dander, may result in altered host susceptibility and allergic sensitivity, respectively. [10] Furthermore, viral infections and subsequent antibiotic administrations, as well as the greater global consumption of fast food, may play a role in triggering asthma in children. [11,12] 1.1.2 Economics of Asthma Due to the high prevalence and chronic nature of asthma, a major economic burden arises at both the individual and societal levels. Missed work, decreased productivity, and adverse psychological effects significantly influence the overall well-being of an individual with asthma.[13] Economic demands, because of increased treatment and physician costs, emergency room visits, and hospitalization due to asthma, present major challenges to the healthcare system and, ultimately, society at large. In 1990, the total Canadian asthma-related expenditures were between $504 and $648 mill ion in 1990 Canadian dollars.[14] The cost of asthma was similar to the cost of all infectious diseases ($772 million), hematological diseases ($486 million), perinatal illnesses ($532 million), and congenital defects ($673 million), but was considerably lower than the cost of cardiovascular diseases ($7.7 billion) and cancers ($11.2 billion).[14] In 1998, the economic burden of asthma in the United States was estimated to be $12.7 billion in 1998 US dollars.[l] Due to their high unit costs and widespread use, prescription P-agonists and anti-inflammatory drugs comprise the largest component of the direct costs of asthma management, followed by hospital inpatient care and physician services. Other direct costs, relating to outpatient services including nurses, and ambulance and emergency services, account for relatively small proportions of the aggregate cost. Asthma-related morbidity and productivity loss are major contributors to the indirect costs. The cost of an individual's asthma is dependent on disease severity.[15] One study found that the average annual cost of asthma from a societal perspective in south-central Ontario was $2,549 per person; however, when adjusted for disease severity, the costs were $1,617, $2,218, and $3,905 for the management of mild, moderate, and severe asthma patients, respectively.[13] Patients with difficult-to-treat or inadequately controlled asthma consume a disproportionate share of asthma healthcare resources, since these individuals have a greater need for unscheduled medical services, emergency department visits, and hospitalizations. [ 16] 1.2 Asthma Management In 1995, the Canadian Asthma Consensus Report was published,[17] with revised versions published more recently [4,18-20]; it is currently the national guideline for asthma management. With appropriate management, nearly all people with asthma should be able to achieve almost complete control and prevention of asthma-related symptoms. Thus, people with asthma should be able to live a nearly normal lifestyle with a good health-related quality of life ( H R Q L ) and a high level of productivity. Current asthma management guidelines propose a multifaceted, stepped-care approach, dependent upon the symptom frequency and disease severity of the individual.[17,21,22] This method improves not only therapeutic management but also environmental control and education awareness. Initial pharmacologic management in patients with mild, intermittent, or exercise-induced symptoms entails the use of short-acting (SA) P-agonists to control symptoms as they occur.[4] As the frequency of asthma symptoms increases, indicative of worsening disease or deteriorating control, low-dose inhaled corticosteroids (ICS) are administered. The dose of ICS is gradually increased until the symptoms are controlled and the use of S A P-agonists is minimized to fewer than four doses per week or fewer than two canisters per year. However, i f symptoms are still persistent in spite of low to moderate doses of an ICS, additional therapies such as long-acting P-agonists or, less commonly, leukotriene receptor antagonists, are required for asthma control. 1.2.1 Outcomes of Inadequate Asthma Management Despite the dissemination of asthma management guidelines and the availability of effective therapies, there still remains a significant number of inadequately managed patients. [23-27] Poor control of asthma symptoms can lead to adverse clinical and economic outcomes.[1] The rate of asthma-related mortality increased from the 1970s to the mid-1980s, especially in the 15-24 and the over-65 age groups.[2] Although the annual asthma-related mortality rate has decreased in Canada since the 1990s, there are still approximately 20 children and 500 adults whose deaths are directly attributed to asthma each year.[2] It is estimated that greater than 80% of asthma-related deaths could be prevented with appropriate management. Previous work in assessing asthma management and outcomes has demonstrated an over reliance on S A P-agonists, resulting in a diminished Q O L and an increased risk for asthma-related morbidity and mortality.[28-30] Anis et al. determined that individuals with uncontrolled asthma tend to have more physician visits and hospital and emergency admissions compared to those with better controlled asthma.[23] A three-year longitudinal analysis of prescription claims data in British Columbia following the publication of the management guidelines in 1995 did not show any decrease in use of S A p-agonist but, unexpectedly, did show a decline in the utilization of ICS.[24] Nearly 13% of the study population filled prescriptions for nine or more canisters of S A P-agonists in 1995, more than twice the recommended annual usage. This excessive use of S A P-agonists puts these patients at greater risk for fatal or near-fatal asthma exacerbations. In a follow-up study, Lynd et al. discovered a strong negative association between socioeconomic status (SES) and the magnitude of S A P-agonist use, independent of disease severity.[25] Individuals receiving social assistance were more likely to use greater amounts of S A p-agonist (odds ratio (OR) 3.4, 95% confidence interval (CI) 1.7 to 6.5). After adjusting for severity', both annual household income and level of education completed were significantly and negatively associated with the magnitude of S A (3-agonist use. Consistent with these findings, they also found an inverse relationship between S A p-agonist use and annual household income (greater than $50,000: O R 0.28, 95% CI 0.13 to 0.60; $20,000-$50,000: O R 0.44, 95% CI 0.21 to 0.96 - both cases relative to incomes of less than $20,000). Also, an inverse relationship between S A P-agonist and education was observed for having a university degree relative to not receiving a high school diploma (OR 0.25, 95% CI 0.14 to 0.71). Various studies have demonstrated the cost-effectiveness of administering ICS to people with asthma. [26,31] Hospitalization is an important outcome in asthma patients because it often reflects the presence of severe, uncontrolled, or progressive disease; the use of ICS has been shown to prevent asthma hospitalization over the long term. [26] In addition, hospitalization for asthma is a significant predictor of asthma-related mortality, as well as of subsequent re-hospitalizations.[28,29] In a nested case-control study of 30,569 subjects, Suissa et al. demonstrated that the regular use of an ICS was associated with a 31% reduction in the rate of hospital admissions for asthma and a 39% reduction in the rate of re-admission. [26] Regular ICS use could potentially prevent approximately five hospital admissions per 100 asthma patients per year. O f patients who have previously been admitted for asthma, the use of ICS prevented 27 re-admissions per 100 asthma patients. Previous work has demonstrated that lower SES asthma patients experience more frequent hospital admissions, emergency room visits, and family physician visits.[32-35] 1.2.2 Etiologies of Inadequate Asthma Management Findings of previous studies have shown that inadequate asthma management is still a significant problem [23-30]; however, the reasons for this are still unclear. Factors relating to the physician, the healthcare system, and the patient may affect this inadequate management but the extent of these factors is unclear. [36] These proposed etiologies may potentially interact in a complex and dynamic manner. Physicians' poor knowledge regarding asthma management strategies may contribute to a patient's poor asthma control. Inappropriate prescribing patterns may result from the physician's lack of awareness for the management guidelines. In addition, the availability of drug therapies, treatment cost, and medical insurance coverage may play significant roles. If the treatment is not readily accessible or easy to administer, or i f medical insurance wi l l not cover the majority o f treatment cost, this may also contribute to the reluctance of the physician to prescribe the appropriate treatment because of the realization that the patient may be non-compliant to this treatment regimen. In addition to their non-compliance to appropriately prescribed management strategies, patients may just choose to rely on S A (3-agonists. There is some evidence of an 'addiction'-like phenomenon from the (3-agonist-related side effects that may lead some individuals to use greater than the recommended doses even in the absence of symptoms.[37] Due to the lack of knowledge regarding appropriate asthma management, and behavioural or social factors, patients' preferences may also affect treatment decisions. Patients' Preferences There is increasing evidence that providing services responsive to the preferences of patients improves treatment adherence, asthma control, and Q O L [38,39]; however, it is unclear i f patients' preferences for asthma management strategies are considered in the clinical decision-making process. Decisions regarding treatments for patients generally follow guidelines based on evidence from clinical trials. These guidelines advocate that the management of chronic diseases, such as asthma, should incorporate a joint decision-making process between the physician and the patient. [40] However, previous studies investigating patient autonomy for making decisions regarding asthma treatments suggest that patients generally do not want a dominant role in selecting their control strategy.[41-43] Currently, there is a lack of evidence in the literature evaluating patients' preferences for asthma treatments. Therefore, there is a need to evaluate patients' preferences for individual aspects of asthma drug therapy to develop a better understanding about what patients prefer and how their preferences may differ. This may provide insight into why patients choose treatments and, ultimately, may improve their compliance with their treatments. 1.2.3 Valuation of Health States in Asthma With the increasing prevalence of both asthma and inappropriate management, there is an ongoing need to improve asthma control and the cost-effectiveness of asthma interventions. Given the budget constraints imposed on healthcare, it is vital that all asthma therapies undergo appropriate economic analyses to ensure that the most cost-effective treatment is used. Thus, it is important to have a valid instrument which can assess the impacts of differing levels of asthma control on an individual level. [44] Given that the primary goal of asthma management is to achieve optimal control and to improve the HRQL of an individual,[45,46] there is a need for an instrument capable both of discriminating across all levels of asthma control and of comparing HRQLs between different diseases. 1.3 Research Needs and Justification The continued inadequate asthma management places a burden on both the afflicted individual and society. Reduced HRQL and impaired productivity are common among poorly-managed asthma patients. This results in an increase in direct costs arising from higher medication use and healthcare utilization, as well as an increase in the indirect costs resulting from work and school absenteeism. It is likely that patients' preferences for different treatment attributes play a significant role in how well patients adhere to their treatment strategy. However, it is unclear whether patients have different preferences regarding treatment-related benefits, risks, and costs. Specifically, the level of treatment benefit a patient is willing to forego in exchange for a better adverse event profile has not been evaluated. In addition, socioeconomic factors figure prominently in overall health and in asthma control. While it is believed that SES can influence preferences, this has not been quantified in asthma. Furthermore, there is a need for a greater understanding of health states in people with asthma. This, in turn, may lead to an improvement in the individuals' HRQL and result ultimately in lower proportions of patients with suboptimal asthma control. Currently, there is an absence of empirical evidence on whether preference-based instruments yielding HRQL values are able to discriminate across levels of asthma control. 1.4 Thesis Objectives, Hypotheses, and Organization This thesis had two overall objectives: an evaluation of the preferences patients have towards the attributes of different asthma therapies and an assessment of the validity of preference-based instruments in predicting the asthma control status of the patients. The specific objectives of this thesis were: 1. To quantify, using a discrete choice experiment, patients' preferences for different aspects of asthma therapy, including treatment efficacy, potential adverse events, ease of use, and cost; 2. To evaluate the influences of SES on the patients' preferences for treatment attributes; and 3. To examine the cross-sectional construct validity of three indirect preference-based instruments (EuroQol, Health Utility Index Mark 3, and Short Form 6D), in terms of their ability to discriminate across levels of asthma control. A gradient was hypothesized to exist among patients' preferences for the different treatment attributes. Specifically, individuals were expected to prefer a treatment that will result in more symptom-free days as opposed to fewer, and fewer episodes of a specific adverse event as opposed to more. It was also proposed that a social gradient will be present for the preference for asthma treatments, such that people of lower social class, in terms of income and education levels, will be less likely to pay greater amounts to receive more treatment benefits or fewer adverse events. In addition, it was assumed that preference-based instruments will be able to discriminate across validated measures of asthma control, so that poorly-controlled individuals will have corresponding poor QOL scores. The thesis is comprised of six chapters. This first chapter provides a brief introduction to asthma, and discusses the economic burden and health outcomes related to inappropriate asthma management. This chapter also provides insights into the motivation and justification for this thesis. Chapter 2 introduces health economics, as well as descriptions of the discrete choice experiment method to elicit patient preferences and the preference-based HRQL instruments to value health states. Chapter 3 identifies the gap in the current literature with a systematic review of published studies investigating patient preferences in asthma treatment. Chapters 4 and 5 present the empirical results, namely the findings from the evaluation of the patients' preferences for asthma therapy and the validity assessment of preference-based instruments, respectively. The final chapter provides a summary of the research findings and outlines the strengths, limitations, and potential impacts of the study findings. Suggestions for future work are also addressed. Chapters 3, 4, and 5 are stand-alone manuscripts, which are either in the process of being submitted for, or currently under review by, major peer-reviewed journals. The work presented in this thesis was conducted solely by the Master's candidate. 1.5 Summary The increasing prevalence of asthma combined with inappropriate asthma management places a substantial burden on the healthcare system. Although other potential causes of poor asthma management need to be evaluated, patients' preferences may be one contributing factor. The preferences patients have for immediate symptomatic relief rather than long-term prevention may play a significant role in suboptimal asthma control. Thus, the primary focus of this research was to develop and administer a discrete choice experiment to quantify the preferences patients may have towards specific attributes of asthma drug therapy, including benefits, harms, and costs. The secondary objective was to examine the cross-sectional validity of preference-based instruments, in terms of their ability to discriminate across validated measures of asthma control. 1.6 References 1. Bousquet J, Bousquet PJ, Godard P, Daures J-P. The public health implications of asthma. Bull WHO 2005;83(7):548-54. 2. Health Canada. The burden of asthma and other chronic respiratory diseases in Canada. Ottawa, ON: Health Canada, Health Protection Brach, Respiratory Division 1998. 3. Weiss K B , Gergen PJ, Wagener DK. Breathing better or wheezing worse? The changing epidemiology of asthma morbidity and mortality. Ann Rev Public Health 1993;14:491-513. 4. Boulet L-P, Becker A , Berube D, Beveridge R, Ernst P. Canadian asthma consensus report, 1999. Can Med Assoc J 1999;161(11 Supp):Sl-62. 5. World Health Organization: World Health Report 2003: Shaping the future [Online]. [2003?] . [cited 2006 Apr 24]; Available from: U R L : http://www.who.int/whr/2003/en/index.html 6. Selgrade M K , Lemanske Jr RF, Gilmour MI, Neas L M , Ward MDW, Henneberger PK, et al. Induction of asthma and the environment: what we know and need to know. Environ Health Perspect 2006; 114(4):615-9. 7. The prevention and management of asthma in Canada [Online]. 2000 Oct 19 [cited 2006 Apr 19]; Available from: URL: http://www.phac-aspc.gc.ca/publicat/pma-pca00/pdf/asthma003 .pdf 8. Chen Y , Helen J. Asthma. Health Rep 2004; 16(2):43-6. 9. Strachan DP. Hay fever, hygiene, and household size. B M J 1989;299:1259-60. 10. Gilmour MI, Jaakkola MS, London SJ, Nel A E , Rogers CA. How exposure to environmental tobacco smoke, outdoor air pollutants, and increased pollen burdens influences the incidence of asthma. Environ Health Perspect 2006; 114:627-33. 11. Marra F, Lynd L, Coombes M , Richardson K, Legal M , FitzGerald JM, et al. Does antibiotic exposure during infancy lead to development of asthma? A systematic review and metaanalysis. Chest 2006:29:610-8. 12. Norback D, Zhao ZH, Wang ZH, Wieslander G, M i Y H , Zhang Z. Asthma, eczema, and reports on pollen and cat allergy among pupils in Shanxi province, China. Int Arch Occup Environ Health; 2006 Jul 5;[Epub ahead of print]. 13. Ungar WJ, Coyte PC. Measuring productivity loss days in asthma patients. The Pharmacy Medication Monitoring Program and Advisory Board. Health 2000;9(l):37-46. - 10-14. Krahn M D , Berka C, Langlois P, Detsky AS. Direct and indirect costs of asthma in Canada, 1990. Can Med Assoc J 1996; 154(6):821 -31. 15. Godard P, Chanez P, Siraudin L, Nicoloyannis N , Duru G. Costs of asthma are correlated with severity: a 1-year prospective study. Eur Respir J 2002;19:61-7. 16. Bootman JL, Crown WH, Luskin AT. Clinical and economic effects of suboptimally controlled asthma. Managed Care Interface 2004;17:31-6. 17. Ernst P, FitzGerald JM, Spier S. Canadian Asthma Consensus Conference: summary of recommendations. Can Respir J 1996;3:101-14. 18. Boulet L-P, Bai TR, Becker A, Berube D, Berveridge R, Bowie D M , et al. What is new since the last (1999) Canadian asthma consensus guidelines? Can Respir J 2001;8(Suppl A):5A-27A. 19. Lemiere C, Bai T, Baiter M , Bayliff C, Becker A , Boulet L-P, et al. Adult asthma consensus guidelines update 2003. Can Respir J 2004; 1 l(Suppl A):9A-18A. 20. Becker A , Lemiere C, Berube D, Boulet L-P, Ducharme F M , FitzGerald JM, et al. Summary of recommendations from Canadian asthma consensus guidelines, 2003. Can Med Assoc J 2005;173(6 Suppl):S3-l 1. 21. Global Initiative for Asthma. Pocket guide for asthma management and prevention. Bethesda (MD):-National Institute of Health, Nov 1998. NIH Pub. No. 96-3569B. 22. British Thoracic Society, British Paediatric Association, Royal College of Physicians of London, King's Fund Centre, National Asthma Campaign, Royal College of General Practitioners, et al. Guidelines on the management of asthma. Thorax 1993;48(Suppl2):Sl-24. 23. Anis A , Lynd L, Wang X , King G, Spinelli J, Fitzgerald M , et al. Double trouble: inappropriate asthma medication use linked to increased use of health care resources. Can Med Assoc J 2001; 164(5):625-31. 24. Lynd L D , Guh DP, Pare PD, Anis A H . Patterns of inhaled asthma medication use: a 3-year longitudinal analysis of prescription claims data from British Columbia, Canada. Chest 2002; 122(6): 1973-81. 25. Lynd LD, Sandford AJ , Kelly EA, Pare PD, Bai TR, FitzGerald JM, et al. Reconcilable differences - a cross-sectional study of the relationship between socioeconomic status and the magnitude of short-acting P-agonist use in asthma. Chest 2004; 126(4): 1161-8. 26. Suissa S, Ernst P, Benzyoun S, Baltzan M , Cai B. Low-dose inhaled corticosteroids and the prevention of death from asthma. N Engl J Med 2000;343(5):332-6. - 11 -27. FitzGerald JM, Boulet L-P, Mclvor RA, Zimmerman S, Chapman KR. Asthma control in Canada remains suboptimal: the Reality of Asthma Control (TRAC) study. Can Respir J 2006;13:253-9. 28. Crane J, Pearce N , Flatt A. Prescribed fenoterol and death from asthma in New Zealand, 1981 - 1983. A case-control study. Lancet 1989;i:917-22. 29. Grainger J, Woodman K, Pearce N , Crane J, Burgess C, Keane A, et al. Prescribed fenoterol and death from asthma in New Zealand 1981 - 1987: a further case-control study. Thorax 1991;46:105-11. 30. Pearce N , Granger J, Atkinson M , Crane J, Burgess C, Keane A. Case-control study of prescribed fenoterol and death from asthma in New Zealand. Thorax 1990;45(3):170-5. 31. Sullivan SD, Buxton M , Andersson LF, Lamm CJ, Liljas B, Chen Y Z , et al. Cost-effectiveness analysis of early intervention with budesonide in mild persistent asthma. J Allergy Clin Immunol 2003; 112:1229-36. 32. Watson J, Cowen P, Lewis R. The relationship between asthma admission rates, routes of admission, and socioeconomic deprivation. Eur Respir J 1996;9:2087-93. 33. Goodman DC, Stukel TA, Chang CH. Trends in pediatric asthma hospitalization rates: regional and socioeconomic differences. Pediatrics 1998; 101 (2):208-13. 34. Boulet L-P, Belanger M , Lajoie P. Characteristics of subjects with a high frequency of emergency visits for asthma. Am J Emerg Med 1996;14(7):623-8. 35. Erzen D, Carrier K, Dik N , Mustard C, Roos L, Mangreda J, et al. Income level and asthma prevalence and care patterns. Am J Resp Crit Care Med 1997;155:1060-5. 36. Chapman KR, Ernst P, Grenville A , Dewland P, Zimmerman S. Control of asthma in Canada: failure to achieve guideline targets. Can Respir J 2001;8(Suppl A):35A-40A. 37. Lynd L D . An evaluation of the determinants of asthma management and asthma control: a study of British Columbia asthmatics [dissertation]. Vancouver (BC): Department of Health Care and Epidemiology, University of British Columbia; 2002. 38. Guadagnoli E, Ward P. Patient participation in decision-making. Soc Sci Med 1998;47:329-39. 39. Coulter A . Partnerships with patients: the pros and cons of shared clinical decision making. J Health Serv Res Policy 1997;35:276-81. 40. Coulter A, Entwistle V , Gilbert D. Sharing decisions with patients: is the information good enough? B M J 1999;318:318-22. - 12-41. Gibson PG, Talbot PI, Toneguzzi PC. Self-management, autonomy, and quality of life in asthma. Chest 1995;107:1003-8. 42. Adams RJ, Smith BJ, Ruggin RE. Patient preferences for autonomy in decision-making in asthma management. Thorax 2001;56:126-32. 43. Caress A - L , Beaver K, Luker K, Campbell M , Woodcock A. Involvement in treatment decision:"what do adults with asthma want and what do they get? Results of a cross sectional survey. Thorax 2005;60:199-208. 44. Streiner DL, Norman GR. Health measurement scales: a practical guide to their development and use. 2 n d ed. New York: Oxford University Press; 1995. 45. O'Byrne P M , Bisgaard H, Godard PP, Pistolesi M , Palmqvist M , Zhu Y, et al. Budesonide/Formoterol combination therapy as both maintenance and reliever. Am J Respir Crit Care Med 2005;171:129-36. 46. Bateman ED, Boushey HA, Bousquet J, Busse WW, Clark TJ, Pauwels RA, et al. Can guideline-defined asthma control be achieved? The Gaining Optimal Asthma ControL study. A m J Respir Crit Care Med 2004; 170,836-44. - 13 -C H A P T E R 2 B A C K G R O U N D 2.1 Health Economics Health economics is the discipline that examines choices between costs and consequences within the context of healthcare. It provides a set of tools for use by researchers and policy analysts to assess the risks and benefits of particular resource allocation options and the trade-offs made in selecting among these alternatives. As health has many different definitions, the scope of health economics depends both on the question and on the questioner. Health economics provides methods to improve health treatments in the face of increasing healthcare costs and constraints on public funds. The results can be used to aid decision-makers in choosing the most cost-effective option amongst different alternatives; however, to improve decision-making, the views of stakeholders, including patients, clinicians, and healthcare providers need to be incorporated. [1] Considering a patient's preferences in decision-making for a specific treatment or service has been shown to improve chronic disease management,[2] which in turn can improve the(patient's disease control and overall well-being.[3,4] As will be discussed further in Chapter 3, patients make decisions by understanding the available options, weighing their individual preferences, and then trading-off between the perceived advantages and drawbacks. The simplest approach to elicit these preferences is a crossover study in which patients try various treatments sequentially and then state which treatment they prefer.[5,6] More complex approaches have been derived from psychological theories: one example of such a method is the discrete choice experiment (DCE).[7,8] DCEs, which are being used increasingly in health economics, can investigate a patient's relative preferences for different aspects of care and provide an estimate of the trade-offs between these aspects.[7] One particular strength of using a DCE is its similarity to real life decision-making situations. [9] Understanding individuals' preferences relating to a specific disease is important, but does not provide broad-scale insight into the relative usefulness of different treatments for different health states. Consequently, decision-makers are increasingly demanding a - 14-validated instrument to assess and compare the relative consequences of different interventions for different disease groups in different populations. Physiological measures often provide useful information to healthcare providers; however, they correlate poorly with a patient's functional capacity and well-being.[10] Better instruments incorporate a patient's health-related quality of life (HRQL) as a single index to measure both the patient's well-being and clinical effectiveness for use in economic evaluations. Numerous instruments have been proposed to evaluate HRQL; however, before being used by decision-makers, they need to be validated. [10] Both the understanding of patients' preferences and effectiveness measures are important in optimizing the management of, and treatments for, a specific disease. In this chapter, the theory and methodology for conducting a D C E and for evaluating HRQL will be presented; the focus is on the development and evaluation of such tools for asthma therapy. 2.2 Discrete Choice Experiments A key component for improving the effectiveness of treatment strategies is to ensure that they meet the preferences of patients. While many techniques can be used to obtain this preference information, one of the most effective methods is a DCE. DCEs are derived from the rating and ranking techniques of conjoint analysis, which were originally developed by Luce and Tukey in the early 1960s.[11] Conjoint analysis was used to evaluate the choices made by consumers between brand, price, and other product attributes in market research. However, the term conjoint analysis is now considered obsolete and has been replaced with more specific names such as random utility choice or discrete choice modeling to describe the various ways to model preferences and choices.[12] These descriptive terms distinguish choice-based experiments from other forms of conjoint theory that are not derived from economic random utility theory. Given their strong foundations in economic theory, DCEs explore preferences, assign monetary valuations, and predict the uptake of many new programs and applications.[1,8] In a DCE, stated preferences are obtained from surveys, in which respondents indicate which option they prefer from a series of hypothetical choice sets described in terms of the attributes of interest.[13,14] Then the relative importance of the different attributes and the willingness-to-trade between attributes can be estimated for each - 15 -respondent. Inferences can be made regarding the average preference of that population by using a representative sample of respondents from a population of interest, 2.2.1 Theoretical Background Adopting Lancaster's view of goods and utility, the random utility theory assumes that goods and services can be described by their characteristics, and that a utility function can be developed for an individual based on these characteristics.[15] The individual's decision-making process is modeled as a difference between two indirect utility functions, where each utility function is associated with a different choice. DCE respondents make a series of pairwise choices and choose the scenario that result in a higher level of expected utility. Thus, the individual would choose the intervention A over B if: where U represents the indirect utility function of the respondent, AA and As the characteristics of intervention A and B, respectively, and Z characteristics of the respondent that influences preferences. While DCE respondents are assumed to implicitly know the nature of their utility function, the researcher will not. This assumption introduces the concept of random utility, where an error term is included to reflect the unobservable factors in the utility function for each respondent. Within the random utility framework, the respondent will choose intervention B over A if: where V is the measurable component of utility estimated empirically and ej (j = A or B) is the unobservable factors in the individual's utility function. Assuming a linear utility function, the change in preferences (AV) of an individual, /, can be estimated in pairwise choice questions in moving from scenario A to B by: U{AA,Z)>U{AB,Z), (2.1) V{AA,Z)+sA>V(AB,Z) + s B ' (2.2) (2.3) Equation 2.3 can then be simplified to: OB (2.4) - 16-where au - (*OB is the constant term in the model, reflecting the overall preference for A over B when there is no difference between the levels of characteristics across A and B, Z.8~,An are the parameters of the model to be estimated for n interventions (n = A or B), and e are the unobservable random factors in the respondent's utility function. There are three main assumptions associated with a DCE. The first is that the intervention of interest can be broken down into specific attributes and each attribute can be further defined by at least two levels. The second assumption is that every respondent has a unique utility for each level of each attribute. The third is that these utilities can be combined across attributes; thus, the overall utility is a function of the individuals' relative preferences for the various attributes. A higher utility for a specific intervention indicates a greater preference for the attribute levels of that intervention. The attribute levels directly impact the respondent's preference, as the intervention may become more attractive as the level of a specific attribute changes. The utility function not only expresses which alternative is the most preferred but also provides an indication of its relative importance. 2.2.2 Experimental Design The design objectives for conducting a DCE are described as: identification, ensuring that the parameters of the utility function can be estimated; precision, ensuring that the statistical efficiency of the experiment allows the parameters to be estimated precisely; cognitive simplicity, ensuring that the exercise does not impose excessive cognitive burden; and realism, ensuring that the experiment realistically represents the actual choice process.[16] Current literature on DCE design has extensively addressed the first two objectives, such that the process for selecting a sample of choice sets that ensures all parameters of the utility functions can be measured precisely is well understood. However, there is less information available regarding cognitive complexity or realism. Further information on the specific components of the design elements is provided below. Identifying the Key Dimensions Based on results, either from previous studies or from qualitative analyses, the key dimensions of the intervention under investigation can be determined and incorporated into the DCE design. The number of attributes varies in the literature from two to 24, with a mode - 17-of six.[8] A recent study suggests a maximum of six attributes to avoid excessive cognitive complexity for the respondent completing the DCE. [17] Furthermore, when relevant attributes to the decision analysis are not specified, the respondents may systematically assign them a preference weight. If, however, the preference for an unassigned attribute is greater than other specific attributes in the DCE design, biased parameter estimates will result. This indicates the necessity of accurately identifying and including what most respondents consider to be the most important attributes. Assigning Levels to the Key Dimensions Once the attributes have been selected, the next crucial step in the DCE development is assigning of levels to the key dimensions. Although the choices presented to the respondents are hypothetical, they need to be as realistic as possible to encourage true responses.[l] Furthermore, the levels of the attributes need to be set so that there is a willingness-to-trade between them. Dominant responses may be encountered if individuals are unwilling to trade between attributes and choose a specific attribute level every time, independent of the other levels.[18] If the combination of attribute levels presented in the choice set suggests that some alternatives are not feasible or credible, increased random variation in responses is likely to occur.[19] In addition, factors that increase random variability, such as cognitive complexity, may also result in biased parameter estimates. [19] This further reinforces the importance both of identifying the most important attribute and of ensuring that the levels are realistic. Presenting the Scenarios In a DCE, each choice set, or scenario, is composed of two or more profiles. Each profile contains one level of each attribute in the DCE, such that all the choice sets have the same attributes but with varying levels between profiles. Evidence suggests that individuals can manage between nine and 16 choice set comparisons before they become fatigued or disinterested.fi 7] The number of choice sets rises as the number of attributes and levels increases; even for a low number of attributes and levels, the number of possible combinations can be large. For example, i f a DCE contains a total of six attributes, of which five attributes consist of four levels and one consists of three levels, there are a total of 3072 possible permutations. In the face of such large numbers, the strengths of using a full factorial design are its robustness to measure the responses of every possible combination of the attributes and levels and its ability to account for all interaction effects. However, when the number of potential choice sets becomes intractable, as shown above, cognitive complexity becomes a problem. A fractional factorial design, on the other hand, yields a carefully prescribed and representative subset of the full factorial design by reducing the total number of profiles to a cognitively acceptable number. The fractional factorial design can be constructed using computer software packages, design catalogues, or design experts. After designing the DCE, its ability to elicit valid preferences has to be assessed; this will be discussed in the next section. Conventional DCEs force respondents to actively select between the two presented options. In real life decision-making situations, individuals may not be willing to select either of the options; they are therefore classed as 'non-demanders'. For healthcare applications, individuals may prefer not to use certain drugs or specific screening programs; it is thus possible that neither option in the choice set is acceptable. As such, it is necessary that a third alternative, for example a 'neither' option, be provided to simulate the 'do nothing' option of a true decision process. Although there have been a few DCEs in healthcare that allow for non-demanders,[8] an identified drawback of the inclusion of an opt-out option is that patients may select it instead of making difficult decisions.[20] Thus, to prevent a high frequency of respondents selecting the opt-out option, and to reduce subject variability, it is vital that the levels of the attributes for this option be clearly specified. An introduction is usually included with the DCE that explains the purpose of the questionnaire and provides instructions on how to complete the exercise. Respondents are informed that the survey is a purely hypothetical exercise designed to determine the most important characteristics of the intervention and to evaluate the decision-making process for each individual. Providing a description of each attribute and its levels ensures that all participants will have the same basic level of understanding, and that no ambiguities exist in the interpretation of the attributes and associated levels. - 19-Design Criteria The literature identifies desirable DCE design features as level balance, minimum overlap, and orthogonality.[19] Level balance is obtained when the levels of each attribute appear with equal frequency. The objective of minimum overlap is to minimize the frequency of an attribute having the same level in both options of a single choice set. However, Ryan suggests that achieving no overlap is impossible; furthermore, some overlap reduces respondent burden.[21] Orthogonality of the selected profiles should be verified, using either %2 or Kendall's TB tests of association.[22] Orthogonality is achieved when a minimum correlation exists between attributes; however, this is no longer guaranteed once profiles are paired to create a scenario.[23] Although this feature is a desirable property in a choice task design, the need to maintain plausibility and realism results in this design feature often being relaxed in practice.[24] Bateman et al. suggest that deciding whether to diverge from orthogonality is best left up to the researcher.[25] There is only limited information on how these experimental design features affect the parameter estimates within DCE applications. [26,27] Selecting designs to achieve these criteria is crucial to the success of the DCE. Determining Parameter Estimates The analysis of most of the DCEs in the health literature employs the multinomial logit (MNL) model. [8] However, the most appropriate analytic method requires that the cross-elasticities of substitution be considered when more than two options within a choice set are included. The cross-elasticities reflect the changes in the probability of choosing an alternative as a function of changes in the level of utility from competing alternatives.[20] This criterion depends on the extent to which the three options are observed to be substitutes for each other. The M N L model is only applicable when the options available are all close substitutes. A survey of cancer screening preferences provides an example of the different cross-elasticities of substitution.[19] In this study, three options were offered: clinic A, clinic B, and no screening. Respondents needed to make a joint decision of whether and where to be screened for cancer. Based on the independence of irrelevant alternatives (IIA) assumption, the M N L models that an increase in the probability of choosing clinic A results in an equally - 2 0 -proportional decrease in the probability of selecting clinic B or of choosing no screening. However, i f clinics A and B are seen as substitutes and the 'no screening' option does not compete with them, changes in the probability of choosing A affect the probability of choosing B but not the probability of choosing the 'no screening' option. In this case, two separate binary logit (SBL) models are appropriate: the decision whether or not to be screened and then, for those who chose to be screened, the decision of which screening clinic to choose (A or B). A third possibility is where clinics A and B are closer substitutes with each other than with the 'no screen' option, but 'no screening' is still seen as a substitute for A and B. Thus, an increase in the probability of choosing A will result in a greater decrease in the probability of choosing B than no screen. This substitution pattern is reflected by the nested logit model ( N L M ) . The inclusive value is used to determine the validity of using the N L M . This parameter, defined by 9, indicates the degree of substitution between alternatives and ranges between zero and one [20]: 0=E(]nle^Y (2'5) where Uic\s is the deterministic component of utility of individual / for clinic c given a decision has been made to be screened S, Z,is the characteristics of individual /', 5 is an array of parameters to be estimated, and E is the expected utility for each individual. The inclusive value indicates whether or not the IIA assumption holds, and thus indicates which model should be used in the analysis. If 6 = 1, the decision of whether or not to be screened depends on both the characteristics of the individual and the expected utility of being screened. This supports the IIA assumption; hence the M N L model would be appropriate for this situation. However if 6 = 0, the expected utility of being screened does not influence the decision regarding whether or not to be screened, and hence the SBL model would be appropriate. If 0 < 6 < 1, the decision of whether or not to be screened is influenced by the expected utility of being screened; in this case, the N L M is most appropriate. The M N L , SBL, and N L M models of behaviour are all based on random utility theory. Using the example discussed in the preceding paragraphs, there are two components in this decision-making process: (i) deciding whether or not to be screened, and, assuming -21 -the individual decides to be screened, (ii) deciding which clinic to choose. [20] The N L M allows the relationships between alternatives to be considered since the SBL and M N L are regarded as special cases of the N L M ; if the IIA assumption holds, the N L M collapses to the M N L . Alternatively, if there is no relationship between the decision of whether or not to be screened and what screening clinic to choose, the N L M collapses to the SBL model. As a result, the appropriate technique, if 9 is unknown, is to conduct the analysis using the N L M to determine the inclusive value. If the value is either 0 or 1, the analysis needs to be repeated using either the SBL or M N L models; otherwise using the N L M is justified. 2.2.3 Methodological Challenges Non-Trading Behaviour The responses of some individuals may not follow compensatory decision rules. If the decision-making of an individual appears to be driven by only one attribute, then an ordering effect, known as a lexicographic order, exists and the utility function for that individual cannot be accurately estimated.[28] These respondents have very strong preferences for a given attribute and are not prepared to trade-off with other attributes. [29] Policy recommendations made on these non-compensatory decisions should be treated with caution, as their inclusion may overestimate the regression estimates for the dominant attributes.[28] However, a recent study does not justify the deletion of 'irrational' responses because it may result in the removal of valid preferences. [30] The non-trading behaviour may be due to a number of factors, including dominant preferences, poor questionnaire design, or inadequate comprehension. However, it is not possible to identify which of these factors is the fundamental cause of non-trading behaviour. While it is important to identify the individuals who are unwilling to trade, uncovering this bias may be possible only through in-person interviews.[23] Correlation A common assumption in choice experiments is that the correlation between the attributes is negligible. This assumption is due to model limitations rather than any theoretical basis. If interaction effects do in fact exist, then a reliable estimation is possible only if these effects are built into the experimental design. These interaction effects may - 2 2 -reduce the power of the model since the regression estimates may be incorrect. Further work is required to utilize other analytical models, such as the mixed logit model, to attempt to account for these correlations. [31] Psychometric Properties Despite the advantages of using DCEs to evaluate and quantify preferences, information regarding their psychometric properties, such as reliability and validity, is limited due to a lack of published studies on these topics. One significant drawback is that the results from a DCE are not generalizable to different populations and hence must be repeated for each population. This concept has not been addressed and additional work is needed to identify and compare real and stated behaviours of respondents when completing a DCE. As the number of relevant characteristics of an intervention is virtually limitless, the structural reliability needs to be evaluated as only a select number of attributes and levels can be incorporated into a DCE. In a previous study, the consistency of the responses between alternative attributes is evaluated by administering two DCEs to the participants, one with four and the other with five attributes.[32] This experiment determines if the additional attribute results in a change in the attribute-trading rates; the results reveal no statistical difference between the regression estimates from the two models. Another study investigated the structural reliability by identifying the consistency of the responses between two DCEs, where one DCE has fewer levels for one attribute; no statistical difference between the two DCEs is reported.[33] Due to the lack of revealed preference data, it can be difficult to determine whether valid responses are obtained from a DCE; therefore, theoretical (internal) validity and internal consistency are used to determine the validity of a DCE. Previous DCEs have emphasized the theoretical validity of the regression estimates, in that they should be consistent with basic economic theory and, more generally, a priori expectations. Based on the axioms of economic theory, internal consistency can be assessed by either monotonicity or stability. Monotonicity postulates that an individual should prefer more rather than less of any good. This is assessed by including a choice set where one profile is dominant, in that the levels are considerably 'better' for all attributes. The concept -23 -of stability asks individuals to consider the same scenario within the same survey. The internal consistency can then be evaluated based on the rationality of the choices made. For example, if one scenario is clearly 'superior' in all aspects to another, individuals are expected to choose the 'better' scenario; furthermore, individuals are expected to make the same selection when the scenario is presented twice in the same DCE. While the majority of studies have used the monotonicity approach when evaluating internal consistency, there is no consensus in the literature on how to deal with inconsistent responses. In some previous studies, inconsistent results were removed on the basis that the subjects may not have understood the questionnaire or were not taking the process seriously.[1,7,17] However, others report that there is no obvious justification for this arbitrary removal as there is no indication that failing to remove the inconsistent responses biases the results.[30,34,35] In either case, it is necessary to remove the dominant choice set result from the regression analysis since it is a consistency check and not a part of the DCE design.[36] Although methodological challenges, such as non-trading behaviour, correlation, and psychometrics, need to be addressed further, DCEs have demonstrated their success for many clinical indications and various healthcare applications. 2.2.4 An Evaluation of Patients' Preferences in Asthma Over the past decade, DCEs have been increasingly used in health applications. [1,8] Their popularity is due to their similarity with day-to-day decision-making processes. Of particular importance in the assessment of patients' preferences for asthma treatments is the ability of DCEs to incorporate a diverse set of attributes in their design. These attributes may consist of a range from benefits to risks, including both clinical and non-clinical outcomes, which are representative of asthma therapies. However, despite their advantages, DCEs alone do not permit decision-makers to compare asthma treatments with treatments for other diseases. To meet this broader spectrum requirement, techniques that measure the overall effectiveness of treatments in units which can be directly compared across diseases, are needed. -24 -2.3 Measures of Utility The requirement for a universal measure of effectiveness stems from the need to address how best to quantify health consequences. Typically, these are expressed in outcome units specific to the disease under investigation, such as the number of symptom-free days, the years of life gained, and the reduction of blood pressure (mm Hg) in the treatments of asthma, thrombolysis, and hypertension, respectively. [37-39] Unfortunately, these approaches do not allow for the comparison of treatments across different disease states. Utility measures reflect the preferences of patients for treatment processes and health outcomes. The key elements of these utilities are that they incorporate preference measurements and that they differentiate health states, from perfect health all the way to death. [10] Utility scores reflect both the absolute health status and the value that patients place on their health status. Utilities reflect the desirability of living in a particular state, anchored by a value of 'one' for full health and 'zero' for death. Some health states, such as time spent in coma,[40] can attract negative values which indicates, from a societal perspective, as states being worse than death. This, of course, cannot be interpreted either to mean that such patients wish to die or that the societal perspective is that such patients be allowed to die; it represents the fact that healthy individuals consider that existing in such health states is worse than death. [41] The use of utilities permits the calculation of a unit widely used in economic evaluation, the quality-adjusted life year (QALY). 2.3.1 Quality-Adjusted Life Year Numerous health technology assessment organizations, including the Canadian Agency for Drugs and Technologies in Health (CADTH) and the National Institute for Clinical Excellence (NICE), indicate that QALYs are the required measure for health outcomes.[42,43] In the Q A L Y approach, the quality adjustment is based on utilities, which reflect the relative desirability of a particular health state. The utilities obtained for each state are then multiplied by the time spent in that state and the sum of the products yields the total QALYs for n time periods, such that: where tt is the time interval of period / in years and w, is the utility of period /. -25 -2.3.2 Direct Methods Various methods have been implemented for measuring preferences; however, some confusion results from the interchangeable use of the terms preference, value, and utility by researchers to describe quality weights. [44] Preference generally describes the desirability of a set of outcomes, while value and utility are specific types of preference. [45] Value is measured under conditions of certainty, while utility is measured under conditions of uncertainty. The three most widely-used methods to directly measure the preferences of individuals for health outcomes are the rating scale (and its variants), the standard gamble (SG), and the time trade-off (TTO). In all these techniques, the respondents directly identify their health state. Rating Scale, Category Scale, and Visual Analogue Scale These scales are the simplest approaches for measuring preferences. Using a two-step process, the respondent initially ranks health outcomes from the most to the least preferred and then places these outcomes on a scale, such that the intervals between placements correspond to the perceived differences in preferences. The fundamental purpose of this approach is to create an interval scale of preferences.[45] There are many variations on the rating scale approach. Rating scales typically have a scale of numbers; category scales consist of a small number of categories that are assumed to be equally spaced; visual analogue scales (VASs) consist of a line on a page, often 10 cm in length, with clearly defined endpoints and with or without other marks along the line. Scores from a rating scale give the researcher a firm indication of the ordinal rankings of the health outcomes and the intensity of those preferences. This technique is often used as a prelude to administering other preference exercises to familiarize the respondents with the descriptions of the outcomes. Standard Gamble The SG is a classic method of measuring cardinal preferences. Grounded in von Neumann-Morgenstern expected utility theory,[45] the SG method requires the respondents to make a choice between outcomes, in which one outcome involves an element of uncertainty. This gamble involves a probability of a better or worse outcome than the certain -26 -outcome.[46] The goal of this approach is to determine the probability in the gamble at which the respondent is indifferent between the certain and uncertain alternatives. Time Trade-Off Designed as an alternative to the SG, the TTO approach aims to overcome the problems encountered when explaining probabilities to respondents.[45] The respondents are asked to choose between a shorter life span in perfect health versus a longer life span in an uncertain health state. The time in full health is varied until the respondent is indifferent between the two alternatives. The TTO choice is not made under uncertainty so the values that it elicits are not considered to be measures of utility as derived from expected utility theory. 2.3.3 Indirect Methods Pre-scored multi-attribute health status classification systems were developed to alleviate the complex and time-consuming nature of collecting health preferences using the direct methods. The three most widely used multi-attribute health status classification systems are the Health Utilities Index, EuroQol, and Short Form 6D. Health Utilities Index The Health Utilities Index (HUI) currently consists of two systems, HUI-2 and HUI-3.[47,48] Each includes a health status classification system and a utility scoring formula. In both cases, the scoring formula is based on the SG utilities measured in the general populace. The HUI uses multi-attribute utility theory for the estimation of the utility formula! [44,45] Using this theory, a single summary index is derived by taking a weighted average across all factors that simultaneously characterize the phenomena; each factor in the model can vary in amount. For most applications, the HUI-3 should be used as the primary analysis. [45] While based closely on the classification system of the HUI-2 (with the elimination of the fertility attribute), the HUI-3 has a more detailed descriptive system.[47,48] It expands the single sensory attribute of the HUI-2 into three attributes: vision, hearing, and speech. Thus, the self-administered HUI-3 survey includes a total of eight attributes: vision, hearing, speech, - 2 7 -ambulation, dexterity, emotion, cognition, and pain, as opposed to the seven on the HUI-2. As the number of levels for each attribute varies from four to six, the total number of possible health states is 972,000. The baseline preferences for the HUT3 were measured on a random sample of adults residing in Hamilton, Canada using both a VAS and a SG instrument. Using a multiplicative model, states worse than death can be measured on the zero (dead) to one (perfect health) scale, with a lower bound of -0.36. EuroQol The EuroQol (EQ-5D) is designed as a cardinal index of health for describing and valuing HRQL. This self-administered survey consists of a descriptive health state classification system with five attributes (mobility, self care, usual activity, pain/discomfort, and anxiety/depression) and a V A S 'health thermometer'.[49] The 'health thermometer' represents a subjective, global evaluation of the respondent's health status on a scale between zero and 100, where zero represents the worst imaginable health state and 100 represents the best imaginable health. Each attribute in the classification system has three levels: no problem, some problems, and major problems, for a total of 243 possible health states. Preferences for the scoring function were measured with the TTO technique from a random sample of the adult population in the United Kingdom.[50,51] The scoring function uses econometric modeling, with scores falling on the zero (dead) to one (perfect health) value scale, with a lower bound of-0.59. Short Form 6D The Short Form 6D (SF-6D) is derived from another popular HRQL questionnaire, the Short Form 36.[52] The SF-6D consists of a multi-attribute health status classification system with six attributes and a scoring table, and like the EQ-5D, uses econometric modeling. There are six attributes in this survey: physical functioning, role limitations, social functioning, pain, mental health, and vitality. The classification system ranges from four to six levels for each of the six attributes for a total of 18,000 unique health states. The scoring model for the SF-6D is based on the SG utilities of 836 members of the general population in the United Kingdom.. The obtained utilities fall on the conventional health utility scale -28 -ranging from zero (dead) to one (full health); the worst state in the SF-6D system has a utility of0.30. There are numerous other methods for determining the utility for use in the Q A L Y framework, including direct methods (i.e. rating scales, SG, and TTO approaches) and indirect methods (i.e. EQ-5D, HUI-2, HUI-3, and SF-6D surveys). Due to their ease and low cost of administration, the indirect methods are most commonly used; however, these questionnaires differ in the dimensions of health they cover, in the number of levels defined for each dimension, in the description of these levels, and in the severity of the most severe level. 2.3.4 An Evaluation of Health-Related Quality of Life in Asthma Overall, utility measures are useful for differentiating between individuals who have a better HRQL and those who have a worse HRQL; however, there is still a need to determine which HRQL instrument is the most discriminative when evaluating health states. Specifically, in asthma, there is a lack of empirical evidence assessing which of the indirect preference-based instruments yield the most valid results. To determine which HRQL instrument to use, the construct validity of the instruments needs to be assessed. Construct validity examines the logical relations that should exist between a measure and characteristics of patients and patient groups.[10] Valid instruments should be able to discriminate across different levels of asthma control. 2.4 Summary The rationale and methodology employed for this thesis draw on the fundamentals of health economics. One technique which is being increasingly used in the preference evaluation of health interventions is the DCE. This framework allows for the exploration of the relative preferences and the assignment of monetary valuations for a specific intervention. Features contributing to the overall success of the DCE design are identified as level balance, minimum overlap, and orthogonality. While there are many issues regarding DCE methodology and psychometry that still need to be addressed, DCEs are being increasingly applied in the healthcare setting. One application has been to elicit patients' preferences for specific components of asthma care. An in depth review of the current - 2 9 -literature evaluating patients' preferences for the treatment of asthma will be discussed in Chapter 3. Using the techniques described here, a DCE was developed to allow a better understanding of patients' preferences in asthma; the results are discussed in Chapter 4. The application of the D C E to asthma treatments demonstrates its ability to incorporate a wide range of benefits and risks. While DCEs are an effective technique to quantify patients' preferences for different aspects of a resource allocation decision, they are limited to being context-specific. As such, there is a need to evaluate other techniques that measures the overall effectiveness of treatments in common units. The Q A L Y is a common metric used in the evaluation of different treatments in cases where a set of utilities for each possible health state needs to be derived. Utility measures reflect the preferences of patients for treatment processes and health outcomes. While there are a number of methods to determine the quality weighting for the QALYs , administering indirect preference-based instruments are preferred due to their ease of administration and low cost. As such, there is a need for assessing the construct validity of these instruments in various disease states, including asthma. In particular, their ability to discriminate across different levels of asthma control needs to be assessed, as will be discussed in Chapter 5. - 3 0 -References Farrar S, Ryan M , Ross D, Ludbrook A. Using discrete choice modeling in priority setting: an application to clinical service developments. Soc Sci Med 2000;50:63-75. Little P, Everitt H , Williamson I, Warner G, Moore M , Gould C, et al. Observational study of effect of patient centredness and positive approach on outcomes of general practice consultations. B M J 2001;323:908-11. Guadagnoli E, Ward P. Patient participation in decision-making. Soc Sci Med 1998;47:329-39. Coulter A . Partnerships with patients: the pros and cons of shared clinical decision-making. J Health Serv Res Policy 1997;35:276-81. Moher D. CONSORT: an evolving tool to help improve the quality of reports of randomized controlled trials. Consolidated Standards of Reporting Trials. J A M A 1998;279:1489-91. Altman DG, Schulz K F , Moher D, Egger M , Davidoff F, Elbourne D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med 2001;134:663-94. Ryan M , Bate A , Eastmond DJ, Ludbrook A. Use of discrete choice experiments to elicit preferences. Qual Health Care 2001 ;10(Suppl I):i55-i60. Ryan M , Gerard K. Using discrete choice experiments to value health care programmes: current practice and future research reflections. Appl Health Econ Health Policy 2003;2(l):55-64. Hensher DA, Rose JM, Greene WH. Applied choice analysis. A primer. Cambridge: Cambridge University Press; 2005. Guyatt GH, Feeny D H , Patrick DL. Measuring health-related quality of life. Ann Intern Med 1993;118:622-9. Luce D, Tukey J. Simultaneous conjoint measurement: A new type of fundamental measurement. J Math Psychol 1964;1:1-27. Louviere JL. Conjoint analysis modeling of stated preferences. A review of theory, methods, recent developments, and external validity. J Transport Econ Policy 1998;22:93-119. Sheldon RJ, Steer JK. The use of conjoint analysis in transport research. Transport Res 1982;145-58. -31 -14. Swallow S, Opaluch J, Weaver T. Siting noxious facilities: an approach that integrates technical, economic, and political consideration. Land Econ 1992;68:283-301. 15. Lancaster K. Consumer Demand: A new approach. New York: Columbia University Press; 1991. 16. Louviere JL, Hensher DA, Swait JD. Stated choice methods: analysis and applications. Cambridge: Cambridge University Press; 2000. 17. Ryan M . Using conjoint analysis to take account of patient preferences and go beyond health outcomes: an application to in vitro fertilization. Soc Sci Med 1999;48:535-46. 18. Ratcliffe J, Buxton M . Patient's preferences regarding the process and outcomes of life-saving technology: an application of conjoint analysis to liver transplantation. Int J Tech Assess Health Care 1999;14:340-51. 19. Viney R, Savage E, Louviere J. Empirical investigation of experimental design properties of discrete choice experiments in health care. Health Econ 2005;14(4):349-62. 20. Ryan M , Skatun D. Modeling non-demanders in choice experiments. Health Econ 2004;11:397-402. 21. Ryan M . Using discrete choice experiments in health economics: theoretical and practical issues. Three-day residential workshop, Lake Louise, Alberta, Canada; 2004. 22. Lloyd A , Mcintosh E, Price M . The importance of drug adverse effects compared with seizure control for people with epilepsy. A discrete choice experiment. Pharmacoeconomics 2005;23(11):1167-81. 23 Scott A , Stuart Watson M , Ross S. Eliciting preferences of the community for out-of-hours care provided by general practitioners: a stated preference discrete choice experiment. Soc Sci Med 2003;56:803-14. 24. Kjaer T. A review of the discrete choice experiment - with emphasis on its application in health care [discussion paper]. Odense: Department of Health Economics, University of Southern Denmark; 2005. 25. Bateman IJ, Carson RT, Day B, Hanemann M , Hanley N , Hett T, et al. Economic Evaluation with Stated Preference Techniques. Cheltenham: Edward Elgar Publishing Limited; 2002. 26. Carlsson F, Martinsson P. Design techniques for stated preference methods in health economics. Health Econ 2003;12:281-94. - 3 2 -27. Maddala T, Phillips K A , Johnson RF. An experiment on simplifying conjoint analysis designs for measuring preferences. Health Econ 2003;12:1035-47. 28. Scott A . Giving things up to have more of others. The implications of limited substitutability in eliciting preferences for health and health care [discussion paper]. Aberdeen: Health Economics Research Unit, University of Aberdeen; 1998. 29. Payne JW, Bettman JR, Johnson EJ. Behavioural decision research: A constructive processing perspective. Ann Rev Psychol 1992;43:87-131. 30. Lanscar E, Louviere J. Deleting 'irrational' responses from discrete choice experiments: a case of investigating or imposing preferences? Health Econ 2006;15:797-811. 31. Revelt D, Train K. Mixed logit with repeated choices: households' choices of appliance efficiency level. Rev Econ Stat 1998,80:1-11. 32. Bryan S, Parry D. Structural reliability of conjoint measurement in health care: an empirical investigation. Appl Econ 2002;34:561-7. 33. Ratcliffe J, Longworth L. Investigating the structural reliability of a discrete choice experiment with health technology assessment. Int J Tech Assess Health Care 2002;18(l):139-44. 34. Phillips K A , Johnson FR, Maddala T. Measuring what people value: a comparison of "attitude" and "preference" surveys. Health Serv Res 2002;37(6): 1659-79. 35. Phillips K A , Maddala T, Johnson FR. Measuring preferences for health care interventions using conjoint analysis: an application to HIV testing. Health Serv Res 2002;37(6):1681-1705. 36. Lloyd A , Mcintosh E, Price M . The importance of drug adverse effects compared with seizure control for people with epilepsy. Pharmacoeconomics 2005 ;23( 11): 1167-81. 37. Sculpher MJ , Buxton MJ. The episode-free day as a composite measure of effectiveness. Pharmacoeconomics 1993;4:345-52. 38. Logan A G , Milne BJ , Achber C, Campbell WP, Haynes RB. Cost-effectiveness of a worksite hypertension treatment programme. Hyptertension 1981 ;3:211-8. 39. Mark D B , Hlatky M A , Califf R M , Martin JS, Weaver WD. Cost-effectiveness of thrombolytic therapy with tissue plasminogen activator as compared with streptokinase for acute myocardial infaraction. N Eng J Med 1995;332:1418-24. 40. Patrick DL, Starks HE, Cain K C , Uhlmann RF, Pearlman RA. Measuring preferences for health states worse than death. Med Decis Making 1994;14:9-18. -33 -41. Hurst NP, Kind P, Ruta D, Hunter M , Stubbings A. Measuring health-related quality of life in rheumatoid arthritis: validity, responsiveness, and reliability of EuroQol (EQ-5D). Br J Rheumatol 1997;36:551-9. 42. Technology appraisal programmed of the National Institute for Clinical Excellence. A review by the World Health Organization [Online]. [2003?] [cited 2006 Jun 6]; Available from: URL:http://www.euro.who.int/ Document/E81254.pdf 43. Canadian Agency for Drugs and Technologies in Health [Online]. 2006 Mar 30 [cited 2006 Aug 1]; Available from: URL:http://www.cadth.ca 44. Neumann PJ, Goldie SJ, Weinstein M C . Preference-based measures in economic evaluation in health care. Ann Rev Public Health 2000;21:587-611. 45. Drummond M F , Sculpher MF, Torrence GS, O'Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programmes. 3rd ed. New York: Oxford University Press; 2005. 46. Tengs TO, Wallace A. One thousand health-related quality of life estimates. Med Care 2000;38:583-637. 47. Torrence GW, Feecy D H , Furlong WJ, Barr RD, Zhang Y , Wang Q. Multiattribute utility function for a comprehensive health status classification system: Health Utility mark 2. Med Care 1996;34:702-22. 48. Feeny D, Furlong W, Torrance GW, Goldsmith C H , Zhu Z, DePauw S, et al. Multiattribute and single-attribute utility functions for the Health Utilities Index mark 3 system. Med Care 2002;40:113-28. 49. The EuroQoL Group. EuroQoL - a new facility for the measurement of health-related quality of life. Health Policy 1990;16:199-208. 50. Dolan P, Gudex C. Time preference, duration and health state valuations. Health Econ 1995;4:289-99. 51. Dolan P, Gudex C, Kind P, Williams A. The time trade-off method: results from a general population study. Health Econ 1996;4:141-54. 52. Brazier J, Roberts J, Deverill M . The estimation of a preference-based measure of health from the SF-36. J Health Econ 2002;21:271-92. - 34 -C H A P T E R 3 PATIENTS' P R E F E R E N C E S IN T H E T R E A T M E N T O F A S T H M A : A R E V I E W OF C U R R E N T EVIDENCE* 3.1 Introduction Asthma is a chronic respiratory condition that poses a serious global health challenge. Worldwide, over 300 million people are afflicted with this disease, and the prevalence has been increasing steadily over the past few decades. [1] Although asthma has no known cure, treatment strategies in accordance with current management guidelines can result in essentially complete control in all but those with the most severe disease.[2,3] The most effective management strategies encourage a multifaceted, stepped-care approach that is dependent on symptom frequency and disease severity. Suboptimal asthma control is still widespread, despite good understanding of asthma control requirements and the availability of effective treatments. Specifically, short-acting (SA) [3-agonists are being overused as control strategies where other treatments would be more effective. [4,5] This excessive use is a potential risk factor for increased asthma-related morbidity and mortality. [6,7] Inappropriate asthma management may be the result of complex interactions between factors, such as available drug therapies, treatment cost, medical insurance coverage, patient and healthcare provider preferences, and healthcare provider knowledge. There is increasing understanding that patients' preferences should be taken into account during medical decision-making to maximize strategy effectiveness.[8] Patients make decisions by understanding the available options, weighing their individual preferences, and then trading-off between the perceived advantages and drawbacks. The simplest approach to elicit patients' preferences is a crossover study, in which patients try various treatments sequentially. [9,10] More complex approaches apply instruments grounded in utility and psychological theories, including standard gamble, time trade-off, conjoint analysis, and discrete choice experimentation.[11-13] *A version of this chapter is currently under review for publication in the European Respiratory Journal. McTaggart-Cowan, H . M . , FitzGerald, J .M., Anis, A . H . , Lynd, L .D. Patients' Preferences in the Treatment of Asthma: A Review of Current Evidence. -35 -Patient's preferences are one factor that may contribute to poor control and inappropriate asthma management. The literature provides various studies investigating different aspects of patient's preferences for asthma treatments. The objective of this work was to systematically compare and critically appraise these studies. 3.2 Methods A systematic review identified all published studies evaluating patients' preference for various aspects of asthma treatments. An electronic search of Medline, Cochrane Collaboration, and E M B A S E databases identified all English articles published between March 1, 1992 and November 30, 2005. The search strategy included the following Medical Subject Heading (MeSH) terms: 'asthma' and 'patient satisfaction' and the key words: 'patient preferences' and 'preferences'. Articles were eligible for inclusion if they specifically evaluated patients' preferences within the context of asthma care. A secondary search of the reference lists of the retrieved articles provided a more thorough list of publications not identified in the primary search. A l l studies that met the inclusion criteria were then categorized based on which component of patients' preferences they evaluated: treatments (i.e. symptom relief, frequency, inhaler type, and administration route), management (i.e. action plans, doctor visits), involvement in treatment decision-making, and symptoms. Extracted descriptive information included details of the study populations, year of study, and study design. Due to the heterogeneity of the study populations and objectives, it was not possible to conduct a formal meta-analysis to estimate a summary measure of patient preferences. Instead, all the study results were compared as published; potential explanations for the variations in preferences between studies are presented and discussed. 3.3 Results One hundred and seventeen studies met the initial search criteria. This review excluded ninety-five of these studies because they did not pertain to asthma (n = 2), were non-English publications (n = 4), only evaluated physician's preferences (n = 5), or did not empirically evaluate patients' preferences (n = 84). The analysis included the remaining 22 studies that met the inclusion criteria. The attributes analyzed in these studies were patients' - 3 6 -preferences for treatments (Table 3.1), management, decision-making, and symptoms (Table 3.2). Although the majority of the studies used questionnaires designed specifically for the study (n = 17), five studies used previously validated instruments, such as the asthma Autonomy Preference Index (API) and the Control Preference Scale (CPS). 3.3.1 Treatment Preferences Of the 22 reviewed studies, 13 studies evaluated patients' preferences pertaining to asthma therapy (Table 3.3). The three main preference themes were hypothetical treatment attributes (n = 3) such as frequency of administration, symptom relief, and monthly cost; inhaler type (n = 7); and administration route (n = 3). The experimental designs in the reviewed studies included crossover (n = 5), cross-sectional (n = 3), and cohort (n = 5). Hypothetical Treatment Attributes Three studies used techniques grounded in utility and psychological theories to evaluate preferences for hypothetical treatment attributes.[21,23,30] Balsbaugh et al. explored the willingness to take a drug in terms of the route and frequency of administration and the need for blood test monitoring for adverse events using a conjoint analysis that involved ranking treatment scenarios on a five-point Likert scale. [23] Patients were most willing to take an inhaled medication once daily with no blood monitoring (score 1.8), while an oral drug taken four times daily that required blood test monitoring was least favoured (score 4.0). With the notable exception of the rarely-used theophylline, there is no treatment requiring blood test monitoring, which restricts the applicability of the results to current practice. Another conjoint analysis assessed the relative preferences for maintenance therapy type (corticosteroids versus long-acting P-agonists), use of SA p-agonists, time to and duration of relief, number of monthly symptom-free days (SFDs), and monthly cost. [30] The results revealed that patients focused on treatment effectiveness by preferring scenarios with greater numbers of SFDs. Approximately half of patients preferred a combination inhaler to using separate inhalers and 78% preferred a reliever that was both rapid and long-acting. Since this conjoint analysis included a cost component, the average monthly willingness-to-pay for the preferred treatment was 328 Swedish krona (2003 US $36). Finally, patients' preferences for a hypothetical drug that was equally effective but safer than currently - 3 7 -available treatments was investigated using a two-stage contingent valuation method. [21] The first phase assessed the trading rate between safety and efficacy; 35% of the respondents indicated a willingness to trade-off safety for efficiency by selecting the more effective but riskier drug. Unfortunately, the absence of a description for the riskier treatment made it unclear whether drug safety related to severe (i.e. life-threatening) or minor (i.e. oral candidiasis) events. In the second stage, the annual willingness-to-pay for complete asthma control ranged from US$1000 to US$3700. Specific Inhaler Types Five of the seven studies investigating patients' preference for different inhalation systems indicated that patients preferred dry powder inhalers (DPI) over metered dose inhalers (MDI).[19,20,26,28,29,31] In a crossover study, Boe et al. assessed 123 adult preferences for a MDI with and without a spacer and for a DPI using a 10 cm long visual analogue scale anchored by a value of zero for an extremely negative answer and ten for a very positive answer. [19] Specifically, the survey focused on the preference for appearance, taste, ease of use, and efficacy, relating to both budesonide, an inhaled corticosteroid and terbutaline, a SA [3-agonist. The DPI was rated significantly higher than the MDI for both budesonide (8.16 versus 6.88, respectively; pO.OOl) and terbutaline (7.87 versus 7.60, respectively; p<0.01). Two similar studies supported these results, such that 91/154 (59%) participants preferred fluticasone delivery by DPI relative to MDI,[26] and 84/114 patients (74%) preferred budesonide delivered by DPI over flunisolide, fluticasone, and beclomethosone delivered via MDI. [29] These latter two studies recorded the proportion of patients preferring which treatment based on survey results. While both studies found that patients preferred DPIs because of their greater ease of use, Welch et al. also found that the lack of taste and shorter administration time associated with DPIs were also important. Three additional studies employing cohort designs, evaluated the preference for inhaler types. Two resulted in conclusions similar to those discussed previously,[28,31] while the results from the third study were equivocal.[20] Specifically, in their unblinded study of preferences for a new DPI with a dose counter and moisture protection relative to a MDI, Liam et al. found that 71% preferred the new DPI due to the presence of the dose counter and the perceived ease of use; these preferences were not associated with age, sex, - 3 8 -education level, occupation, or duration of using MDI.[28] Unfortunately the presence of the dose counter does not permit the generalization of these findings to all DPI. Using an adapted version of Boe's satisfaction questionnaire,[19] the second cohort study found a greater preference for budesonide delivered via DPI over triamcinolone acetonide administered via MDI (5.52 versus 4.93, respectively; p<0.001).[31] The final cohort study did not find any difference in the proportion of participants that preferred the DPI to the MDI. [20] In an attempt to improve drug delivery and treatment compliance, ongoing advances in inhaler technology have resulted in the development of DPI with different drug delivery methods. To evaluate some of these methods, Majahan et al. used a randomized, double-blind, double-dummy, placebo-controlled study of patient preferences for fluticasone delivery via two different DPI delivery devices, the Diskhaler® and the Diskus®.[22] After 12 weeks of use, 153/213 (72%) of the participants expressed a high level of overall satisfaction with the performance of the Diskhaler® on all attributes, which included convenience to carry, durability, ease of use (loading, holding, operating, and cleaning), and the ability to determine the number of remaining doses. Sixty-one percent of respondents preferred the Diskus® inhaler, 25% preferred the Diskhaler® inhaler, with the remainder being indifferent. Administration Modes In a series of randomized crossover studies, patients consistently indicated a strong preference for orally-administered over inhaled treatments.[24,25,27] Weinberg and Naya administered a questionnaire on preference and ease of use to compare opinions for zafirlukast tablets relative to beclomethasone dipropionate MDI in adolescents (age 12 to 17 years).[24] The results revealed that 86/123 (70%) respondents preferred the oral tablets (p<0.01). Two similar studies supported these results but using a population of children with asthma (age 6 to 11 years) and their caregivers.[25,27] One study using interviews found that 273/333 (82%) of the children and 290/333 (87%) of the caregivers had a significantly greater preference for oral montelukast sodium over inhaled cromolyn sodium. Using a previously validated questionnaire addressing symptom relief, ease of use, lifestyle interference, taste, social concern, and overall satisfaction, Maspero et al. investigated the preferences of oral montelukast versus inhaled beclomethasone (n = 124). [14,25] Both - 3 9 -children (p = 0.001) and caregivers (p = 0.05) reported a significantly greater overall satisfaction with orally-administered montelukast over beclomethasone. 3.3.2 Asthma Management Patient's preferences for personal action plans and other factors for improving asthma management can directly impact asthma control. Four crossover studies, of which two used qualitative approaches, one a mixed-method design, and one conjoint analysis, investigated these preferences (Table 3.4).[32-35] Studies indicate that written action plans, a central component of management guidelines, are associated with better asthma control.[2,15,16] Jones et al. evaluated the views of action plans held by general practitioners (n = 11), nurses (n = 13), and patients (n = 32) using separate focus groups.[32] A l l three groups displayed limited enthusiasm for action plans due to a general distrust of their usefulness or relevance. Healthcare professionals agreed that each plan had to be "individually constructed" and "regularly reviewed", but felt that continuing education was more important for maintaining control. Non-compliant patients felt the plans could be useful for people with "more serious" or "proper" asthma, whereas compliant patients felt it was "pointless for them" or they already had "a full understanding of the issues". Contradictory results were observed in Douglass' study, which used semi-structured in-depth interviews to evaluate the importance of asthma plans (n = 62).[33] Although most of the patients did not have action plans, many considered them potentially useful for management of their asthma. Haughney et al.'s study provided insights into patients' preferences for asthma management. [3 5] Using a mixed-method design, 40 in-depth interviews were conducted to explore demographic details, asthma history, treatment, control, and preferences. Both qualitative and quantitative results showed that many patients were opposed to a treatment regime that was more complicated than their current two inhaler system. While 80% of respondents did not have personal action plans, the majority (68%) indicated that they would have preferred to have such a plan. Asthma management is influenced by many variables; as a result, Ratcliffe et al. used a conjoint analysis to evaluate patients' preferences for the amount of time spent with the doctor, the possibility of having the same doctor each visit or of having the doctor treating - 4 0 -the patient as a whole, symptom relief, treatment-related side effects, and travel costs associated with a physician's clinic.[34] A l l but the probability of experiencing side effects significantly influenced preferences (p<0.05) for asthma management. This result indicated, from the patients' perspective, that side effects were not important when selecting an asthma treatment, but symptom relief and an improved doctor's visit (i.e. more time with the doctor, having the same doctor each visit, and having the doctor treating the patient as a whole) were important management strategies. 3.3.3 Decision-making Four cross-sectional studies (three quantitative and one qualitative) evaluated the role of the patient in decision-making (Table 3.5).[36-39] Two studies utilized the asthma Autonomy Preference Index (API), an instrument composed of vignettes representative of the management guidelines.[36,37] The API is a 26-item preference scale with two domains: information-seeking and decision-making. Each item is scored on a five-point Likert scale and linearized for a maximum score of 100, which represents the strongest desire for either knowledge or management independence. In another study, a mean API score of 90 was determined from a sample of 133 patients. Older participants expressed less interest in being involved in, or making, decisions.[36] In a separate cross-sectional study, Adams et al. utilized an adapted version of the API and found that only 78/212 (37%) of respondents preferred total autonomy for treatment decisions. [37] Sixty-four percent of the respondents expressed stronger preferences for autonomy when dealing with a moderate attack compared to either arranging routine doctor's visits or for managing a severe episode requiring hospitalization and admission to an intensive care unit (p<0.001). However, the preferences for the moderate attack focused more on autonomy in scheduling physician visits rather than for altering medications. Multiple regression analysis indicated stronger autonomy preferences for treatment decision-making in patients who were more concerned about treatment-related adverse events, made greater use of active coping strategies to become further informed and involved in asthma management, had higher levels of formal education, had greater concerns about costs resulting from delays in seeking care, or had a physician that encouraged participatory decision-making. -41 -A pair of studies conducted by Caress et al. evaluated patients' preferences for decision making using the Control Preference Scale (CPS).[38,39] Composed of five vignettes, the CPS is based on a theoretical participation model from a continuum of active, through collaborative, to passive participation. [17] The pilot study used focused, conversation-style interviews to explore the preferred decision-making roles of 32 patients.[38] Fourteen patients (44%) preferred a passive role, involving their physician to make the final decision because of the "doctor knows best" belief. They also felt that they lacked the knowledge, experience, or confidence to participate in the decision-making. In the full study, 82/230 (36%) of the respondents preferred a collaborative role, where the healthcare professional and the individual share decision-making responsibilities, while 55 (24%) and 93 (41%) preferred active and passive roles, respectively. [39] A greater preference for participation in decision-making was associated with younger age (pO.OOl), higher education (p = 0.003), and higher socioeconomic status (p<0.001). There was no association between the interest in decision-making role and sex (p = 0.16), disease duration (p = 0.54) or asthma severity (p = 0.22). A l l the studies that investigated decision-making preferences generally agreed that most patients preferred some degree of involvement in the decision-making process. 3.3.4 Symptom Preferences In addition to Ratcliffe et al.'s study, which indicated a strong preference among asthma patients for the avoidance of symptoms,[34] Osman et al. evaluated the relative importance of cough, breathlessness, wheeze, chest tightness, and sleep disturbance in 272 patients with moderately severe asthma. [40] The respondents chose between pairs of scenarios comprised of varying levels of severity of each symptom. In this sample, patients were willing to accept higher levels of wheeze, sleep disturbance, and chest tightness to achieve lower levels of cough and breathlessness; this suggests that for patients some symptoms are more important than others. 3.4 Discussion The main findings from the 22 reviewed studies suggest that treatment aspects differ in importance to patients. In terms of medication delivery, oral is preferred over inhaled -AT-administration; for inhaled therapies, DPIs are preferred over MDIs. The asthma management studies indicated that the patient preferred to have action plans implemented, to use fewer inhalers, and to have an improved and a more interactive visit with the doctor. Patients also desired at least a limited role in self-management decisions, although the degree of decision-making autonomy that participants desired varied substantially. For symptoms, a reduction in the severity of cough and breathlessness was more important than improvements in wheeze, sleep disturbances, and chest tightness. A common theme among the studies was that a therapy's ease of use and convenience are integral to the patient's choice [24,26,29]; this, effect was demonstrated by a consistent preference for oral versus inhaled therapies. The ease and quickness of administration, as well as confidence in knowing the amount of drug taken and that it was taken properly were the primary reasons identified for this preference.[24,25,27] Similarly, in comparing dry power and MDIs, the DPIs were always favoured due to their ease of use, promptness of effect, and lack of taste. [29] Some patients preferred a passive role in decision-making because that they believed they lacked the knowledge, experience, or confidence to participate. Medical, emotional, and social needs, health beliefs, and individual attitudes may influence patient compliance and the decisions that patients make to stay on a specified prescribed treatment. Many patients believed that if they had accurate knowledge of treatment side effects and knew the benefits of late (instead of omitted) dosing, their medication adherence would improve. Therefore, it is evident that properly informing patients about available treatment options is essential to ensure that they make decisions in accordance with the management guidelines. [3] This would reduce the excessive use of SA (3-agonists and ultimately improve asthma control. [4,5] The majority of the reviewed studies used a simple survey to determine what treatment characteristics or treatments were preferred and reported the proportion of patients expressing a preference for a specific attribute or treatment. Preferences obtained in this manner are not grounded in utility or psychological theories, and as such do not accurately represent a true decision process. To assess preferences that are representative of an actual decision process, the respondent needs to be forced to trade-off benefits against drawbacks. Decision-making involves making judgments regarding the value of a particular health state -43 -or a treatment option. A number of preference-based techniques, such as the standard gamble and time trade-off, are available to formally quantify preferences; however, only five of the 22 reviewed studies utilized preference-based techniques.[21,23,30,34,40] Conjoint analysis or discrete choice experimentation is the process of asking decision-makers (in this case, the patient), to make trade-offs replicating real life decision-making; four of the reviewed studies utilized this technique.[23,30,34,40] By making trade-offs between hypothetical scenarios, the relative importance of an individual attribute of a product or service and the willingness-to-trade between the different attributes can be estimated. While only one of the conjoint analysis studies incorporated an element of risk into its design, Ratcliffe et al. showed that experiencing treatment side effects was not significant, and that other aspects of asthma management (i.e. symptom relief and an improved doctor's visit) were more important to patients.[34] However, in this study the side effects (weight gain, throat irritation, voice changes) were grouped as a single attribute, which may have made it challenging for respondents to make a true trade-off between potential risks and benefits. Furthermore, the side effects selected may not have been the most important for patients. Providing services and asthma management consistent with patients preferred treatments has been shown to improve treatment adherence, asthma control, and quality of life.[18] If patients feel that their treatments have unfavourable aspects (such as inconvenience, complications, etc.), they are less likely to adhere to it. As no treatment regimen is likely to be 'perfect', it is important to consider patients' preferences when developing a management plan. However, given that the majority of the studies reviewed did not use a preference-based technique, caution is needed when interpreting the results for application to asthma management strategies. There is an increasing need to understand the process of making trade-offs between different aspects of drug therapy, in terms of treatment-related risks and benefits. It is important to understand the preferences for individual treatment attributes; however determining i f receiving a benefit or avoiding a risk is preferred may be more important for improving asthma control and quality of life for patients. While the contingent valuation by O'Conor is the only study evaluating the patients' willingness to accept risk in exchange for reduced asthma symptoms,[21] the lack of a description for the risk component may have -44 -made it difficult for the patients to appropriately assess the risks and make the corresponding risk-benefit trade-offs. As such, an instrument needs to be developed to accurately represent the reality of making decisions involving the trade-offs between positive and negative attributes with a particular focus on the assessment of specific treatment side effects. This will provide further understanding for incorporating patients' preferences in clinical decision-making. -45 -3.5 Tables Table 3.1: Characteristics of studies evaluating patients' preferences for different aspects of asthma treatments Authors, Year Country Population* Study Design Method Boe et al.[19], 1992 Norway Adults using ongoing treatment with ICS and S A B A Crossover Self-administered, study-specific questionnaire LaForce et al.[20], 1993 U S A Adolescents and adults using S A B A daily Cohort Self-administered, study-specific questionnaire O 'Conore t al.[21], 1997 Sweden Adults with physician-diagnosed asthma Cross-sectional Contingent valuation Majahan et al.[22], 1997 U S A Adolescents and adults with F E V | of 50-80% Cohort Self-administered, study-specific questionnaire .Balsbaugh et al.[23], 1999 U S A Adults with physician-diagnosed asthma Cross-sectional Conjoint analysis Weinberg et al.[24], 2000 South Africa, U K , Finland, Czech Republic Adolescents with F E V ] of greater than 75% Crossover Self-administered, study-specific questionnaire Maspero et al.[25], 2001 Argentina, Colombia, Israel, Mexico , Hungary, U S A Children with F E V , of 60-85% Cohort Previously validated questionnaires Sheth et al.[26], 2003 U S A Adolescents and adults using ongoing treatment with ICS and S A B A Crossover Self-administered, study-specific questionnaire Bukstein et al.[27], 2003 U S A Children with F E V , of 60-85% Crossover •Self-administered, study-specific questionnaire Liam et al.[28], 2004 Malaysia Adolescents and adults using ongoing treatment with ICS and S A B A Cohort Self-administered, study-specific questionnaire Welch et al.[29], 2004 U S A Adults with stable, mild to moderate asthma Crossover Self-administered, study-specific questionnaire Johansson et al.[30], 2004 Sweden Adults using ongoing treatment with ICS and S A B A Cross-sectional Conjoint analysis Weiss et al.[31], 2005 U S A , Sweden Adults with F E V , of 40-90% Cohort Self-administered, study-specific questionnaire *FEV,: forced expired volume in the first second; ICS: inhaled corticosteroids; S A B A : short-acting P-agonist. -46 -Table 3.2: Characteristics of studies evaluating patients' preferences for management, decision-making, and symptoms in asthma Authors, Year Country * Population Study Design Method Asthma Management Jones et al.[32], South Adolescents and adults Cross- Focus groups 2000 Wales with physician-diagnosed asthma sectional Douglass et Australia Adults who presented at Cross- Semi-al.[33], 2002 emergency department with asthma over a two month period sectional structured in-depth interviews Ratcliffe et U K Adults with physician- Cross- Conjoint al.[34], 2002 diagnosed asthma sectional analysis Haughney et U K Adults with mild to Mixed- Qualitative al.[35], 2004 moderate asthma method structured interview Decision-making Gibson et Australia Adults with mild to Cross- Validated al.[36], 1995 severe asthma sectional questionnaire Adams et Australia Adults with moderate to Cross- Conversation-al.[37], 2001 severe asthma sectional style interviews Caress et U K Adults with mild to Cross- Validated al.[38],2002 moderate asthma sectional questionnaire Caress et U K Adults with physician- Cross- Validated al. [39], 2005 diagnosed asthma sectional questionnaire Symptom Preferences Osman et U K Adults with moderately Cross- Conjoint al.[40], 2001 severe asthma sectional analysis - 4 7 -Table 3.3: Significant findings from treatment preferences studies Authors Sample Si/e . Findings* Hypothetical Treatment Attributes Balsbaugh et al. [23] 394 Patients are not willing to use medications that require blood test monitoring and high dosing frequency. Johansson et al.[30] 298 79% preferred a reliever that is both rapid and long acting and 25% preferred combination inhaler for both maintenance and reliever use. The monthly WTP was US$36. O'Conor et al.[21] 216 Annual WTP for complete asthma control is US$1000 to US$3700. Specific Inhaler Type Boeetal.[19] 179 DPI is preferred to MDI with and without a spacer (p<0.05). Sheth et al.[26] 154 60% preferred Diskus® (DPI) overall compared to MDI (p<0.025). Welch et al.[29] 114 Significant preference for Pulmicort Turbuhaler® (DPI) compared to MDI's (pO.OOl). Liam et al.[28] 48 71% preferred Accuhaler® (DPI) over MDI. Weiss et al.[31] 925 Budesonide DPI is preferred to triamcinolone acetonide MDI (pO.OOl). LaForce et al.[20] 1235 DPIs are an acceptable alternative to MDIs. Mahajan et al.[22] 213 61% preferred the Diskus® inhaler, 25% preferred the Diskhaler® inhaler, and 13% expressed no preference. Administration Modes Weinberg et al.[24] 113 70% preferred the orally-administered Accolate® to inhaled beclomethasone dipropionate. Bukstein et al.[27] 333 82% preferred the orally-administered montelukast (82%) over the inhaled cromolyn. Maspero et al. [25] 124 Significant satisfaction for orally-administered montelukast over inhaled beclomethasone was observed (p = 0.001). *DPI: dry powder inhaler; MDI: metered dose inhaler; WTP: willingness-to-pay -48 -Table 3.4: Significant findings from asthma management studies Authors Sample Size Findings Jones et al.[32] 68 Almost all participants were ambivalent about usefulness or relevance of action plans. Most health professionals opposed the use of action plans. Douglass et al.r.331 62 Patients viewed action plans positively. Haughney et al.[35] 40, 517 81% preferred to use fewer inhalers, 68% preferred to have action plans. Ratcliffe et al.[34] 142 Patients preferred: treatments with a greater potential to relieve symptoms (p = 0.40), a doctor to treat them as a whole person (p = 0.23), seeing the same doctor during the course of a treatment (P = 0.15), more time with the doctor (p = 0.14), and lower travel costs to attend asthma consultation (P = -0.01) Table 3.5: Significant findings from decision-making studies Authors Sample Size Findings Gibson et al.[36] 85 Asthmatics want to be informed about their condition but they do not wish to be the prime decision makers during an exacerbation. Adams et al.[37] 212 Patients preferred clinicians to assume the major role in most decision-making about their management; however, patients wished to remain in control in choosing when to seek care. Caress etal.[38] 32 Most patients wished to contribute to or feel involved in treatment decision-making but not necessarily to control it Caress et al. [39] 230 24% preferred an active role, 36% a collaborative role, 40% a passive role. - 4 9 -3.6 References 1. World Health Organization: World Health Report 2003: Shaping the future [Online]. [2003?] [cited 2006 Apr 24]; Available from: URL: http://wvvvv.who.int/whr/2003/en/index.html 2. Boulet L-P, Becker A , Berube D, Beveridge R, Ernst P. Canadian asthma consensus report, 1999. Can Med Assoc J 1999;161 :S1-S62. 3. Global Initiative for Asthma. Asthma management and prevention: a practical guide for public health officials and health care professionals. Imperial College, London: GINA, 2001. 4. Pearce N , Grainger J, Atkinson M , Crane J, Burgess C, Culling C, et al. Case-control study of prescribed fenoterol and death from asthma in New Zealand, 1977-1981. Thorax 1990;45:170-5. 5. Anis A , Lynd L, Wang X , King G, Spinelli JJ, FitzGerald M , et al. Double trouble: inappropriate asthma medication use linked to increased use of health care resources. Can Med Assoc J 2001;164:625-31. 6. Lynd LD, Guh DP, Pare PD, Anis A H . Patterns of inhaled asthma medication use: a 3-year longitudinal anlaysis of prescription claims data from British Columbia, Canada. Chest 2002;122:1973-81. 7. Suissa S, Ernst P, Benayoun S, Baltzan M , Cai B. Low-dose inhaled corticosteroids and the prevention of death from asthma. N Engl J Med 2000;343:332-6. 8. Little P, Everitt H , Williamson I, Warner G, Moore M , Gould C, et al. Observational study of effect of patient centredness and positive approach on outcomes of general practice consultations. B M J 2001;323:908-11. 9. Moher D. CONSORT: an evolving tool to help improve the quality of reports of randomized controlled trials. Consolidated standards of reporting trials. J A M A 1998;279:1489-91. 10. Altman DG, Schulz K F , Moher D, Egger M , Davidoff F, Elbourne D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med 2001;134:663-94. 11. Von Neuman J, Morgensterm O. Theory of games and economic behaviour. New York Wiley; 1953. 12. Torrence GW, Thomas WH, Sackett DL. A utility maximization model for evaluation of health care programs. Health Serv Res 1972;7:119-33. 13. Maas A , Stalpers L. Assessing utilities by means of conjoint measurement: an application in medical decision analysis. Med Decis Making 1992;12:288-97. - 5 0 -14. Volovitz B, Dunas-Meza E, Chmielewska-Szewczyk DA, Kosa L, Astafieva NG, Villaran C, et al. Comparison of oral montelukast and inhaled cromolyn with respect to preference, satisfaction, and adherence: a multicenter, randomized, open-label, crossover study in children with mild to moderate persistent asthma. Curr Ther Res 2000;61:490-506. 15. Abramson M , Bailey M , Couper F, Driver J, Drummer OH, Forbes A, et al. Are asthma medications and management related to deaths from asthma? Am J Respir Crit Care Med 2001;163:12-18. 16. Gibson PG, Powell H , Coughlan J, Wilson AJ, Abramson M , Haywood P, et al. Self-management education and regular practitioner review for adults with asthma. Cochrane Database Syst Rev 2003:CD001117. 17. Degner LF, Sloan JA, Venkatesh P. The Control Preferences Scale. Can J Nurs Res 1997;29:21-43. 18. Guadagnoli E, Ward P. Patient participation in decision-making. Soc Sci Med 1998;47:329-39. 19. Boe B, Stiksa G, Svensson K, Asbrink E. New method of evaluating patient preference of different inhalation delivery system. Ann Allergy 1992;68:255-60. 20. LaForce CF, Ellis EF, Kordansky DW, Cocchetto D M , Sharp JT. Use and acceptance of ventolin Rotacaps® and the Rotahaler® in 1235 asthmatic patients. Clin Ther 1993;15:321-9. 21. O'Conor R M , Blomquist GD. Measurement of consumer-patient preferences using a hybrid contingent valuation method. J Health Econ 1997;16:667-83. 22. Majahan P, Okamoto L. Patient satisfaction with the Diskhaler and the Diskus inhaler, a new multidose powder delivery system for the treatment of asthma. Clin Ther 1997;19:1126-34. 23. Balsbaugh TA, Chambers CV, Diamond JJ. Asthma controller medications: what do patients want? J Asthma 1999;36:591-6. 24. Weinberg EG, Naya I. Treatment preferences of adolescent patients with asthma. Pediatr Allergy Immunol 2000; 11:49-55. 25. Maspero JF, Duefias-Meza E, Volovitz B, Pianacho Daza C, Kosa L, Vrijen F, et al. Oral montelukast versus inhaled beclomethasone in 6- to 11-year-old children with asthma: results of an open-label extension study evaluating long-term safety, satisfaction, and adherence with therapy. Curr Med Res Opin 2001;17:96-104. 26. Sheth K, Bernstein JA, Lincourt WR, Merchant K K , Edwards LD, Crim CC, et al. Patient perceptions of an inhaled asthma medication administered as an inhalation -51 -powder via the Diskus or as an inhalation aerosol via a metered-dose inhaler. Ann Allergy Asthma Immunol. 2003;1:55-60. 27. Bukstein DA, Bratton DL, Firriolo K M , Estojak J, Bird SR, Hustad C M , et al. Evaluation of parental preference for the treatment of asthmatic children aged 6 to 11 years with oral montelukast or inhaled cromolyn: a randomized, open-label, crossover study. J Asthma 2003;40:475-85. 28. Liam CK, Lim K H , Wong C M M . Acceptance of the accuhaler, a multi-dose powder inhaler, among asthmatic patients: a comparison with the pressurized metered-dose inhaler. Asian Pac Allergy Immunol 2000;18:135-40. 29. Welch MJ , Nelson HS, Shapiro G, Bensch GW, Sokol WN, Smith JA, et al. Comparison of patient preference and ease of teaching inhaler technique for pulmicort Turbuhaler® versus pressurized metered-dose inhalers. J Aerosol Med 2004;17:129-39. 30. Johansson G, Stallberg B, Tornling G, Andersson S, Karlson G, Fait K, et al. Asthma treatment preference study: a conjoint analysis of preferred drug treatments. Chest 2004;125:916-23. 31. Weiss K B , Paramore L C , Liljas B, Revicki DA, Luce BR. Patient satisfaction with budesonide TurbuhalerTM versus triamcinolone acetonide administered via pressurized metered-dose inhaler in a managed care setting. J Asthma 2005 ;42:769-76. 32. Jones A , Pill R, Adams S. Qualitative study of views of health professionals and patients on guided self managements plans for asthma. B M J 2000; 321:1507-10. 33. Douglass J, Aroni R, Goeman D, Stewart K, Sawyer S, Thien F, et al. A qualitative study of action plans for asthma. B M J 2002;324:1003-5. 34. Ratcliffe J, Van Haselen R, Buxton M , Hardy K, Coleman J, Partridge M . Assessing patients' preferences for characteristics associated with homeopathic and conventional treatment of asthma: a conjoint analysis study. Thorax 2002;57:503-8. 35. Haughney J, Barnes G, Partridge M , Cleland J. The living and breathing study: a study of patients' views of asthma and its treatment. Primary Care Respir J 2004;13:28-35. 36. Gibson PG, Talbot PI, Toneguzzi RC. Self-management, autonomy, and quality of life in asthma. Chest 1995;107:1003-8. 37. Adams RJ, Smith BJ, Ruggin RE. Patient preferences for autonomy in decision-making in asthma management. Thorax 2001;56:126-32. - 5 2 -Caress A - L , Luker K, Woodcock A, Beaver K. A qualitative exploration of treatment decision-making role preference in adult asthma patients. Health Expect 2002;5:223-39. Caress A - L , Beaver K, Luker K, Campbell M , Woodcock A. Involvement in treatment decisions: what do adults with asthma want and what do they get? Results of a cross sectional survey. Thorax 2005;60:199-208. 40. Osman L M , McKenzie L, Cairns J, Friend JAR, Godden DJ, Legge JS, et al. Patient weighting of importance of asthma symptoms. Thorax 2001;56:138-42. -53 -C H A P T E R 4 A N E V A L U A T I O N O F PATIENTS' P R E F E R E N C E S IN A S T H M A T H E R A P Y USING A DISCRETE C H O I C E E X P E R I M E N T * 4.1 Introduction Asthma is a prevalent and chronic respiratory condition that afflicts over 8% o f Canadian adults [1]; globally, prevalence has been increasing steadily over the past few decades.[2] B y using management strategies from published guidelines, a person with asthma can achieve a high degree o f control and a nearly perfect quality o f life (QOL).[3-5] However, previous studies have revealed that many asthma patients are not treated in accordance with these guidelines; even i f they are prescribed the appropriate treatment, patients may not adhere to their management strategy.[6-9] Furthermore, patients of lower socioeconomic status (SES), independent of disease severity, are more likely to have suboptimal asthma control.[10,11] Although the factors leading to inappropriate management have not yet been elucidated, potential factors include the patient's perception of treatment side effects, the patient's SES, or the patient and physician's knowledge or preferences for different aspects of asthma treatments. Evidence indicates that considering patients' preferences in treatment decisions can improve disease management.[12] Market research techniques, such as conjoint analysis and discrete choice experiments (DCEs), are seeing increasing use in preference elicitation in healthcare.[13-17] D C E s can evaluate the individual's preferences for different treatment attributes using hypothetical treatment scenarios; thus they are effective for eliciting preferences and understanding the components involved in decision-making. Patients reveal their preferences by assessing the available options and then trading-off between the perceived advantages and drawbacks by simultaneously considering varying levels of different treatment attributes. *A version of this chapter will be submitted to Thorax. McTaggart-Cowan, H . M . , Shi, P., FitzGerald, J .M. , Anis, A . H . , Kopec, J.A., Bai, T.R., Currie, G., Soon, J.A., Lynd, L .D. An Evaluation of Patients' Preferences in Asthma Therapy Using a Discrete Choice Experiment. - 5 4 -The premises behind DCEs are: (i) the investigated treatment can be broken down into specific attributes, for which a range of values, or levels, can be defined; (ii) every respondent has a unique preference (utility) for each attribute level; and (iii) these utilities, can be combined across attributes. The attribute levels can directly impact the individual's preferences, as the attractiveness of the treatment option may depend on the level of a specific attribute. The results from the analysis identify the relative importance of the treatment attributes. To include patients' preferences in clinical decision-making for asthma treatments, accurate information is needed regarding the nature of patients' preferences for attributes such as cost, efficacy, ease of use, and potential for adverse events. This study was conducted to evaluate the importance of such treatment attributes in asthma patients using a DCE. Specifically, we wanted to assess the number of symptom-free days (SFDs) a patient was willing to forego in exchange for a reduction in treatment-related risks. A further objective was to evaluate the influence of SES on treatment preferences. 4.2 Methods 4.2.1 Patient Recruitment and Study Sample We conducted a cross-sectional study composed of 157 English-speaking patients with physician-diagnosed asthma, between 19 and 49 years of age, with no other concurrent respiratory conditions, and residing in the Vancouver metropolitan area in British Columbia (BC), Canada. A general poster campaign recruited 114 (73%) participants, a further 17 were recruited from respiratory clinics, and the remainder was recruited by community pharmacists. Each subject was assessed at a pulmonary research clinic and received $20 to defray time and travel costs. The Research Ethics Boards of University of B C , Providence Health, and Fraser Health approved the study protocol (Appendix I), and informed written consent was obtained from each participant (Appendix II). 4.2.2 Discrete Choice Experiment We administered a DCE to elicit patients' preferences for six specific attributes of asthma therapy. While the number of attributes used in the literature varies,[18] a maximum of six attributes has been suggested to avoid excessive cognitive complexity for the - 5 5 -respondent.[19] In consultation with an expert panel, consisting of respirologists and asthma educators, the most important treatment attributes selected were the number of monthly SFDs, monthly out-of-pocket cost for medications, frequency of administration, number of required inhalers, and the number of oral thrush and tremors/heart palpitation episodes. Upon further discussion, the range of possible values for each attribute was broken down into either three or four levels (Table 4.1). Treatment profiles were developed based on combinations of attribute levels. A set of 25 pairs of profiles were assembled through a fractional factorial design using a design catalogue. [20] Patients were asked to choose between these pairs of hypothetical treatment scenarios; inclusion of a third choice, 'neither', provided a no treatment option. Following development, the questionnaire was tested for orthogonality, using Kendall's X B test of association, to ensure that there was minimal correlation between attributes; level balance, to verify that levels occurred at equal frequency; and minimum overlap, to check that the attributes did not appear at the same level with the same choice. One extra treatment scenario was included with a dominant option, where attribute levels of one profile were better than the alternative, to test the validity of the responses. An example of one choice set is shown in Figure 4.1 (the complete DCE is provided in Appendix III). An initial pilot study of 30 participants assessed D C E comprehension and the validity of the treatment attributes. 4.2.3 Patient Information Data pertaining to demographic information and asthma status were obtained using a questionnaire from a previous study and included: SES, quantified using self-reported annual household income, years of post-secondary education, highest education attained, and receipt of social assistance; and annual utilization of all asthma-related drugs prior to the evaluation (Appendix IV). [10] Spirometry was measured in accordance with American Thoracic Society criteria.[21,22] The forced expired volume in the first second (FEVi) was expressed as a percent of the predicted value based on the individual's age, height, and sex. As the individual's QOL may influence treatment preferences, the standardized Asthma Quality of Life Questionnaire (AQLQ(S)) was also administered (Appendix V).[23-25] This valid and reliable measure of asthma-specific QOL encompasses four domains (symptoms, activity limitation, emotional function, and environmental stimuli) and consists of 32 questions each scored on a seven-point Likert scale ranging from one (worst - 56 -possible QOL) to seven (best possible QOL). The average score of all questions yields overall QOL score. 4.2.4 Statistical Analysis Relative Preferences A multinomial discrete choice model, which uses the maximum likelihood method for estimation, allows for the analysis of scenarios consisting of multiple alternatives. [26] The random utility function is used to explain the stated preference data, where the utility of a specific choice can be interpreted as the relative preference that the decision-maker derives from that choice relative to the other alternatives in a finite choice set. For this current work, dummy-coded variables were created for each level of all categorical attributes; only the monthly number of SFDs and treatment cost were evaluated on a continuous scale. The resulting model is: Uij ~ P.SFDX],ij + Po>.siX2,ij + Pooxe\X3,ij + PDoxe2X + Ppose3X5Jj + P'/hnxh\X6,ij + Prhrush2Xl,ij + P'fhruxhiXS,ij + Prrcmi,r]X9,ij ' (4-1) + Prremor2X\Qjj PrremoriX\\Jj + P lnhaler2X\2,ij + P Inhaler yX\i,ij where PSFD denotes the preference for monthly symptom free days, Pcost the preference for treatment cost, Poose* the preference for a treatment requiring x daily dose(s) relative to a treatment requiring as needed dosing (where x = 1, 2, 3), pThrusiu the preference for a treatment with x annual episode of oral thrush relative to a treatment with no annual oral thrush episodes (where x = 1, 2, 3), Piremoa the preference for a treatment with x monthly episode of tremors/heart palpitations relative to a treatment with no monthly tremors/heart palpitations episodes (where x = 1, 2, 3), Pinhaiea the preference for a treatment requiring x inhalers relative to one inhaler (where x = 2, 3), and e,y is the unobservable error term of individual / for choice j-This analysis is dependent on the cross-elasticities of substitution across the alternatives. The cross-elasticities of substitution, measured by an inclusive value, reflect the changes in the probability of choosing an alternative as a function of changes in the level of utility from competing alternatives.[27] Based on random utility theory, the nested logit model (NLM) assumes treatments A and B are closer substitutes with each other than with the 'neither' option; however, the opt-out option is still a potential substitute for A and B. Thus, - 5 7 -an increase in the probability of choosing A results in a greater decrease in the probability of choosing B than of choosing no treatment. The two components in this decision-making process therefore are first deciding whether or not to be treated and then, assuming the individual decides to be treated, which treatment to choose.[27] An inclusive value between zero and one suggests that the decision of whether or not to be treated is influenced by the expected utility of being treated with either treatment A or treatment B relative to no treatment, thereby representing the appropriateness of the N L M model. [28] We hypothesized that the covariates that influence the preference to not be treated (ultimately yielding a 'neither' response) were age, sex, AQLQ(S) score, and receipt of social assistance. This N L M model is represented as: Not Treated = K + SAgB + SSex + SAQLQ(S) + SAssilllance + UjJ, (4.2) where Not Treated represents the preference to not be treated (where one represents the selection of the 'neither' option and zero represents a preference to be treated), K is the constant, 5 A g e is the preference to not be treated influenced by being older, 5sex is the preference to not be treated influenced by being female, SAQLQ(S) is the preference to not be treated influenced by a higher AQLQ(S) score, §Assistance is the preference to not be treated influenced by receiving social assistance from the provincial government (individuals with annual incomes of less than $20,000), and uy denotes the unobservable error term of individual i for choice j. Odds ratios and 95% confidence intervals are reported for each association. Marginal Rate of Substitution The regression coefficients (/J) of the N L M (equation 4.1) are used to evaluate the relative preferences and the marginal rates of substitution (MRS) for all attributes. The MRS represents the rate at which individuals are willing to forego amounts of one specific attribute in exchange for greater amounts of another attribute: MRS = -1B A H attribute 1 (4.3) \^ Pattribute 2 J Using the regression coefficient for cost in the denominator of equation 4.3 determines the willingness-to-pay (WTP) for each attribute. [29] - 5 8 -Sub-population Preferences Patients' preferences may differ across levels of SES: evaluating preferences for sub-populations determines preference variations with income and education levels. Since social class influences asthma control,[10,11] variations in treatment preferences with medication use was also explored. In accordance with asthma management guideline definitions, we identify good asthma control as using fewer than four short-acting (SA) p-agonist canisters annually.[3-5] To evaluate the presence of a differential in preferences, interaction terms were created between the cost attribute and income, the risk attributes thrushx and tremorx and education level, and the SFD attribute and the magnitude of SA P-agonist use. The population was subdivided by annual income (three groups, 'up to $20,000', '$20,000 to $50,000', and 'greater than $50,000'), education level (two groups, 'less than high school, high school and/or trades diploma' and 'at least a university Bachelor's degree') and asthma control (defined by annual SA P-agonist use, 'four or fewer canisters', 'five to 12 canisters', and '13 or more canisters'). The mean WTP were estimated for all interaction terms to evaluate whether or not treatment preferences for cost, frequency of adverse events, and number of SFDs are influenced by an individual's annual income, education, and asthma control, respectively. A l l analyses were performed using the SAS statistical software package, version 8.2 (SAS Institute, Cary, NC). Significance was defined as p<0.05. 4.3 Results 4.3.1 Sample Characteristics Table 4.2 displays information regarding demographic and asthma-specific variables of the study participants. The sample included all levels of disease control and severity. Twenty-eight patients either did not know or preferred not to provide their annual household income and were therefore excluded from sub-population analyses involving incomes. Forty-three patients (28%) reported incomes of greater than $70,000, of which 24 were greater than $100,000. One hundred twenty-nine (83%) and 65 patients (43%) completed at least one and five year(s) of post-secondary education, respectively. The mean value of 5.4 on the AQLQ(S) showed that the sample had a moderately high QOL. Overall, the sample had a high SES (55% with a university degree, 56% with an annual income of greater than $50,000) and well-- 5 9 -controlled asthma, as indicated by 70% annually using fewer than four SA P-agonists canisters. 4.3.2 Unadjusted Model As there were 26 choices in the DCE and 157 participants, a total of 4082 observations were possible in the study; however, after removing the five missing responses and the question included as the non-design consistency check, the final analysis consisted of 3920 observations. One hundred and fifty-four (98%) respondents answered the consistency check correctly, demonstrating a high level of comprehension for the questionnaire design. The inconsistent responses are included in the final analysis as removal of such responses may lead to biased results.[30] A l l attribute coefficients had the appropriate signs and are generally in the hypothesized directions, which contributes to the overall validity of the results. This demonstrates that the patient's ideal treatment would result in more SFDs, fewer daily doses and inhalers, lower cost, and fewer episodes of oral thrush and tremors/heart palpitations. The inclusive value of 0.40 indicates that the N L M is appropriate for the analysis. Statistically significant covariates affecting patients' preferences to not have their asthma treated are being female, older, having a higher AQLQ(S) score, and receiving social assistance (Table 4.3). This indicates that individuals who are male, younger, have poorer asthma QOL, or are not receiving social assistance have greater preferences for treatment of their asthma. For those patients who preferred to be treated, all treatment attributes are significant predictors of choice except for the preference to take one daily dose compared to as-needed dosing (p = 0.09) (Table 4.4). The insignificance of this attribute indicates that there is an indifference in preference between taking a medication once daily versus on an as needed basis. For the categorical attributes, there are linear gradients between all attribute levels except for the frequency of tremor episodes and the number of inhalers. Insignificant differences are observed between a treatment regimen that will result in one and two episodes of tremors/palpitations (p = 0.73), and a treatment regimen requiring two and three inhalers (p = 0.84). This suggests that patients are indifferent to whether a treatment induces one or two monthly episodes of tremor/palpitations; however, there is a strong preference for none over any tremor/palpitation episodes (p<0.001) and a strong aversion to three events over one or - 6 0 -two (pO.OOl). Patients prefer treatments requiring once daily administration relative to higher frequencies, and a single rather than multiple inhalers. There is no difference in preferences between treatment regimens with two or three inhalers. Participants are willing to pay, on average, $14 per month for a treatment that results in one additional SFD per month. For the thrush attribute, the monthly WTP is $26 to avoid one annual episode, $79 to avoid two annual episodes, and $112 to avoid three annual episodes, when compared to a treatment that does not result in any episodes. The results demonstrate a willingness-to-trade asthma control for a reduction in adverse events and in the frequency of daily administration. Specifically, patients are willing to forego 1.8, 5.5, and 7.9 monthly SFDs to avoid one, two, and three annual episodes of oral thrush, respectively. 4.3.3 Sub-population Preferences Stratifying the population by income revealed that people with a higher income had a higher WTP for all beneficial treatment attributes compared to the low and moderate incomes groups. This overall trend is in the expected direction and provides support for the theoretical validity of the WTP estimates (Table 4.5). Using education as a proxy for SES results in similar findings, in that participants with university degrees consistently have higher WTP for treatments with lower frequencies of adverse events (thrush and tremor/heart palpitations) versus people without university degrees (Table 4.6). While the MRS for individuals with university degrees are statistically different from those without university degrees, a WTP of $0 is unexpectedly observed for individuals without university degrees for all levels of the adverse events. In general, individuals using more SA P-agonists have a greater preference for a treatment that results in more monthly SFDs compared to individuals using fewer SA P-agonists (Table 4.7). There was no significant difference between patients using four or fewer and five to 12 SA P-agonist canisters annually; however, those using more than 13 had a significantly higher WTP. 4.4 Discussion This study evaluates patients' preferences for their asthma treatments using a DCE. We found that individuals who are male, younger, have poorer asthma QOL, or are not receiving social assistance have greater preferences for treatment of their asthma, compared to those -61 -who are female, older, have better asthma QOL, or are receiving social assistance. Overall, the results reveal that asthma patients prefer a treatment regimen that will result in more monthly SFDs but are willing to forego them in exchange for a reduction in adverse event frequency or treatment convenience. Treatment preferences vary across sub-populations, such that individuals with lower incomes, less education, and better asthma control have a lower WTP to receive benefits and avoid risks relative to those with higher incomes, more education, and poorer asthma control. The study results add to the current literature, as only a few published studies have used DCEs to investigate patients' preferences for asthma treatments; however, none of the identified studies included a risk component.[31,32] As patients' risk attitudes may contribute to treatment preferences; failing to include risk thereby reduces the realism of the simulated decision process. Our results suggest that patients attach a significant value to the avoidance of adverse events. Although the probabilities associated with experiencing thrush and tremor/heart palpitation episodes are low, the patients are still willing to trade symptom relief for a reduced risk of adverse events. Preferences for treatment attributes differ between sub-populations. Socioeconomic differences figure prominently in overall health and in asthma control [10]; lower SES asthma patients experience more frequent hospital admissions, emergency room visits, and physician visits.[7,33,34] Using income and education as proxies for SES, a greater WTP for more treatment-related benefits and less adverse events are associated with individuals with higher income and completion of a university degree. Our results reveal that, on average, individuals without university degrees had a WTP of $0 for all levels of the investigated adverse events. The causes of this null effect are unclear, and may only be identifiable through in depth interviews. We also propose that if an individual is already well-controlled, more SFDs are not as important as reductions in adverse events or improvements in convenience (i.e. daily doses and number of inhalers). As over two-thirds of the study population reported using fewer than four canisters of SA (3-agonists annually, they are identified as well-controlled. On average, patients have a monthly WTP of $14 for one SFD; for an uncontrolled patient, the monthly WTP would significantly increase for each additional SFD desired. While this value could potentially be quite high, this may be related to patients not having to pay the full cost of their - 6 2 -medications due to various pharmaceutical coverage programs. In the province of BC, there is a drug co-payment plan, which provides comprehensive first-dollar coverage for all residents on social assistance and the general population (less than 65 years of age) whose annual family pharmaceutical expenditure excess $800.[35] Between this co-payment plan and drug coverage from private insurance agencies, many BC residents have no direct experience of paying the full cost of their prescription medications. Even if this effect biases the absolute magnitudes of the WTP results, it should not affect the relative weightings of the individual treatment attributes. A linear and proportionate gradient is not always observed between each attribute level; this implies that patients do not interpret the attributes in a linear and continuous manner. Furthermore, the evaluation of risk is investigated using specific variables (i.e. number of oral thrush and tremor/palpitation episodes). A previous asthma DCE used an attribute related to collective side effects [36]; however, it is unlikely that individuals will have the same preferences for different potential risks. Potential limitations of the current study are primarily associated with the questionnaire design. Evidence suggests that individuals can cope with between nine and 16 pairwise comparisons before they become fatigued or disinterested; in this study, participants responded to 26 scenarios.[19] Although respondent burden was initially a concern, pilot study respondents did not identify this as a problem. This concern is further alleviated by the lack of missing responses, the 98% internal consistency rate, the fact that the preferences are in the hypothesized directions, and the consistency of results across different measures of SES. The use of patient self-report for income level may be complicated by varying interpretations of household and indirect assets; the high non-response rate for this variable is also a concern. The use of population SES measures, such as neighbourhood income, could overcome missing values by providing a less precise but more objective measure. Although the study sample was heterogeneous and encompassed all levels of disease control and severity, it may not accurately reflect the preferences of poorly-controlled patients. Despite these potential limitations, the results demonstrate that our DCE is capable of evaluating patients' preferences for asthma treatments and analyzing preference variations in different sub-populations. Overall, the study reveals that while people with asthma desire treatments that offer more benefits, they are willing to accept some decrement in benefits to -63 -avoid potential adverse events; the magnitude of these preferences vary with SES. Specifically, people of lower SES have a lower WTP for increasing SFDs or avoiding adverse events than do people with higher SES. These findings provide useful information for the development of targeted asthma education and treatment programs, particularly for the design of individualized management plans aimed at individuals with suboptimal asthma control. Furthermore, these results will provide clinicians with insight into desirable asthma treatment attributes, thereby allowing them to help select management plans more aligned with the preferences of their individual patients. By thus incorporating patients' preferences in clinical decision-making, treatment adherence may be improved. -64 -4.5 Tables Table 4.1: Attributes and levels used in the study Treatment Attribute Variable Name Levels Number of SFDs per month SFD 10, 15,20,25 or more Daily dosage of medication Dosex as needed, 1, 2, 3 or more Out-of-pocket cost per month ($)* Cost 20, 40, 80, 160 Number of thrush episodes per year Thrushx none, 1, 2, 3 Number of tremors/palpitation episodes per month Tremor* none, 1, 2, 3 or more Number of inhalers used daily Inhalerx 1,2,3 *Out-of-pocket cost in Canadian funds -65 -Table 4.2: Demographic and asthma-specific information related to the participants in the survey Parameter* Frequency (%) or Mean (+SD)T Age 35.0 (7.9) Female 110(70.1) FEV, 90.9(19.9) AQLQ(S) Score (1-7) 5.4(1.1) Highest education level attained Less than high school, high school diploma, trades 70 (45.5) diploma University degree 84 (54.6) Years of post-secondary education completed Less than 1 25 (16.5) 1-4 61 (40.1) Greater than 5 65 (43.3) Annual household income Less than $20,000 22(17.1) $20,000-50,000 35 (27.1) Greater than $50,000 72 (55.8) Receiving social assistance No 135 (87.7) Yes 18 (11.8) Number of short-acting (3-agonist canisters used in the past year None 15 (9.6) 1-4 95 (60.5) 5-12 35 (22.3) 13 or more 12 (7.6) Number of ICS canisters used in the past year None 63 (40.6) 1-4 56 (36.1) 5-12 28 (18.1) 13 or more 8 (5.2) Frequency of long-acting (3-agonist use Never 139 (89.1) Less than once per week 8(5.1) Everyday 9 (5.8) AQLQ(S): standardized Asthma Quality of Life Questionnaire; F E V , : forced expired volume in the first second; ICS: inhaled corticosteroid. ^ D : standard deviation. - 6 6 -Table 4.3: Patient characteristics affecting the preference to not have their asthma treated Parameter Regression Coefficient (SE)* Odds Ratio (95% CI) Constant -0.57 (0.32) Female 0.64 (0.09) 1.90 (1.01,3.55) Age 0.02 (0.01) 1.02 (1.00, 1.04) AQLQ(S) score 0.11 (0.04) 1.12 (1.08, 1.21) Receipt of social assistance 1.08 (0.17) 2.95 (2.11,4.11) *SE: standard error. A l l parameters p<0.005, except constant p=0.08. ^CI: confidence interval. - 6 7 -Table 4.4: Relative preferences and marginal rates of substitution for the nested logit model* Treatment Regression Marginal Rates of Substitution Attributes ' Coefficient ( S E ) + 4 . Willingness-to- Willingness-to-pay trade S F D 1 ($/month) § Symptom-free day 0.18 (0.01) Ref 14.31 Cost -0.012 (0.001) 0.07 Ref Frequency of administration As needed Ref Ref Ref Once per day -0.16(0.09) 0.89 -12.76 Twice per day -0.65 (0.10) 3.64 -52.05 Three times per -1.24 (0.12) 6.96 -99.69 day Number of oral thrush episodes None Ref Ref Ref One annual -0.32 (0.09) 1.80 -25.80 episode Two annual -0.98 (0.16) 5.52 -79.00 episodes Three annual -1.39(0.15) 7.82 -112.00 episodes Number of tremor/heart palpitation episodes None Ref Ref • Ref One monthly -0.33 (0.11) 1.89 -27.02 episode Two monthly -0.31 (0.13)11 1.72 -24.63 episodes Three monthly -1.42 (0.15) 7.98 -114.18 episodes Number of inhalers One inhaler Ref Ref Ref Two inhalers -0.42 (0.10) 2.34 -33.56 Three inhalers -0.38 (0.12)" 2.16 -30.90 *Log likelihood function: -3183; McFadden's log-likelihood ratio: 0.63. ^SE: standard error; SFD: symptom-free day. *A11 attributes p<0.02, except for once per day daily administration p = 0.092. §Willingness-to-pay estimates in Canadian funds. \ = 0.73 using one monthly tremor/palpitation episode as referent group. **p = 0.84 using two inhalers as referent group. - 6 8 -Table 4.5: Willingness-to-pay estimates for the stratified income model Treatment Attribute Annual Income* <S20,000 $20,000-50,000 >$50,000 Symptom-free day 12.46 12.46 16.20 Frequency of administration As needed Ref Ref Ref Once per day -10.95 -10.95 -14.24 Twice per day -41.56 -41.56 -54.07 Three times per day -83.12 -83.12 -108.12 Number of oral thrush episodes None Ref Ref Ref One annual episode -20.33 -20.33 -26.44 Two annual episodes -63.22 -63.22 -82.25 Three annual episodes -96.54 -96.54 -125.58 Number of tremor/heart palpitation episodes None Ref Ref Ref One monthly episode -22.19 -22.19 -28.87 Two monthly episodes -15.22 -15.22 -27.84 Three monthly episodes -94.07 -94.07 -122.37 Number of inhalers One inhaler Ref Ref Ref Two inhalers -28.31 -28.31 -36.85 Three inhalers -21.65 -21.65 -28.16 Income in Canadian funds. - 6 9 -Table 4.6: Marginal rates of substitution for the stratified education model ! Treatment Attribute Marginal Rates of Substitution Willingness-to-trade SFDs Willingness-to-pay ($/month) Symptom-free day Ref 9.33 Cost 0.11 Ref Frequency of administration As needed Ref Ref Once per day 2.68 -25.03 Twice per day 1.95 -18.15 Three times per day 5.01 -46.74 Number of inhalers One inhaler Ref Ref Two inhalers 1.95 -18.19 Three inhalers 1.80 -16.83 Number of oral thrush episodes Less than high school or a high school/trades diploma^ None Ref Ref One annual episode 0 0 Two annual episodes 0 0 Three annual episodes 0 0 University degree None Ref Ref One annual episode 2.09 -19.46 Two annual episodes 3.50 -32.63 Three annual episodes 4.88 -45.53 Number of tremor/heart palpitation episodes Less than high school or a high school/trades diploma^ None Ref Ref One monthly episode 0 0 Two monthly episodes 0 0 Three monthly 0 0 episodes University degree None Ref Ref One monthly episode -0.21 1.94 Two monthly episodes 2.27 -21.15 Three monthly 3.51 -32.78 episodes SFD: symptom-free day. tp<0.0001 when compared to individuals with a university degree. - 7 0 -Table 4.7: Willingness-to-pay estimates for the stratified short-acting P-agonist model Treatment Attribute* Willingness-to-Pay (S/month)T SFD Magnitude of short-acting p-agonist use 4 or less canisters per year 14.53* 5 to 12 canisters per year 12.29* 13 or more canisters per year 18.02* Frequency of administration As needed Ref Once per day -12.50 Twice per day -52.20 Three times per day -99.44 Number of oral thrush episodes None Ref One annual episode -25.85 Two annual episodes -79.15 Three annual episodes -111.74 Number of tremor/heart palpitation episodes None Ref One monthly episode -27.29 Two monthly episodes -24.85s Three monthly episodes -114.34 Number of inhalers One inhaler Ref Two inhalers -33.13 Three inhalers -30.62s *SFD: symptom-free day. +A11 p<0.02, except for once per day daily administration p = 0.09. *p<0.0001 between each level of short-acting P-agonist use. §p>0.7 using one monthly tremor/palpitation episode and two inhalers as referent group, respectively. -71 -4.6 FIGURES Figure 4.1: Example of a choice set T R E A T M E N T T R E A T M E N T Question A B Number of symptom free days (per month) 20 10 Dose (per day) 2 as needed Cost (per month) $40 $20 Thrush (episodes per year) 1 3 Tremors/palpitations (episodes per month) 3 or more 2 Number of inhalers 1 2 Which treatment would you Treatment A Treatment B Neither prefer {check one box only)? • • • - 7 2 -) 4.7 References 1. Statistics Canada. 2001 National population health survey [Online]. [2001?] [cited 2006 Jun 20]; Available from: URL: http://\vww40.statcan.ca/101/cst01/health50a.htm 2. World Health Organization: World Health Report 2003: Shaping the future [Online]. [2003?] [cited 2006 Apr 24]; Available from: URL: http://www, who. int/whr/2003/en/index. html 3. Global Initiative for Asthma. Pocket guide for asthma management and prevention. Bethesda (MD): National Institute of Health, Nov 1998. NIH Pub. No. 96-3569B. 4. Boulet L-P, Becker A , Berube D, Beveridge R, Ernst P. Canadian asthma consensus report, 1999. Can Med Assoc J 1999;161(11 Supp):Sl-62. 5. British Thoracic Society, British Paediatric Association, Royal College of Physicians of London, King's Fund Centre, National Asthma Campaign, Royal College of General Practitioners, et al. Guidelines on the management of asthma. Thorax 1993;48(Suppl 2):Sl-24. 6. Chapman K R , Ernst P, Grenville A, Dewland P, Zimmerman S. Control of asthma in Canada: failure to achieve guideline targets. Can Respir J 2001;8(Suppl A):35A-40A. 7. Anis A , Lynd L , Wang X , King G, Spinelli JJ, FitzGerald M , et al. Double trouble: inappropriate asthma medication use linked to increased use of health care resources. Can Med Assoc J 2001;164(5):625-31. 8. Lynd LD, Guh DP, Pare PD, Anis A H . Patterns of inhaled asthma medication use: a 3-year longitudinal analysis of prescription claims data from British Columbia, Canada. Chest 2002; 122(6)-.1973-81. 9. FitzGerald JM, Boulet L-P, Mclvor RA, Zimmerman S, Chapman KR. Asthma control in Canada remains sub optimal: the Reality of Asthma control (TRAC) study. Can Respir J 2006;13:253-9. 10. Lynd LD, Sandford AJ, Kelly EA, Pare PD, Bai TR, Fitzgerald JM, et al. Reconcilable differences - a cross-sectional study of the relationship between socioeconomic status and the magnitude of short-acting-agonist use in asthma. Chest 2004; 126(4): 1161-8. 11. Suissa S, Ernst P, Benayoun S, Baltzan M , Cai B. Low-dose inhaled corticosteroids and the prevention of death from asthma. N Engl J Med 2000;343(5):332-6. 12. Little P, Everitt H , Williamson I, Warner G, Moore M , Gould C, et al. Observational study of effect of patient centredness and positive approach on outcomes of general practice consultations. B M J 2001;323:908-11. 13. Streiner DL, Norman GR. Health measurement scales. A practical guide to their development and use. 2nd ed. New York: Oxford University Press; 1989. -73 -14. Von Neuman J, Morgensterm O. Theory of games and economic behaviour. New York: Wiley; 1953. 15. Torrence GW, Thomas WH, Sackett DL. A utility maximization model for evaluation of health care programs. Health Serv Res 1972;7:119-33. 16. Maas A , Stalpers L. Assessing utilities by means of conjoint measurement: an application in medical decision analysis. Med Decis Making 1992 12(4):288-97. 17. Ryan M , Bate A , Eastmond DJ, Ludbrook A. Use of discrete choice experiments to elicit preferences. Qual Health Care 2001;10(Suppl l):i55-60. 18. Ryan M , Gerard K. Using discrete choice experiments to value health care programmes: current practice and future research reflections. Appl Health Econ Health Policy 2003;2(l):55-64. 19. Ryan M . Using conjoint analysis to take account of patient preferences and go beyond health outcomes: an application to in vitro fertilization. Soc Sci Med 1999;48:535-46. 20. Kocur G, Adlter T, Hyman W, Aunet B. Guide to forecasting travel demand with direct utility assessment. Washington, DC: United States Department of Transportation; 1982. Report No.: UMTA-NH-11-0001-82-1. 21. Crapo RO, Morris A H , Gardner R M . Reference spirometric values using techniques and equipment that meet ATS recommendations. Am Rev Respir Dis 1981;123(6):659-64. 22. American Thoracic Society. Standardization of spirometry—1987 update. Statement of the American Thoracic Society. Am Rev Respir Dis 1987; 136(5): 1285-98. 23. Juniper EF, Guyatt GH, Ferrie PJ, Griffith LE . Measuring quality of life in asthma. Am Rev Respir Dis 1993;147(4):832-8. 24. Juniper E, Guyatt G, Epstein R. Evaluation of impairment of health related quality of life in asthma: development of a questionnaire for use in clinical trials. Thorax 1992;47:76-83. 25. Juniper EF, Buist AS, Cox F M , Ferrie PJ, King DR. Validation of a standardized version of the Asthma Quality of Life Questionnaire. Chest 1999; 115(5): 1265-70. 26. SAS: The M D C procedure. [2006?] [cited 2 Apr 2006]; Available from: URL: http: //support. sas. com/rnd/app/papers/mdc .pdf 27. Ryan M , Skatun D. Modeling non-demanders in choice experiments. Health Econ 2004;11:397-402. 28. Farrar S, Ryan M , Ross D, Ludbrook A. Using discrete choice modeling in priority setting: an application to clinical service developments. Soc Sci Med 2000;50:63-75. - 74 -29. Lanscar E, Louviere J. Deriving welfare measures from discrete choice experiments: inconsistency between current methods and random utility and welfare theory. Health Econ. 2004; 13(9):901-7 30. Lanscar E, Louviere J. Deleting 'irrational' responses from discrete choice experiments: a case of investigation or imposing preferences? Health Econ 2006;15:797-811. 31. Balsbaugh TA, Chambers CV, Diamond JJ. Asthma controller medications: what do patients want? J Asthma 1999;36(7):591-6. 32. Johansson G, Stallberg B, Tornling G, Andersson S, Karlson G, Fait K, et al. Asthma treatment preference study: a conjoint analysis of preferred drug treatments. Chest 2004;125:916-23. 33. Watson J, Cowen P, Lewis R. The relationship between asthma admission rates, routes of admission, and socioeconomic deprivation. Eur Respir J 1996;9:2087-93. 34. Boulet L-P, Belanger M , Lajoie P. Characteristics of subjects with a high frequency of emergency visits for asthma. Am J Emerg Med 1996; 14(7):623-8. 35. Government of British Columbia, Ministry of Health: Pharmacare [Online]. [2006 Jun 21] [cited 2006 Aug 8]; Available from URL: http://www.heal thservices.gov. be. ca/pharme/plani/planiinfo.html#M 36. Ratcliffe J, Van Haselen R, Buxton M , Hardy K, Coleman J, Partridge M . Assessing patients' preferences for characteristics associated with homeopathic and conventional treatment of asthma: a conjoint analysis study. Thorax 2002;57:503-8. -75 -C H A P T E R 5 A C O M P A R I S O N O F H E A L T H - R E L A T E D Q U A L I T Y O F L I F E M E A S U R E S IN A S T H M A * 5.1 Introduction Asthma is a prevalent and chronic respiratory disease that affects approximately one in 12 Canadian adults.[1] Although it is a controllable disease, asthma poses a burden on its patients, often impairing the individual's overall well-being (or quality of life, QOL). To improve asthma management strategies, it is vital to have a valid instrument which can assess these impacts on an individual.[2] Given that the primary goal of asthma management is to achieve optimal control and ultimately to improve the individual's health-related quality of life (HRQL),[3,4] the instrument must be capable of discriminating across all levels of asthma control rather than levels of disease severity. Asthma severity has been defined to reflect the untreated status of the disease for an individual, whereas asthma control describes the adequacy of treatment strategies.[5] Currently, there is a lack of empirical evidence regarding which HRQL instruments have the sensitivity to provide a valid representation of asthma control status. HRQL can be evaluated using either disease-specific or generic preference-based instruments. An example of a disease-specific instrument that estimates asthma-specific QOL is the Asthma Quality of Life Questionnaire (AQLQ).[6] While the advantage of using these measures is their capacity to detect minimal changes in a disease, they are not applicable for comparisons across different disease states. Generic measures, on the other hand, can .be used to compare HRQL between diseases; these involve the determination of a utility value and often encompass a wide range of health-related attributes. A strength of using generic instruments that generate utility values is that they integrate different aspects of health into a single index anchored by a value of 'one' for full health and 'zero' for death. Some health states can attract negative values, which identify, from a societal perspective, A version of this chapter is currently under review for publication in Quality of Life Research. McTaggart-Cowan, H.M. , Marra, C.A., Yang, Y . , Brazier, J.E., Kopec, J.A., FitzGerald, J .M., Anis, A . H . , Lynd, L.D. A Comparison of Health-Related Quality of Life in Asthma. -76 -states being worse than death. Using the calculated utility value for a particular health state and the length of time in that state, the quality-adjusted life year (QALY), a metric used in economic evaluations, can be determined. Determining utility scores normally involves valuing individuals' preferences, using direct measurement approaches such as the standard gamble (SG) and time trade-off (TTO), or indirect measurement techniques such as the multi-attribute health classification systems. The Health Utility Index Mark 3 (HUI-3),[7] the EuroQol (EQ-5D),[8] and the Short Form 6D (SF-6D)[9] are frequently used as indirect measures. The availability of a valid generic, indirect preference-based instrument capable of quantifying HRQL has significant applications in health policy and resource allocation decision-making. [10,11] Validity is defined as the extent to which an instrument measures the property that it is intended to measure. [2] Construct validity assesses the extent to which the scores of an instrument correlate with other hypothesized measures of the health concepts under investigation. [2,12] While most previous economic evaluations of asthma therapies have used either the cost per symptom day avoided or the cost per 0.5 unit change on the A Q L Q as a measure of effectiveness,[13] only a limited number of studies have measured effectiveness in terms of QALYs.[14-19] A recent study, exploring the relationship between asthma control and HRQL, found that both the EQ-5D and the Short Form 36 were able to distinguish across levels of control. [19] These authors classified asthma control status using international guidelines and the St. George's Respiratory Questionnaire (SGRQ).[20] While the SGRQ commonly assesses chronic obstructive pulmonary disease outcomes, studies have also shown it to be reliable in asthmatic populations. [21,22] However, the influence of a validated measure of control, such as the score from the Asthma Control Questionnaire (ACQ),[23] on the HRQL has not been evaluated. There remains a gap in the literature regarding the assessment of construct validity for indirect preference-based instruments in asthma. Therefore, the objective of this study was to evaluate the construct validity of the HUI-3, the EQ-5D, the SF-6D, and the Asthma Quality of Life 5D (AQL-5D),* a condition-specific preference-based measure,[24] in terms V.Yang and J.E. Brazier at the School of Health and Related Research, University of Sheffield developed the condition-specific preference-based measure and converted the study results to the A Q L - 5 D values. - 7 7 -of their ability to distinguish between different levels of asthma control. Asthma control was quantified using the validated A C Q , as well as the magnitude of short-acting (SA) [3-agonist use and self-reported control status. 5.2 Methods 5.2.1 Study Population English-speaking patients with physician-diagnosed asthma between 19 and 49 years of age with no other concurrent respiratory conditions residing in the Vancouver metropolitan area in British Columbia (BC), Canada participated in this cross-sectional study. Each subject was assessed in a pulmonary research clinic and received $20 to defray time and travel costs. The research ethics boards of the University of BC, Providence Health, and Fraser Health approved the study protocol (Appendix I), and each participant provided informed consent (Appendix II). 5.2.2 Data Collection Clinical Measures Data pertaining to demographic information and asthma status were obtained using a self-administered questionnaire: annual utilization of all asthma-related drugs prior to the evaluation, self-perception of their overall QOL measured using a 10 cm horizontal visual analogue scale (VAS) anchored by one for perfect health and zero for death, and self-assessment of their asthma control and their asthma severity, both measured using a five-point Likert scale (Appendix IV).[25] In addition, subjects performed three spirometric tests, with the best measurement recorded. The forced expired volume in the first second (FEVi) is expressed as a percent of predicted FEVi based on the individual's age, height, and sex. [26,27] Asthma Quality of Life and Asthma Control The standardized version of the Asthma Quality of Life Questionnaire (AQLQ(S)) measures the individual's asthma-specific QOL (Appendix V) [6,28,29]; however, the AQLQ(S) cannot be directly used in economic evaluation because it does not incorporate preference information. The instrument consists of 32 questions each scored on a seven-point - 7 8 -Likert scale ranging from one (worst possible QOL) to seven (best possible QOL). It encompasses four domains of asthma-specific QOL: symptoms, activity limitation, emotional function, and environmental stimuli. Averaging the scores for all the questions yields an overall QOL score. Juniper et al. also developed the A C Q to assess asthma control status (Appendix VI).[23] Comprised of seven questions (five relating to asthma symptoms and one each assessing the magnitude of SA p-agonists use and FEVj), the A C Q is both valid and reliable, and possesses strong evaluative and discriminative properties. Scaling each question on a seven-point Likert scale, from zero (best) to six (worst), provides an ordinal characterization of asthma control status. The results are rescaled to coincide with the AQLQ(S) to ensure that higher scores from both instruments indicate better control and better QOL. Preference-based Instruments HUI-3. The HUI-3 includes a health status classification system and a utility scoring formula, derived from multi-attribute utility theory. [7] The scoring formula is based on the SG utilities measured in the general populace. The self-administered HUI-3 survey includes eight attributes: vision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain. As the number of levels for each attribute varies from four to six, the total number of possible health states is 972,000. The baseline preferences for the HUI-3 were measured on a random sample of adults residing in Hamilton, Canada using both a V A S and a SG instrument. Using a multiplicative model, states worse than death can be measured on the zero (dead) to one (perfect health) scale, with a lower bound of -0.36. EQ-5D. The EQ-5D is designed as a cardinal index of health for describing and valuing HRQL.[8] This self-administered survey consists of a descriptive health state classification system with five attributes (mobility, self care, usual activity, pain/discomfort, and anxiety/depression) and a V A S 'health thermometer'. The 'health thermometer' represents a subjective, global evaluation of the respondent's health status on a scale between zero and 100, where zero represents the worst imaginable health state and 100 represents the best imaginable health. Each attribute in the classification system has three levels (no problem, some problems, and major problems) for a total of 243 possible health states. Preferences for - 7 9 -the scoring function used in this study were measured with the TTO technique from a random sample of the adult population in the United Kingdom.[30] The scoring function uses econometric modeling, with scores falling on the zero (dead) to one (perfect health) value scale, with a lower bound of -0.59. SF-6D. The SF-6D is derived from another popular HRQL questionnaire, the Short Form 36.[9] The SF-6D consists of a multi-attribute health status classification system with six attributes and a scoring table, and like the EQ-5D, uses econometric modeling. There are six attributes in this survey: physical functioning, role limitations, social functioning, pain, mental health, and vitality. The classification system ranges from four to six levels for each of the six attributes for a total of 18,000 unique health states. The scoring model for the SF-6D is based on the SG utilities of 836 members of the general population in the United Kingdom. The obtained utilities fall on the standard zero to one (death to perfect health) scale; the worst state in the SF-6D system has a utility of 0.30. AQL-5D. Another method for evaluating HRQL involves developing a condition-specific preference-based measure, which converts the results from the AQLQ(S) into preference-based scores.[24] Using Rasch analysis,[31] the original questionnaire is reduced to a five-dimension classification system (AQL-5D), consisting of the following domains: concern about asthma, shortness of breath, weather and pollution stimuli, sleep impact, and activity limitation. The health states obtained from the AQLQ(S) are directly mapped onto this newly defined classification system and valuations of intermediate AQL-5D health states are obtained. From this, an econometric model is constructed to predict values for all health states. 5.2.3 Analysis Descriptive statistics characterized the sample in terms of age, sex, pulmonary function, HRQL scores, perceived asthma severity and control, asthma medication use, and concomitant chronic illnesses. Continuous variables are presented as means and standard deviations while categorical variables are presented as the proportion of the sample within each group. - 8 0 -Construct validity is assessed based on the ability of each instrument to discriminate between patients with different levels of asthma control, represented by the A C Q score, magnitude of SA P-agonist use, and self-reported control status. Current management guidelines characterize good asthma control as requiring four or fewer SA p-agonist canisters per year.[32] Therefore, we expect that patients who use excessive doses of SA P-agonist as their mainstay of therapy probably require more healthcare services, an outcome that suggests poor asthma control and potentially lower QOL.[25] Other variables assessing construct validity include the presence of other chronic diseases and self-reported asthma severity. Patients with poorer asthma control are hypothesized to have lower scores for all measures. Analysis of variance (ANOVA) evaluates the differences among the utility scores when stratified by magnitude of SA P-agonist use, number of chronic illnesses, A C Q scores, and self-reported asthma severity and control. Convergent validity is assessed by comparing the relationships between FEV], preference-based utilities, and the disease-specific and V A S scores. The potential skewedness of the data requires the use of non-parametric correlations (Spearman's rho (p)). A correlation coefficient of greater than 0.5 or less than -0.5 is considered to be strong, 0.30 to 0.49 or -0.49 to -0.30 is considered to be moderate, and values between 0.30 and -0.30 are considered to be weak.[33] A l l analyses have been performed using the SAS statistical software package, version 8.2 (SAS Institute, Cary, NC). Significance is defined a priori as p<0.05. 5.3 Results 5.3.1 Sample One hundred and fifty-seven respondents (110 females (70%), mean age of 35.0 years) participated in the study. A general poster campaign recruited 114 (73%) participants, respiratory clinics recruited 17, and community pharmacists recruited the remainder. The recruitment methodology resulted in a heterogeneous sample of patients with a full spectrum of HRQL and representative of all levels of disease severity, disease control, and drug use (Table 5.1). -81 -5.3.2 Description of Global Utilities Figure 5.1 displays a summary of the results from the preference-based instruments. A l l the instruments achieve a maximum score of 1.0 but the minimum values vary. The EQ-5D, SF-6D, and AQL-5D are similar in range, with lower bounds of 0.45, 0.49, and 0.50, respectively. Conversely, the range of the HUI-3 is much wider, due to a smaller lower bound (0.16 to 1.0). Figure 5.2 shows that higher scores are obtained with the EQ-5D and the HUI-3, when 79 (50%) and 86 (55%) respondents, respectively, reported a utility of greater than 0.9 for these instruments. Only 19 (12%) respondents reported utilities of this value for the SF-6D. 5.3.3 Construct Validity Table 5.2 illustrates the relationships between the scores obtained from the preference-based instruments and all investigated measures of asthma control. The relationships between the magnitude of SA P-agonist use and all instruments generally demonstrate a monotonic gradient, such that a lower HRQL is associated with greater medication use. This expresses the ability of the instruments to discriminate between different levels of asthma control, thereby supporting construct validity for all instruments. However, the A N O V A reveals that the SF-6D is not able to differentiate between levels of SA P-agonist use (p = 0.7). A l l instruments are able to discriminate between the presence of other chronic diseases (p<0.03). The assessment of construct validity using the patients' perception of their asthma severity and control level conflicted amongst the preference-based instruments with only the AQL-5D being able to discriminate across these variables (pO.OOOl). Both the disease-specific instruments (pO.OOOl) and the VASs (p<0.0008) are able to differentiate across these variables. Furthermore, the comparisons of the stratified ACQ scores with the HUI-3, the EQ-5D, the SF-6D, and the AQL-5D generate a positive gradient, such that higher utilities are associated with better asthma control (Figure 5.3). The A N O V A reveals that all preference-based instruments are able to discriminate across the ACQ scores. The correlations between the F E V i and the majority of the instruments are positive but weak (p = 0.06-0.26), although A C Q strongly correlates with this physiological measurement (p = 0.55) (Table 5.3). The three indirect preference-based instruments are - 8 2 -strongly correlated (p = 0.62-0.73) but the correlations are progressively weaker when compared to the VASs, AQL-5D, and disease-specific instruments. The AQL-5D is moderately correlated with the indirect, generic measures and the A C Q , although it strongly correlates with the AQLQ(S). The disease-specific measures are moderately to strongly correlated with both VASs. 5.4 Discussion This study evaluates the construct validity of the indirect preference-based instruments in asthma, specifically their ability to distinguish between measures of asthma control. Overall, the results reveal that these instruments are able to discriminate across the A C Q scores, providing evidence that preference-based instruments can detect changes in asthma states. However, when using the magnitude of SA p-agonist use as a subjective measure of asthma control, all preference-based measures but the SF-6D are able to differentiate across levels of this medication use. The disease-specific instruments and VASs can distinguish across levels of self-reported asthma control and asthma severity. Of the preference-based measures, however, only the AQL-5D is able to detect this distinction. The preference-based instruments display convergent validity with most measures of control, as shown by the moderate correlations; however, the HUI-3 weakly correlates with the A C Q (p = 0.20). In agreement with previous work,[17,19] the F E V i is only weakly correlated with the preference-based instruments and the AQLQ(S). As this pulmonary measurement is a component of the A C Q , it is not surprising that it. correlates strongly with this specific instrument (p = 0.55). The inability of the indirect preference-based instruments to discriminate across levels of self-reported asthma control and severity may imply that either poorly-controlled or severe asthma patients perceive their health status more favourably than patients with well-controlled or mild disease. The former individuals may gradually learn to cope and adapt to their limitations in a number of ways such that, over time, the perception of the impact of their disease is reduced. Future work may be to investigate potential psychosocial factors that contribute to this adaptation, as well as the mechanisms by which individuals' cope with their health condition. Understanding the degree to which patients adapt to their condition -83 -could provide useful information in obtaining more valid health state valuations by modifying societal valuations to incorporate these effects. In addition to these potential behavioural effects, the valuation methods and the psychometric properties. of the preference-based instruments can explain the differences between utility scores.[34] Both the SF-6D and the HUI-3 use the SG technique for valuation, while the EQ-5D and the AQL-5D, use the TTO approach. The HUI-3 uses multi-attribute utility theory and multiplicative scoring models, while the SF-6D and EQ-5D use empiric and additive scoring methods.[35] In addition, the scoring functions for the EQ-5D and SF-6D are derived based on responses from the United Kingdom population; as a result, they may differ from the HUI-3 preferences, which are derived from Canadian respondents.[9,36] There is a concern that population valuations tend to differ from the values of patients actually in that specific health state. [3 7] Both the EQ-5D and HUI-3 are known to have strong ceiling effects, such that they are hampered in discriminating between patients with nearly perfect health.[35] A probable reason for the ceiling effect of the EQ-5D is the presence of only three response levels. This may result in a lack of sensitivity in detecting differences in health states, preventing patients from expressing minor health problems. [12] For this study, over half of the study population reported utilities of greater than 0.9 on the EQ-5D. While both the EQ-5D and the SF-6D include a domain that represents usual activity, the description of this domain is more specific in the EQ-5D, encompassing work, study, housework, family, or leisure activities, such that there is no ambiguity for the patient when interpreting this question. As asthma is a health condition that may result in detrimental effects on many areas of life, including physical, psychological, and social functioning, it is necessary to have a domain to capture these aspects. [38] While this study is cross-sectional in nature, the longitudinal properties of the preference-based instruments need to be evaluated for their responsiveness. If a treatment strategy results in an important difference in HRQL, researchers need to be confident that the instruments will be able to detect that difference. The evaluative nature of these instruments needs to be assessed, as it would be beneficial not only to measure improvements in HRQL with treatments of asthma but also to compare asthma-based HRQL scores with those for other chronic diseases. - 84 -Since health is a function of both quality and length of life, the Q A L Y is often used to measure health outcomes in economic evaluations. The Q A L Y provides a foundation for cost-utility analysis, which may be used to compare the relative efficiency of different health programs. The utility scores obtained from the different preference-based instruments need to be incorporated into a Q A L Y measure to evaluate any potential differences that may arise. These may or may not be statistically different and may or may not call for different policy decisions. In conclusion, the HRQL scores derived from the preference-based instruments, the disease-specific instruments, and the VASs are capable of distinguishing between differing degrees of asthma control. While the study sample may not be representative of the general asthmatic population, the sample was heterogeneous and encompassed all levels of disease control and disease severity; this contributes substantially to the validity of these findings. Although the preference-based instruments demonstrate construct validity with the validated ACQ, these instruments are unable to discriminate across levels of subjective control measures, which include the magnitude of SA (3-agonist use and the individual's self-perception of their control status. While disease-specific measures are able to differentiate across all investigated measures of asthma control, they, unlike preference-based instruments, cannot directly compare across different disease states. This study also provides evidence that information obtained from disease-specific instruments can be successfully converted into preference-based single-index equivalents as shown by the significant relationships between the scores from the condition-specific preference-based instrument, AQL-5D, and all measures of asthma control, including subjective measures. This result has considerable implications in the economic assessment of asthma management strategies, since it permits the prediction of Q A L Y s in clinical situations even if only disease-specific instruments are being administered. - 85 -5.5 Tables Table 5.1: Characteristics of study population Parameter Frequency (%) or Mean ( ± S D ) t Age 35.0 (7.9) Female 110(70) F E V , 90.9 (19.9) H R Q L scores EQ-5D 0.84 (0.23) HUI-3 0.84 (0.20) SF-6D 0.79 (0.10) AQL-5D 0.85 (0.11) EQ-5D VAS 0.74(0.18) VAS 0.73 (0.20) AQLQ(S) (1-7) 5.4(1.1) ACQ, (1-7) 5.7 (0.9) Self-reported asthma severity Very mild 21 (13) Mild 59 (38) Moderate 51 (33) Severe 20(13) Very severe 5(3) Self-reported asthma control Very well controlled 37 (24) Well controlled 43 (28) Adequately controlled 54 (35) Not well controlled 19 (12) Not controlled at all 3(2) Number of short-acting P-agonist canisters used in the past year None 15(10) 4 or fewer 95 (61) 5-12 35 (22) 13 or more 12(8) Number of ICS canisters used in the past year None 63 (41) 4 or fewer 56 (36) 5-12 28(18) 13 or more 8(5) Current use of oral steroids Yes 8(5) No 147 (95) Concomitant chronic illnesses Yes 57 (37) No 99 (63) A C Q : Asthma Control Questionnaire; AQL-5D: Asthma Quality of Life 5D; AQLQ(S): standardized version of the Asthma Quality of Life Questionnaire; EQ-5D: EuroQol; EQ-5D V A S : 'health thermometer' from EuroQol; F E V i : forced expired volume in the first second; HRQL: health-related quality of life; HUI-3: Health Utility Index Mark 3; ICS: inhaled corticosteroids; SF-6D: Short Form 6D; V A S : visual analogue scale. SD: standard deviation. - 8 6 -Table 5.2: Relationship between asthma control and the H R Q L instruments Score, mean (SD)*~ EQ-5D HUI-3 SF-6D AQL-5D A C Q AQLQ(S) EQ-5D V A S V A S Number of short-acting B-agonist canisters used in the past year 4 or less 0.87(0.21)§ 0.86(0.19)§ 0.79 (0.10)1 0.87(0.10)* 5.95 (0.82)* 5 - 12 0.85 (0.14)§ 0.81(0.20)§ 0.80 (0.10)11 0.83 (0.09)* 5.51(0.71)* 12 or more 0.69 (0.38)§ 0.74(0.23)§ 0.76 (0.11)* 0.76 (0.1 1)* 4.67(1.01)* 5.64 (0.98)* 5.15 (1.04)* 4.21 (1.10)* 0.78 (0.18)* 0.76 (0.10)* 0.61 (0.22)* 0.75 (0.20)* 0.77 (0.17)§ 0.57 (0.20)§ Other chronic diseases None 0.90 (0.11)1 0.90(0.11)* 0.81 (0.08)* 0.86(0.10)§ 5.82(0.88)* lormore 0.73 (0.28)* 0.72(0.25)* 0.74 (0.12)* 0.82 (0.11)§ 5.48 (1.02)* 5.55 (1.03)* 5.05 (1.20)* 0.80 (0.15)+ 0.67 (0.21)* 0.78 (0.1 iy 0.64 (0.22)* ^ 1 Self-reported asthma severity Very mild Mi ld Moderate Severe Very severe 0.84 (0.29)1 0.89 (0.18)1 0.81 (0.21)1 0.78 (0.25)1 0.67 (0.38)1 0.82 (0.22)1 0.88 (0.18)11 0.84 (0.15)' 0.73 (0.28)1 0.81 (0.15)11 0.80 (0.22)1 0.80 (0.09)1 0.78 (0.08)11 0.74 (0.13)11 0.79 (0.07)1 0.92 (0.08)* 0.87 (0.08)* 0.83 (0.09)* 0.74 (0.14)* 0.75 (0.15)* 6.34 (0.76)+ 6.01 (0.74)* 5.56 (0.72)* 4.84 (1.09)* 4.23 (1.18)* 6.20 (0.81)* 5.68 (0.79)* 5.10(1.06)* 4.46(1.35)* 4.29(1.47)* 0.79 (0.18)* 0.80 (0.15)* 0.76 (0.15)* 0.64 (0.22)* 0.45 (0.36)* 0.80 (0.22)+ 0.78 (0.15)* 0.73 (0.16)* 0.55 (0.22)* 0.45 (0.33)* Self-reported asthma control Very well 0.90 (0.23)f controlled Well controlled Adequately controlled Not well controlled Not controlled 0.66 (0.42)1 at all 0.84 (0.20)1 0.81 (0.22)1 0.83 (0.18)1 0.88 (0.18)11 0.83 (0.20)1 0.84 (0.15)1 0.84 (0.16)1 0.85 (0.20)1 ,1 0.82 (0.11)' 0.79 (0.09)1 0.78 (0.08)' 0.76 (0.10)11 0.83 (0.08)11 0.92 (0.07)+ 0.87 (0.09)* 0.81 (0.10)* 0.78 (0.12)* 0.77 (0.08)* 6.32 (0.66)* 5.95 (0.78)* 5.29 (0.90)* 5.38 (1.00)* 4.43 (0.29)* 6.16(0.73)* 0.81(0.15)* 0.81(0.18)* 5.67 (0.86)* 0.78 (0.15)* 0.77 (0.17)* 4.92 (1.15)* 0.73 (0.19)* 0.69(0.20)* 4.64 (1.05)* 0.72 (0.18)* 0.65 (0.21)* 4.32 (0.63)* 0.40 (0.35)* 0.29 (0.13)* *ACQ: Asthma Control Questionnaire; AQL-5D: Asthma Quality of Life 5D; AQLQ(S): standardized version of the Asthma Quality of Life Questionnaire; EQ-5D: EuroQol; EQ-5D V A S : 'health thermometer' from EuroQol; HRQL: health-related quality of life; HUI-3: Health Utility Index Mark 3; SF-6D: Short Form 6D; V A S : visual analogue scale. + SD: standard deviation. Comparison (using A N O V A ) of mean values, :p<0.001, sp<0.05, \>Q.Q9. Table 5.3: Correlations (Spearman's rho) for the instruments EQ-5D HUI-3 SF-6D AQL-5D EQ-VAS V A S A C Q AQLQ(S) FEV, EQ-5D 1.00 HUI-3 0.73 1.00 SF-6D 0.63 0.62 1.00 AQL-5D 0.39 0.26 0.41 1.00 EQ-VAS 0.58 0.52 0.50 0.48 1.00 V A S 0.53 0.45 0.49 0.49 0.79 1.00 A C Q 0.36 0.20 0.36 0.81 0.48 0.45 1.00 AQLQ(S) 0.45 0.30 0.46 0.92 0.54 0.53 0.81 1.00 F E V i 0.14 0.06 0.15 0.26 0.21 0.23 0.55 0.26 1.00 ' A C Q : Asthma Control Questionnaire; AQL-5D: Asthma Quality of Life 5D; AQLQ(S): standardized version of the Asthma Quality of Life Questionnaire; EQ-5D: EuroQol; EQ-5D V A S : 'health thermometer' from EuroQol; FEV, : forced expired volume in the first second; HUI-3: Health Utility Index Mark 3; SF-6D: Short Form 6D; V A S : visual analogue scale. 5.6 Figures Figure 5.1: Distribution of global utility values across the H R Q L instruments*^ 00 o CD o o CN O O O i i i i EQ-5D HUI-3 S F - 6D AQL-5D Instrument Outliers are marked by lines f AQL-5D: Asthma Quality of Life 5D; EQ-5D: EuroQol; HUI-3: Health Utility Index Mark 3; SF-6D: Short Form 6D. - 8 9 -Figure 5.2: Distribution of global utility values across the H R Q L instruments 0to 0.09 0.1 to 0.19 0.2to 0.29 0.3 to 0.39 0.4to 0.49 0.5 to 0.59 0.6 to 0.69 0.7 to 0.79 0.8 to 0.89 0.9 to 1.0 Global Utility Values EQ-5D: EuroQol: HUI-3: Health Utility Index Mark 3; SF-6D: Short Form 6D. - 9 0 -Figure 5.3: Relationship between Asthma Control Questionnaire ( A C Q ) score and preference-based instruments ' f 2 to 2.99 6 to 7 *AQL-5D: Asthma Quality of Life 5D; EQ-5D: EuroQol; HUI-3: Health Utility Index Mark 3; SF-6D: Short Form 6D. Comparison (using A N O V A ) of mean values stratified by A C Q score (pO.OOl). -91 -5.7 References 1. Statistics Canada. 2001 National population health survey [Online]. [2001?] [cited 2006 Jun 20]; Available from: URL: http://www40.statcan.ca/101/cst01/health50a.htm 2. Streiner DL, Norman GR. Health measurement scales: a practical guide to their development and use. 2 n d ed. New York: Oxford University Press; 1995. 3. Bateman ED, Boushey HA, Bousquet J, Busse WW, Clark TJ, Pauwels RA, et al. Can guideline-defined asthma control be achieved? The Gaining Optimal Asthma ControL study. A m J Respir Crit Care Med 2004;170:836-44. 4. O'Byrne P M , Bisgaard H, Godard PP, Pistolesi M , Palmqvist M , Zhu Y , et al. Budesonide/Formoterol combination therapy as both maintenance and reliever medication in asthma. Am J Respir Crit Care Med 2005;171:129-36. 5. Cockcroft DW, Swystun V A . Asthma control versus asthma severity. J Allergy Clin Immunol 1996;98:1016-8. 6. Juniper EF, Guyatt GH, Ferrie PJ, Griffith LE . Measuring quality of life in asthma. Am Rev Respir Dis 1993;147:832-8. 7. Feeny D, Furlong W, Torrance GW, Goldsmith C H , Zhu Z, DePauw S, et al. Multiattribute and single-attribute utility functions for the Health Utilities Index mark 3 system. Med Care 2002;40:113-28. 8. The EuroQol Group. EuroQoL - a new facility for the measurement of health-related quality of life. Health Policy 1990;16:199-208. 9. Brazier J, Roberts J, Deverill M . The estimation of a preference-based measure of health from the SF-36. J Health Econ 2002;21:271-92. 10. Weinstein M C , Siegel JE, Gold MR, Kamlet MS, Russell L B . Recommendations of the panel on cost-effectiveness in health and medicine. J A M A 1996;276:1253-8. 11. Neumann PJ, Goldie SJ, Weinstein M C . Preference-based measures in economic evaluation in health care. Ann Rev Public Health 2000;21:587-611. 12. Brazier J, Deverill M . A checklist forjudging preference-based measures of health related quality of life: learning from psychometrics. Health Econ 1999;8:41-51. 13. Sculpher M . Using economic evaluations to reduce the burden of asthma and chronic obstructive pulmonary disease. Pharmacoeconomics 2001;19:21-5. - 9 2 -14. Paltiel A D , Fuhlbrigge A L , Kitch BT, Liljas B, Weiss ST, Neumann PJ, et al. Cost-effectiveness of inhaled corticosteroids in adults with mild-to-moderate asthma: results from the asthma policy model. J Allergy Clin Immunol 2001;108:39-46. 15. van den Boom G, Rutten-van Molken MP, Molema J, Tirimanna PR, van Weel C, van Schayck CP. The cost effectiveness of early treatment with fluticasone propionate 250 microg twice a day in subjects with obstructive airway disease. Results of the D I M C A program. Am J Respir Crit Care Med 2001;164:2057-66. 16. Schermer TR, Thoonen BP, van den Boom G, Akkermans RP, Grol RP, Folgering HT, et al. Randomized controlled economic evaluation of asthma self-management in primary health care. Am J Respir Crit Care Med 2002;166:1062-72. 17. Stahl E, Postma DS, Svensson K, Tattersfield A E , Eivindson A , Schreurs A, et al. Formoterol used as needed improves health-related quality of life in asthmatic patients uncontrolled with inhaled corticosteroids. Respir Med 2003;97:1061-6. 18. Moy M L , Fuhlbrigge A L , Blumenschein K, Chapman RH, Zillich AJ , Kuntz K M , et al. Association between preference-based health-related quality of life and asthma severity. Ann Allergy Asthma Immunol 2004;92:329-34. 19. Szende A , Svensson K, Stahl E, Meszaros A , Berta GY. Psychometric and utility-based measures of health status of asthmatic patients with different disease control level. Pharmacoeconomics 2004;22:537-47. 20. Global Initiative for Asthma. Pocket guide for asthma management and prevention. Bethesda (MD): National Institute of Health, Nov 1998. NIH Pub. No. 96-3569B. 21. Jones PW, Quirk FA, Baveystock C M , Littlejohns P. A set-complete measure of health status for chronic airflow limitation. The St. George's Respiratory Questionnaire. A m Rev Respir Dis 1992;145:1321-7. 22. Sanjuas C, Alonson J, Prieto L, Ferrer M , Broquetas JM, Anto JM. Health-related quality of life in asthma: a comparison between St. George's respiratory questionnaire and asthma quality of life questionnaire. Qual Life Res 2002; 11:729-38. 23. Juniper EF, O'Byrne P M , Roberts JN. Measuring asthma control in group studies: do we need airway caliber and rescue beta2-agonist use? Respir Med 2001 ;95:319-23. 24. Yang Y , Tsuchiya A , Brazier JE, Young TA. Estimating a preference-based index from the Asthma Quality of Life Questionnaire (AQLQ). Paper presented at the Health Economists' Study Group Meeting, City University, London, United Kingdom; 2006. 25. Lynd L D , Guh DP, Pare PD, Anis A H . Patterns of inhaled asthma medication use: a 3-year longitudinal analysis of prescription claims data from British Columbia, Canada. Chest 2002;122(6): 1973-81. -93 -26. Crapo RO, Morris A H , Gardner R M . Reference spirometric values using techniques and equipment that meet ATS recommendations. Am Rev Respir Dis 1981; 123:659-64. 27. American Thoracic Society. Standardization of spirometry—1987 update. Statement of the American Thoracic Society. Am Rev Respir Dis 1987;136:1285-98. 28. Juniper E, Guyatt G, Epstein R. Evaluation of impairment of health related quality of life in asthma: development of a questionnaire for use in clinical trials. Thorax 1992;47:76-83. 29. Juniper EF, Buist AS, Cox F M , Ferrie PJ, King DR. Validation of a standardized version of the Asthma Quality, of Life Questionnaire. Chest 1999; 115:1265-70. 30. Dolan P, Gudex C. Time preference, duration and health state valuations. Health Econ 1995;4:289-99. 31. Rasch G. Probabilistic models for some intelligence and attainment tests. Chicago: University of Chicago; 1980. 32. Boulet L-P, Becker A , Berube D, Beveridge R, Ernst, P. Canadian asthma consensus report, 1999. Can Med Assoc J 1999;161, S1-S62. 33. Cohen J. A power primer. Psychol Bull 1992;112:155-9. 34. Bleichrodt H, Johannesson, M . Standard gamble, time trade-off and rating scale: Experimental results on the ranking properties of QALYs . J Health Econ 1997;16:155-5. 35. Kopec JA, Willison K D . A comparative review of four preference-weighted measures of health-related quality of life. J Clin Epidemiol 2003;56:317-32. 36. Drummond M F , Sculpher MJ, Torrance GW, O'Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programmes. 3 r d ed. New York: Oxford University Press; 2005. 37. Brazier J, Akehurst R, Brennan A, Dolan P, Claxton K, McCabe C, et al. Should patients have a greater role in valuing health states? Appl Health Econ Health Policy 2005;4:201-8. 38. Marra CA, Woolcott JC, Kopec JA, Shojania K, Offer R, Brazier JE, et al. A comparison of generic, indirect utility measures (the HUI2, HUI3, SF-6D, and the EQ-5D) and disease-specific instruments (the RAQoL and the HAQ) in rheumatoid arthritis. Soc Sci Med 2005;60:1571-82. - 94 -C H A P T E R 6 S U M M A R Y , CONTRIBUTIONS, AND R E C O M M E N D A T I O N S 6.1 Summary of Key Research Findings The increasing prevalence of asthma and its continuing widespread inappropriate management presents a pressing public health issue for both the patient and society at large.[l] The results of the present study provide significant insights into both the preferences that patients have towards specific components of asthma treatments and the health-related quality of life (HRQL) of patients relating to the control of their asthma. A survey of asthma patients reveals that various aspects of asthma treatment had different levels of importance for patients. In terms of medication delivery, oral is preferred over inhaled administration; for inhaled therapies, dry power inhalers are preferred over metered dose inhalers. Patients also prefer fewer inhalers. Management studies indicate that asthma patients prefer to have action plans and to have some role in self-management decisions. The degree of autonomy desired, however, varied substantially. With reference to symptoms, a reduction in the severity of cough and breathlessness was more important than improvements in wheeze, sleep disturbances, and chest tightness. Results from the discrete choice experiment (DCE) demonstrate that the patient's sex and income level significantly influences the treatment preferences. Women are 1.90 times more likely than men to decline asthma treatment; similarly, those that receive social assistance are 2.95 times more likely than individuals not receiving social assistance to not have their asthma treated. Furthermore, the individual's age and asthma-specific quality of life also, impact this preference. There is a 1.02 and a 1.12 times increase in the likelihood for not treating asthma for every yearly increase in age and for every one unit increase in the score of the standardized Asthma Quality of Life Questionnaire (AQLQ(S)), respectively. Thus, these results reveal that individuals who are male, who are, younger, who have poorer asthma quality of life (QOL), or who are not receiving social assistance have greater desire for treatment of their asthma, compared to those who are female, who are older, who have better asthma QOL, or who are receiving social assistance. -95 -Overall, the DCE results reveal that, as would be expected, patients prefer asthma treatments with more symptom-free days (SFDs), lower cost, greater convenience (in terms of requiring fewer inhaler and fewer daily administration), and reduction in the frequency of adverse events. The study findings also show that while patients prefer a treatment that results in more SFDs, they are willing to forego these benefits in exchange for a reduction of negative treatment impacts, in terms of convenience and adverse events. Specifically, patients prefer to take a treatment on an as needed or daily basis compared to a treatment requiring multiple daily doses; they are, however, indifferent to taking a medication on an as needed versus a daily basis. This suggests that individuals are not totally opposed to a management plan consisting of an inhaled corticosteroid (ICS), which requires a more regimented dosing schedule. A preference for using one inhaler is also detected, indicative of a treatment strategy of only a short-acting (SA) P -agonist; however, patients report no statistically significant preferences between treatments involving two or three inhalers. This provides some evidence that the proposed stepped-care management approach is less desirable from a patients' perspective,[2] as they would need to be on at least two inhalers (an inhaler for control and an inhaler for symptomatic relief). Furthermore, asthma patients prefer treatments that will result in no episodes of adverse events. In particular, patients are interested in treatments with no probability of oral thrush, a side effect specific to ICS. Not unexpectedly, a treatment that results in less episodes of thrush is preferred to a treatment resulting in more frequent events. Similarly, patients prefer treatments with no episodes of tremors/heart palpitations; this side effect is specific to SA P-agonists. However, there is no statistically significant difference in preferences between treatments that result in one or two events of this particular side effect. The increased frequency of this adverse event compared to oral thrush (monthly versus annually, respectively) and the indifference in preference between the levels of tremor/palpitation episodes potentially suggests that patients are somewhat willing to accept this adverse event over thrush episodes. This may provide evidence for the causes of inappropriate of asthma management, such that individuals are more likely to overuse SA p-agonists as their control therapy due to greater simplicity and avoidance of oral thrush episodes. Treatment preferences also vary across sub-populations, in that individuals with lower income, less education, and better asthma control have a lower willingness-to-pay - 9 6 -(WTP) for a treatment that will result in greater benefits and a reduction of risks, relative to those with higher income, more education, and worse asthma control. Annual income and education levels significantly influence treatment preferences in regards to treatment cost and adverse event frequency, respectively. In particular, individuals annually earning greater than $50,000 are willing to pay more to receive an increase in treatment benefits (for example more SFDs, less episodes of oral thrush). As well, individuals with more formal education (at least a university Bachelor's degree) have stronger preferences for treatments with a better adverse event profile compared to individuals without university degrees. In terms of asthma control status, the preferences of individuals who have less frequent symptoms are different from those who frequently experience symptoms. If one takes the magnitude of SA P-agonist use as a surrogate for asthma control, individuals using greater amounts of SA P-agonists have greater preferences for treatments resulting in more monthly SFDs compared to those using fewer SA P-agonists. While DCEs provide valuable information regarding the preferences of individual patients for various aspects of asthma therapy, they do not provide insights into the relative usefulness or the cost-effectiveness of specific treatments. To provide valid economic evaluations of asthma treatment options for policy analysis and resource allocation purposes, the validity of HRQL measurements in asthma also needs to be addressed. A series of HRQL instruments (including those that are preference-based or disease-specific) and visual analogue scales (VASs), display construct validity by discriminating across levels of asthma control, as identified by the Asthma Control Questionnaire (ACQ).[3] However, when using the magnitude of SA P-agonist use as a measure of asthma control, all instruments but the Short Form 6D are able to differentiate across levels of this medication use. A gradient is observed for both the disease-specific instruments (standardized version of the Asthma Quality of Life Questionnaire, AQLQ(S), and the ACQ) and the VASs with the individual's perception of his/her control status, such that a lower HRQL corresponds to poor control. The absence of a consistent trend between the preference-based instruments and self-reported disease control suggests that poorly-controlled asthma patients may perceive their health status more favourably than well-controlled asthma patients. The former individuals may gradually learn to cope and adapt to their limitations such that their perception of the disease impacts is reduced over time. The majority of the preference-based instruments also - 9 7 -display convergent validity with most measures of asthma control, as shown by the moderate correlations; however, the Health Utility Index Mark 3 only weakly correlates with the ACQ. In agreement with previous work,[4,5] the forced expired volume in one second (FEVi) weakly correlates with the preference-based instruments. 6.2 Unique Contributions, Impact, and Implications This is the first DCE that has incorporated two elements of risk specific to asthma therapies. The absence of a risk component in previous studies of patients' preferences for asthma treatment fails to simulate a true decision-making response in the evaluation of treatment preferences. Consequently, the willingness to accept risk as a natural part of treatment and the willingness to trade-off treatment-related benefits for the avoidance of perceived risk have not been previously evaluated. In the only previous asthma DCE assessing the influence of treatment side effects, only a single, collective attribute represented multiple side effects. [6] This lack of a clear and consistent definition of the side effects most likely resulted in potential ambiguity from the patient's perspective. The non-significance of the results may be attributed more to the lack of definition of the side effects, rather than to any true treatment preferences. The methodology used in this study effectively captures patients' preferences for a treatment that will result in fewer risks. It also demonstrates that patients are willing to trade improvements in symptom relief for a reduction in the frequency of specific adverse events and of treatment administration frequency. It reveals that risk preferences also differ amongst social classes: individuals with higher incomes or university degrees have a greater WTP to avoid treatment-related adverse events compared to those individuals without university degrees. This finding provides useful information for the development of targeted asthma education and treatment programs, particularly for the design of individualized management plans for individuals with suboptimal asthma control. Furthermore, these results will aid clinicians and researchers for future work aimed at improving treatment efficacy and health outcomes by incorporating patients' preferences for their asthma treatments. In addition, these findings will aid decision-makers in improving access to treatments for all patients. The empirical study results show that patients do not always interpret the levels of treatment attributes continuously. Thus, analysis using a linear additive regression model - 9 8 -might not provide a valid representation of the individual's true decision-making process. While the majority of DCEs use the linear regression additive approach predicated on the assumption that there is a linear and proportionate gradient between each attribute level, the results of this study indicate that this assumption is not valid. This suggests that dummy-coded attribute levels may be a better technique to apply in any regression modeling. This is the first study comparing HRQL with a validated measure of asthma control, such as the A C Q score. A consistently strong relationship between all HRQL instruments and this measure is observed; however, the relationships between the preference-based instruments and the magnitude of SA P-agonist use and the patients' self-perception of their asthma status are less consistent. Specifically, low HRQL scores from preference-based instruments do not necessarily correspond to perceived poorer asthma control. 6.3 Study Limitations As with any study, this one has its limitations; however, none of them should significantly affect its main findings. There is a potential for the misclassification of socioeconomic status (SES) since the study depends on patient self-reporting for this variable. It is more likely that subjects in lower social classes will be upwardly misclassified, as they may report more education or higher income than actually is the case.[7] However, across the different measures of SES, the relationships between the relative preferences and all treatment attributes are in the anticipated direction, suggesting that any misclassification of SES is likely minimal. In addition, the self-reported quantities of SA p-agonists used over the past year are subject to recall bias, although differentiation in this regard between individuals from different social classes is not expected. In relation to the final results, the misclassification of SES and SA P-agonist use will result in conservatively biased estimates of this association; therefore, the true relationship should not be less than, and may actually be greater than, what was identified. This study included 157 participants recruited from metropolitan Vancouver, British Columbia (BC); as such, the results may not be generalizable. Although the sample was heterogeneous and encompassed all levels of disease control and severity, the study sample may not be representative of the general asthma population in BC. The results may be biased if, for some reason, those who chose to participate had different preferences and outcomes - 9 9 -than those who chose not to participate. This may result in a potential for differences in health behaviour and in the likelihood of study participation; however, there is no evidence that this bias is present in the sample. The only incentives for participation were $20 compensation for time and travel costs and the provision of the spirometry test results to the participant's physician; no intervention or education was offered to improve the level of asthma control. Social class differences in health behaviours suggest that uncontrolled asthma patients in higher social classes are more likely to seek care and to participate if they perceive that the provision of their test results to their physician would be beneficial.[8] The small financial incentive was not expected to provide differential participation in controlled or uncontrolled asthma patients of lower social classes. 6.4 Recommendations The results from this thesis have the potential to significantly impact asthma outcomes research and clinical decision-making. It would be beneficial to discover if preferences for treatment attributes differ over the course of the year. While the single DCE reported here provides significant information regarding patients' preferences for specific treatment attributes, these results are cross-sectional. By implementing the same DCE questionnaire several times in a year, changes in the individual's asthma control status could be evaluated and the influence of these changes on treatment preferences could be assessed. This experimental design would also provide more information regarding the reliability of DCEs, such as the test-retest reliability. While the profiles for the present study have been combined randomly and tested for orthogonality, a DCE constructed to maximize its D-efficiency would produce a more statistically efficient design. [9] D-efficiency is maximized when the size of the covariance matrix of the estimated parameters is minimized. This is achieved by incorporating prior information about the respondents' preferences into the design process. Further work is needed to see if significant differences exist in the regression estimates of experimental designs where profiles are randomly combined compared to those where the covariance matrix is minimized. - 100-Analysis of the majority of DCEs in the literature is conducted using the multinomial logit model. In spite of this common practice, this model suffers from the restrictive independence from irrelevant alternatives (IIA) assumption, which states that the ratio of two choice probabilities is independent of the other alternatives in the model. Thus, using the multinomial model implies that a change in an attribute of one alternative will have the same proportional impact on the probability of each of the other alternatives being chosen. The current study used the nested logit model (NLM) which relaxes the IIA property by dividing the alternatives into nests. However, the IIA property is not completely avoided, as the assumption does not hold for alternatives in different nests. While this analysis uses a N L M for simplicity in estimation, the use of a mixed logit model may have wider applicability in future work. [10] By explicitly accounting for correlations in unobserved utility over repeated choices by each decision-maker, the mixed logit model would not violate the IIA assumption. The longitudinal properties of HRQL instruments also need to be evaluated for their ability to be responsive to changes. If a treatment strategy results in an important difference in HRQL, researchers need to be confident that the instruments will be able to detect such a difference. The evaluative nature of these instruments needs to be assessed, as it would be beneficial not only to measure improvements in HRQL with treatments of asthma but also to compare asthma-based HRQL scores with those for other chronic diseases. Since health is a function of both the length of life and utility, the Q A L Y is normally used to measure health outcomes in economic evaluations. The Q A L Y provides a foundation for cost-utility analysis, which is used in a general health policy model to compare the efficiency of different programs or to assess the relative contribution of different programs and providers in the healthcare system. The utility, scores obtained from the different preference-based instruments need to be incorporated into a Q A L Y measure to evaluate any potential differences that may arise. These may or may not be statistically different and may or may not call for different policy decisions. Finally, the study results show that individuals with poorly-controlled asthma perceive their health status more favourably than those with well-controlled asthma. Thus, there may be a limit in the ability of patients to self-assess their HRQL, particularly i f they have adapted to either their own pain levels or symptoms, or if they have developed tendencies to minimize the impact of these problems on their overall well-being.[11] As - 101 -responses obtained from the general population are traditionally used as surrogates in the valuation of health states, it has been suggested that these valuations tend to differ from the values of patients actually in that specific state. [12] In particular, understanding the degree to which patients adapt to their condition could provide extremely useful information in developing more valid health state valuations by modifying current societal valuations to incorporate these effects. To do this, potential psychosocial factors contributing to this adaptation, as well as the mechanisms to which individuals cope with their asthma, need to be investigated. 6.5 Conclusions A gap in the literature has been addressed by developing an elicitation technique to evaluate the preferences which patients have towards specific attributes of asthma treatments. This has been done with a particular focus on how patients trade-off between treatment-related benefits and risks. Patients have preferences for a treatment that will result in a greater number of SFDs; however, they are willing to sacrifice treatment benefits to avoid treatment-related adverse events, despite the low probabilities associated with oral thrush and tremor/palpitation episodes. The sub-population analyses reveal that preferences for treatment attributes differ across levels of SES and asthma control. Individuals with lower incomes, less education, and better asthma control have a lower WTP to receive benefits and avoid risks when compared to those with higher incomes, more education, and poorer asthma control. Indirect, generic preference-based instruments appear to have construct validity for asthma. Because they demonstrate an ability to discriminate across levels of validated measures of asthma control at least as well as disease-specific instruments, they clearly have a valuable role to play in policy analysis and resource allocation decision-making. - 102-6.6 References 1. World Health Organization: World Health Report 2003: Shaping the future [Online]. [2003?] [cited 2006 Apr 24]; Available from: URL: http://www.who.int/whr/2003/en/index.html 2. Boulet L-P, Becker A , Berube D, Beveridge R, Ernst P. Canadian asthma consensus report, 1999. Can Med Assoc J 1999;161(11 Supp):Sl-S62. 3. Juniper EF, O'Byrne P M , Guyatt GH, Ferrie PJ, King DR. Development and validation of a questionnaire to measure asthma control. Eur Respir J 1999;14(4):902-7. 4. Stahl E, Postma DS, Svensson K, Tattersfield A E , Eivindson A, Schreurs A, et al. Formoterol used as needed improves health-related quality of life in asthmatic patients uncontrolled with inhaled corticosteroids. Respir Med 2003 ;97(9): 1061-6. 5. Szende A, Svensson K, Stahl E, Meszaros A, Berta G Y . Psychometric and utility-based measures of health status of asthmatic patients with different disease control level. Pharmacoeconomics 2004;22(8):537-47. 6. Ratcliffe J, Van Haselen R, Buxton M , Hardy K, Colehan J, Partridge M . Assessing patients' preferences for characteristics associated with homeopathic and conventional treatment of asthma: a conjoint analysis study. Thorax 2002;57:503-8. 7. Wilkinson RG. Unhealthy societies. The afflictions of inequality. New York: Routledge; 1996. 8. Gochman DS. Health behavior: emerging research perspectives. New York: Plenum Press; 1988. 9. Hensher DA, Rose JM. Respondent behavior in discrete choice modeling with a focus on the valuation of travel time savings. J Transport Stat 2005;8(2): 17-30. 10. Hensher DA, Rose JM, Greene WH. Applied choice analysis. A primer. Cambridge: Cambridge University Press; 2005. 11. Pickard S, Knight SJ. Proxy evaluation of health-related quality of life: a conceptual framework for understanding multiple proxy perspectives. Med Care 2005; 43,493-9. 12. Brazier J, Akehurst R, Brennan A, Dolan P, Claxton K, McCabe C, et al. Should patients have a greater role in valuing health states? Appl Health Econ Health Policy 2005;4:201-8. - 103 -APPENDIX I ETHICS CERTIFICATES - 104-APPENDIX II PATIENT CONSENT F O R M - 109-Subject Information and Consent Form Evaluating Health-Related Quality of Life and Patient's Preferences for Treatment in As thma PRINCIPAL INVESTIGATOR: Dr. Larry Lynd, PhD, Assistant Professor Department of Pharmaceutical Sciences, University of British Columbia (604)875-4111, ext. 63467 CO-INVESTIGATOR(S): Dr. As lam Anis, PhD, Associate Professor Department of Health Care and Epidemiology, University of British Columbia (604)822-5550 Dr. Tony Bai, MD, Professor Department of Medicine, Division of Respiratory Medicine, University of British Columbia (604)806-8346, ext. 62489 Dr. Gillian Currie, PhD, Assistant Professor Department of Economics, University of Calgary (403)220-5602 Dr. J . Mark FitzGerald, MD, Professor Department of Medicine, Division of Respiratory Medicine, University of British Columbia (604)875-4111, ext. 54565 Dr. Carlo Marra, PhD, Assistant Professor Department of Pharmaceutical Sciences, University of British Columbia (604)875-4111, ext. 61732 Ms. Helen McTaggart -Cowan, B S c , Graduate Student Department of Health Care and Epidemiology, University of British Columbia (604)875-4111, ext. 66643 S P O N S O R S : B C Lung Association and the Michael Smith Foundation for Health Research E M E R G E N C Y T E L E P H O N E N U M B E R : Dr. Larry Lynd, (604)875-4111, ext. 63467 - 1 1 0 -INTRODUCTION - "The invitation to participate" You are being invited to take part in this research study because of your current health status, age, and ability to read and write English. YOUR PARTICIPATION IS VOLUNTARY Your participation is entirely voluntary, so it is up to you to decide whether or not to take part in this study. Before you decide, it is important you to understand what the research involves. This consent form will tell you about the study, why the research is being done, what will happen to you during the study, and the possible benefits, risks, and discomforts. If you wish to participate, you will be asked to sign this form. If you decide to take part in this study, you are still free to withdraw at any time, without giving any reasons for your decision. If you do not wish to participate, you do not have to provide any reasons for your decision not to participate, nor will you lose the benefit of medical care to which you are entitled or are presently receiving. WHO IS CONDUCTING THIS STUDY? The Principle Investigator has received financial support from the B C Lung Association for the completion of this study. You are entitled to request any details concerning this compensation from the Principle Investigator. BACKGROUND Inhaled bronchodilators, known as beta-agonists (Ventolin, for example), are a commonly prescribed drug to relax the airways and to relieve the symptoms of asthma. Although no cure is yet available for asthma, the currently prescribed drugs (inhaled steroids and inhaled bronchodilators) are quite effective in minimizing symptoms, improving lung function, and possibly preventing d isease progression. Recently, there has been a concern regarding the safety of the excessive and prolonged use of inhaled beta-agonists. Although high doses may be appropriate to control severe disease, excessive use may make asthma worse. - I l l -WHAT IS THE PURPOSE OF THE STUDY? The objective of this study is to evaluate patient preferences for asthma therapy. The results from this study will allow us to inform clinicians about how to improve patients' ability to maintain their drug therapy and to aid asthma educators to tailor their programs to meet your needs. WHO SHOULD PARTICIPATE IN THE STUDY? Subjects diagnosed with asthma by a physician are eligible to participate. In addition, subjects must be between the ages of 19-49 years and fluent in reading and writing English to be eligible to participate in the study. WHO SHOULD NOT PARTICIPATE IN THE STUDY? Subjects not diagnosed with asthma by a physician are ineligible to participate. Subjects who currently have other respiratory conditions, such as upper or lower respiratory tract infections, tuberculosis, bronchiectasis or cystic fibrosis are ineligible to take part in this study. In addition, subjects younger than 19 or older than 49 years old and are not fluent in reading and writing English, are also ineligible to participate in the study. WHAT DOES THE STUDY INVOLVE? Overview of the Study During your visit at one of the following locations: the Respiratory Clinic at Vancouver General Hospital or St. Paul's Hospital, the Asthma Education Centre at St. Paul's Hospital, Eagle Ridge Hospital, or Surrey Memorial Hospital, or a participating pharmacy within the Greater Vancouver Regional District, a study coordinator, an asthma educator or a pharmacist will approach you to determine if you are willing to participate in our study. You will be required to visit one of the Respiratory Clinics to complete seven questionnaires. O n e of these questionnaires has been developed specifically to evaluate what characteristics of asthma therapy you consider are important. The remaining questionnaires will a s s e s s your asthma control and quality of life. In addition, there will be a set of questions that will evaluate your drug therapy, health system use, and socioeconomic status. Questions include those designed to see if asthma interferes with your activities of daily living, and to collect information on your (and your partner's) income, occupation, education, and residence. Y o u will also be asked for your Personal Health Number for inspection of medical records, pharmacy records of asthma drug use, and records of utilization of medical services. This information will enable the Principal Investigator to determine the significance of the findings of the study. - 112-In addition, you will be asked to perform a simple lung function test, called spirometry, which is done with a spirometer, which consists of a mouthpiece and disposable tubing attached to a machine. T o perform spirometry, you will inhale deeply, with your mouth tightly around the tube and then blowing out the air for 3-4 seconds, while measurements are being taken. T h e s e results allow us to measure how fast and how much air can be breathed out of your lungs. T h e risks are minimal for most people. Because the test involves forced and rapid breathing, some people may experience temporary shortness of breath, light headedness, and possibly some discomfort at the end of the breathing out process. You will spend approximately 45 minutes to complete the spirometry test and questionnaires. If your participation is at the Respiratory Clinics at Vancouver General Hospital or at St. Paul's Hospital, you will be immediately reimbursed $20 for your time and transportation costs. If you will complete the study at the Asthma Education Centres at St. Paul's Hospital, Eagle Ridge Hospital, or Surrey Memorial Hospital, an equivalent reimbursement will be mailed to your home address upon completion of the study. W H A T A R E THE P O S S I B L E H A R M S A N D SIDE E F F E C T S OF PARTICIPATING? There may be minimal risk involved with performing spirometry. S o m e people may experience temporary shortness of breath, light headedness, and possibly some discomfort at the end of the breathing out process. WHAT A R E THE B E N E F I T S OF PARTICIPATING IN THIS S T U D Y ? There is no direct benefit for participating, but your participation will allow us to develop a better understanding of asthma management which will assist us in the development of better management and educational programs in the future. W H A T H A P P E N S IF I DECIDE TO WITHDRAW M Y C O N S E N T TO PARTICIPATE? Your participation in this research is entirely voluntary. Y o u may withdraw from this study at any time. If you decide to enter this study and to withdraw at any time in the future, there will be no penalty or loss of benefit to which you are entitled, and your future medical care will not be affected. If you chose to enter the study and then decide to withdraw at a later time, all data collected about you during your enrolment in the study will be retained for analysis. By law, this data cannot be destroyed. -113 -WHAT H A P P E N S IF SOMETHING G O E S W R O N G ? You signing this consent form in no way limits your legal rights against the sponsor, investigators, or anyone else. WHAT WILL THE S T U D Y C O S T M E ? You will not provide any direct costs for participating in this study. All that is asked is that you volunteer your time. WILL M Y TAKING P A R T IN THE STUDY B E K E P T CONFIDENTIAL? Your confidentiality will be respected. No information that discloses your identity will be released or published without your specific consent to the disclosure. However, research records and medical records identifying you may be inspected in the presence of the Investigator or his or her designate by representatives of Health C a n a d a and the U B C Research Ethics Board for the purpose of monitoring the research. However, no records which identify you by name or initials will be allowed to leave the Investigators' offices. WHO DO I C O N T A C T IF I H A V E QUESTIONS A B O U T THE S T U D Y DURING MY PARTICIPATION? If you have any questions or desire further information with respect to this study, you may contact the Study Coordinator, Ms. Helen McTaggart -Cowan at (604)875-4111, ext. 66643 or the Principal Investigator, Dr. Larry Lynd at (604)875-4111, ext. 63467. WHO DO I C O N T A C T IF I H A V E QUESTIONS A B O U T M Y RIGHTS A S A S U B J E C T DURING T H E S T U D Y ? If you have any concerns about your treatment or rights as a research subject, you may contact the Research Subject Information Line in the U B C Office of Research Services at (604)822-8598. - 114-S U B J E C T C O N S E N T T O P A R T I C I P A T E I understand that participation in this study is entirely voluntary and that I may decline to participate in the study, or that I may withdraw from it any time. My decision to participate or withdraw will not affect any aspect of my present or future medical treatment. I hereby give permission for follow-up information to be gathered by the investigator during the course of the study. This information will include inspection of medical records, pharmacy records of asthma drug use, and records of utilization of medical services. This information will enable the principal investigator to determine the significance of the findings of the study. I have had the opportunity to have my questions answered. A copy of this consent form has been given to me. I agree to participate in this research study under the direction of the Principal Investigator, Dr. Larry Lynd. Subject Signature Date Subject Name Witness Signature Date Principal Investigator Signature Date Larry Lynd -115 -APPENDIX III PATIENTS' PREFERENCES FOR ASTHMA TREATMENT DISCRETE CHOICE EXPERIMENT (SELF-ADMINISTERED) -116-QUESTIONNAIRE TO ASSESS PATIENTS' VIEWS ON ASTHMA TREATMENTS Researchers at the University of British Columbia are carrying out a survey on the views of patients on asthma treatments. We are asking you to participate in this study by completing the attached questionnaire. Knowledge of your views will allow doctors and researchers to provide better care to you in prescribing asthma treatments. All responses will be strictly confidential and only used for the purpose of this study. The choices you make on this questionnaire will not affect your visit today. You do not have to take part in this survey, participation is completely voluntary. If you decide not to take part, it would be much appreciated if you would return the blank questionnaire to reception. Please return completed questionnaires to reception after your examination. Thank you for your time and co-operation. Yours sincerely, Dr. Larry Lynd, PhD Helen McTaggart-Cowan, BSc (Principal investigator) (Study coordinator) - 1 1 7 -BACKGROUND INFORMATION This booklet contains 26 questions. Each of the following questions ask you to state which treatment you would choose (A or B) when considering your own individual asthma condition and lifestyle preferences. Please be aware that the choices are not real treatments that are currently available but are hypothetical scenarios used to evaluate your decision-making process. You may also choose to select neither, however by selecting this option, you are saying you prefer not to have your asthma treated at all. Treatments A and B differ according to: Number of symptom free days (per month) - the options consist of 10, 15, 20, and 25 or more. This value will represent the number of times you will experience no asthma symptoms during a month. Dose (per day) - the options consist of as needed, 1, 2, and 3 or more. This value will represent the number of times you will have to take your asthma treatment during a 24-hour period. Cost (per month) - the options consist of $20, $40, $80, and $160. Because of the characteristic differences between the treatments we present to you, their costs are also different. To assess what value you would place on these different options, you are asked how much, in theory, you would be prepared to pay for them (out of your pocket), without a third party payment system. There is no possibility you would be asked to pay this amount. Thrush (episodes per year) - the options consist of none, 1, 2, and 3. Thrush is a potential side effect that may occur with use of an asthma treatment. It is an infection in the mouth that could occur up to 10 days, and remedied by visiting your family physician and using antifungal medication. -118 -Tremors/palpitations (episodes per month) - the options consist of none, 1, 2, and 3 or more. Tremors/heart palpitations are side effects that may occur with use of an asthma treatment. Tremors are uncontrolled shakings that may affect the whole body or just certain areas. Palpitations are irregular heart beats that can feel pounding, racing, skipped, or stopped. These are not potentially life-threatening but an episode could last for several hours. Number of inhalers - the options consists of 1, 2, and 3. This value will represent the number of inhalers you will use during a 24-hour period. All other factors are the same in the different treatments. This questionnaire should take you approximately 20 minutes to complete. -119-Question 1 TREATMENT A TREATMENT B Number of symptom free days (per month) 25 or more 20 Dose (per day) 1 2 Cost (per month) $80 $20 Thrush (episodes per year) 2 none Tremors/palpitations (episodes per month) 1 none Number of inhalers 3 1 Which treatment would you Treatment A Treatment B Neither prefer {check one box only)! • • • TREATMENT TREATMENT Question 2 A B Number of symptom free days (per month) 25 or more 15 Dose (per day) 1 2 Cost (per month) $40 $20 Thrush (episodes per year) 2 3 Tremors/palpitations (episodes per month) 3 or more 2 Number of inhalers 1 1 Which treatment would you Treatment A Treatment B Neither prefer {check one box only)? • • • - 120-TREATMENT TREATMENT Question 3 A B Number of symptom free days (per month) 25 or more 10 Dose (per day) 2 1 Cost (per month) $160 $40 Thrush (episodes per year) 2 2 Tremors/palpitations (episodes per month) 3 or more 3 or more Number of inhalers 2 1 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • TREATMENT TREATMENT Question 4 A B Number of symptom free days (per month) 20 15 Dose (per day) 2 as needed Cost (per month) $80 $160 Thrush (episodes per year) 2 2 Tremors/palpitations (episodes per month) 2 1 Number of inhalers 3 2 Which treatment would you Treatment A Treatment B Neither prefer [check one box only)? • • • - 121 -Question 5 TREATMENT A TREATMENT B Number of symptom'free days (per month) 20 20 Dose (per day) 1 as needed Cost (per month) $160 $80 Thrush (episodes per year) 3 2 Tremors/palpitations (episodes per month) none 2 Number of inhalers 2 3 Which treatment would you prefer {check one box only)? Treatment A • Treatment B • Neither • TREATMENT TREATMENT Question 6 A B Number of symptom free days (per month) 15 20 Dose (per day) as needed 2 Cost (per month) $40 $80 Thrush (episodes per year) 1 none Tremors/palpitations (episodes per month) 1 1 Number of inhalers 2 3 Which treatment would you Treatment A Treatment B Neither prefer {check one box only)? • • • - 122-TREATMENT TREATMENT Question 7 A B Number of symptom free days (per month) 15 25 or more Dose (per day) 2 1 Cost (per month) $160 $80 Thrush (episodes per year) none 1 Tremors/palpitations (episodes per month) 2 1 Number of inhalers 1 1 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • TREATMENT TREATMENT Question 8 A B Number of symptom free days (per month) 20 10 Dose (per day) 2 as needed Cost (per month) $40 $20 Thrush (episodes per year) 1 3 Tremors/palpitations (episodes per month) 3 or more 2 Number of inhalers 1 2 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 123 -Question 9 TREATMENT A TREATMENT B Number of symptom free days (per month) 10 10 Dose (per day) 3 or more 1 Cost (per month) $80 $20 Thrush (episodes per year) none 1 Tremors/palpitations (episodes per month) 3 or more 2 Number of inhalers 2 3 Which treatment would you prefer (check one box only)? Treatment A • Treatment B • Neither • TREATMENT TREATMENT Question 10 A B Number of symptom free days (per month) 25 or more 25 or more Dose (per day) as needed 3 or more Cost (per month) $80 $40 Thrush (episodes per year) 1 none Tremors/palpitations (episodes oer month) 3 or more 2 1 — Number of inhalers 1 1 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 124-TREATMENT TREATMENT Question 11 A B Number of symptom free days (per month) 15 10 Dose (per day) as needed 2 Cost (per month) $160 $20 Thrush (episodes per year) none none Tremors/palpitations (episodes per month) 3 or more none Number of inhalers 3 1 Which treatment would you Treatment A Treatment B Neither prefer [check one box only)? • • • TREATMENT TREATMENT Question 12 A B Number of symptom free days (per month) 25 or more 15 Dose (per day) 3 or more 3 or more Cost (per month) $40 $160 Thrush (episodes per year) none 1 Tremors/palpitations (episodes per month) 2 none Number of inhalers 2 3 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 125 -TREATMENT TREATMENT Question 13 A B Number of symptom free days (per month) 20 20 Dose (per day) 3 or more 2 Cost (per month) $80 $20 Thrush (episodes per year) 3 1 Tremors/palpitations (episodes per month) 1 3 or more Number of inhalers 1 2 Which treatment would you Treatment A Treatment B prefer {check one box only)? • • Neither • TREATMENT TREATMENT Question 14 A B Number of symptom free days (per month) 15 20 Dose (per day) 1 3 or more Cost (per month) $80 $40 Thrush (episodes per year) 3 none Tremors/palpitations (episodes per month) none none I ' — ' Number of inhalers 1 3 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 126-Question 15 TREATMENT A TREATMENT B Number of symptom free days (per month) 10 15 Dose(per day) as needed 3 or more Cost (per month) $40 $20 Thrush (episodes per year) 3 2 Tremors/palpitations (episodes per month) none none Number of inhalers 2 3 Which treatment would you Treatment A Treatment B Neither prefer {check one box only)? • • • TREATMENT TREATMENT Question 16 A B Number of symptom free days (per month) 25 or more 15 Dose (per day) 1 as needed Cost (per month) $160 $40 Thrush (episodes per year) 3 1 Tremors/palpitations (episodes per month) 3 or more 1 Number of inhalers 3 2 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 127-TREATMENT TREATMENT Question 17 A B Number of symptom free days (per month) 10 25 or more Dose (per day) 2 as needed Cost (per month) $40 $160 Thrush (episodes per year) 3 3 Tremors/palpitations (episodes per month) 2 3 or more Number of inhalers 3 3 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • TREATMENT TREATMENT Question 18 A B Number of symptom free days (per month) 15 20 Dose (per day) 3 or more 3 or more Cost (per month) $20 $20 Thrush (episodes per year) 2 3 Tremors/palpitations (episodes oer month) none 1 i *• — Number of inhalers 3 1 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 128-TREATMENT TREATMENT Question 19 A B Number of symptom free days (per month) 25 or more 10 Dose(per day) 3 or more 2 Cost (per month) $80 $160 Thrush (episodes per year) none 2 Tremors/palpitations (episodes per month) none 2 Number of inhalers 1 3 Which treatment would you Treatment A Treatment B Neither prefer [check one box only)? • • • Question 20 TREATMENT A TREATMENT B Number of symptom free days (oer month) 25 or more 10 u L — Dose (per day) 2 as needed Cost (per month) $160 $160 Thrush (episodes per year) 2 2 Tremors/palpitations (episodes per month) none 1 Number of inhalers 2 1 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 129-TREATMENT TREATMENT Question 21 A B Number of symptom free days (per month) 20 15 Dose (per day) 2 1 Cost (per month) $80 $40 Thrush (episodes per year) none none Tremors/palpitations (episodes per month) 1 3 or more Number of inhalers 3 3 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • . • • TREATMENT TREATMENT Question 22 A B Number of symptom free days (per month) 25 or more 10 Dose (per day) as needed 3 or more Cost (per month) $20 $160 Thrush (episodes per year) none 3 Tremors/palpitations (episodes per month) none 3 or more Number of inhalers 1 3 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 130-TREATMENT TREATMENT Question 23 A B Number of symptom free days (per month) 15 10 Dose (per day) 1 as needed Cost (per month) $80 $20 Thrush (episodes per year) 3 none Tremors/palpitations (episodes per month) 2 none Number of inhalers 2 1 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • TREATMENT TREATMENT Question 24 A B Number of symptom free days (per month) 25 or more 10 Dose (per day) 3 or more 1 Cost (per month) $160 $20 Thrush (episodes per year) 1 3 Tremors/palpitations (episodes per month) 3 or more 2 Number of inhalers 3 2 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 131 -TREATMENT TREATMENT Question 25 A B Number of symptom free days (per month) 20 25 or more Dose(per day) 3 or more 1 Cost (per month) $40 $20 Thrush (episodes per year) 1 none Tremors/palpitations (episodes per month) none 1 Number of inhalers 3 3 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • TREATMENT TREATMENT Question 26 A B Number of symptom free days (per month) 10 15 Dose (per day) 1 as needed 5J *•—t Cost (per month) $160 $40 Thrush (episodes per year) none 1 Tremors/palpitations (episodes per month) 1 1 Number of inhalers 2 2 Which treatment would you Treatment A Treatment B Neither prefer (check one box only)? • • • - 132-APPENDIX IV ASTHMA ASSESSMENT QUESTIONNAIRE (SELF-ADMINISTERED) -133 -S E C T I O N I: A S T H M A a n d H E A L T H C A R E U S E A S S E S S M E N T This first section starts by asking some general questions about yourself. Then there will be some questions dealing specifically with your asthma, your use of the health care system and complementary medicines, and about some of the expenses you might have incurred because of your asthma. Y o u may decline to answer any question, however please remember that it is very important that we get the most accurate and complete information we can, and that all the information you provide is completely confidential. H C 1 . What is your current marital status? • Single • Married • Married and separated • Common- law • Divorced • Widowed HC2 . What type of health insurance coverage do you have?. [Please c h e c k all that apply] • I don't currently have medical insurance • Plan C (Social assistance) • Plan E (Basic M S P ) - s e l f paid • Plan E (Basic M S P ) - employer paid • Extended medical - self paid • Extended medical - employer paid • Prescription drug plan (3 r d party coverage) • Other If O T H E R , please specify: ^ HC3 . When were you first diagnosed with asthma? / ^ Approximate date of asthma diagnosis [month / year] • I don't know • I prefer not to answer this question - 134-HC4. Do you currently, or have you previously, smoked cigarettes, cigars, or a pipe? G Never smoked • [Go to Question HC7] • Currently smoke • Quit smoking • Other • I prefer not to answer this question If OTHER, please specify: • H C 5 . If you have quit smoking, how long has it been since you last smoked? • < 3 months • 3 - 6 months • 6 - 1 2 months • 1 - 5 years • > 5 years • I don't know HC6. How much do you, or did you previously smoke? ^ Amount smoked [per day, or per week] [specify pipe, cigarette, cigar] • I don't know • I prefer not to answer this question HC7 . Over the past year, have you seen a doctor because of your asthma, other than in the emergency department or hospital? • Y e s ; How many times? • • No • I don't know • I prefer not to answer this question - 135 -HC8. Over the past year, have you had to visit an emergency department in a hospital because of your asthma? • Y e s ; How many times? • • No • I don't know • I prefer not to answer this question H C 9 . Over the past year have you been admitted to hospital due to your asthma? • Y e s ; How many times? • • No • I don't know • I prefer not to answer this question [If NO, go to Question HC12] HC10. During any of these hospital admissions, were you admitted to a critical care unit [e.g. ICU or CCU] because of your asthma? • Y e s ; How many times? ^ • No • I don't know • I prefer not to answer this question [If NO, go to Question HC12] HC11. During any of these admissions to a critical care unit, did you have to be placed on a life support system (e.g. a ventilator)? • Y e s • No • I don't know • I prefer not to answer this question - 136-H C 1 2 . H a v e you ever been enrolled in an asthma education program, or received specific instruction on asthma management from an asthma educator? HC13 . Have you ever used a peak flow meter to monitor your asthma? HC14. Do you currently use a peak flow meter to monitor your asthma? HC15 . Do you have an emergency treatment plan for the initial management of an asthma attack? HC16. Have you ever been tested for allergies using a skin test? CONTINUE TO SECTION II - 137-SECTION II: PERSONAL DATA This section is made up of questions asking about yourself and anyone who might be living with you. This involves questions about education, income, and employment of yourself and anyone who might be sharing expenses with you. P1. What was your m a i n activity during the past 12 months? [P lease c h e c k on ly one] • Working at a job • Looking for work • Unable to work due to health reasons • Going to school • Keeping house • Retired • Other If O T H E R , please specify: • P2. If your main activity was W O R K I N G A T A J O B , what is your field of employment (e.g. secretary, nurse, laborer, store clerk, plumber, dentist, etc.)? Brief description: ^ P3. If you worked, even if your main activity wasn't W O R K I N G A T A J O B , on average approximately how much did you work over the past 12 months? • Full time (>35 hours / week) • 3 0 - 3 5 hours / week • 2 0 - 2 9 hours / week • 1 0 - 1 9 hours / week • < 10 hours / week • Casua l -138-P4. If someone in your field of employment lost the income from a missed work day, what would be the value of O N E days lost income (before deductions)? $ Value of one days lost income • I don't know • Not applicable • I prefer not to answer this question P5. What is the highest grade (or year) of secondary (high school) or elementary school you have successfully completed? A Number ( 1 - 1 3 ) of grades of secondary and / or 1 ^ elementary school successfully completed • Never attended school, or attended kindergarten only • I prefer not to answer this question P6. How many years of post-secondary education (after high school) have you completed? ^ Number of years of post-secondary education • None • Less than one year PI. What certificates, degrees, or diplomas have you ever obtained? [Please check all that apply] Q None • High school diploma • Trades or non-university diploma • Undergraduate (bachelor's) university degree • Degree in medicine, dentistry, veterinary medicine, optometry, chiropractory, etc. • Masters or Doctorate degree (M.A., M . S c , Ph.D. , D . S c , D.Ed.) - 139-P8. What was your approximate total household income from all sources for the previous year, before income tax deduction? [ including only your family members, not roommates that you don't share daily expenses with] • less than $20,000 • $60,001 — $70,000 • $20,000 —$30 ,000 • $70,001 — $80,000 • $30,001 —$40 ,000 • $80,001 — $90,000 • $40,001 —$50 ,000 • $90,001 — $100,000 • $50,001 —$60 ,000 • greater than $100,000 • I don't know • I prefer not to answer this question P9. Do you have any children? • Y e s • No P10. Continued: If Y E S : How many children do you have? Total number or children How many children currently live with you? ^ Number of children currently living with you CONTINUE TO SECTION III - 140-SECTION III: A S T H M A MEDICATION U S E This section asks some very specific questions regarding any asthma medications that you use. It is very important that we get the most accurate information that we possibly can. S o m e of the questions ask for very specific details, and deal with some things that may have happened over the past year, so please take your time and try and answer the questions as accurately as possible. O n c e again, please remember that any answers you give are completely confidential. M1. In the past year, have you used a short-acting bronchodilator (blue) inhaler, such as Ventolin® or salbutamol, for the treatment of your asthma? • Y e s • No • [Go to Question M10] • I don't know • I prefer not to answer this question Y o u have indicated that you currently use, or have used, a bronchodilator (blue) inhaler for the treatment of your asthma. The next few questions ask specifically about this/these inhaler(s). P lease try and answer all of the next few questions as they pertain to your bronchodilator (blue) inhaler only. M2. C a n you identify which bronchodilator (blue) inhalers you currently use, or have used in the past year? (please check all that apply) Drug name Currently use Used in past year fenoterol (e.g. Berotec®) • • salbutamol (e.g. Ventolin®, Ventodisk®) • • terbutaline (e.g. Bricanyl®) • • Other: please specify. • • - 141 -Is your bronchodilator (blue) inhaler a multi-dose inhaler, like a puffer, or a dry powder inhaler such as rotohaler capsules or disc, or do you use both types? • Multi-dose inhaler (Puffer) • Dry powder inhaler (capsule or disc) • I use both types How regularly do you use your bronchodilator (blue) inhaler(s)? • Every day Q Less than daily, but more than once a week • Approximately once / week • Less than once a week • Only when I exercise Q Only when I have an asthma attack • Other > Go to Question M6 If O T H E R , please specify: • If you use your bronchodilator (blue) inhaler daily, how many puffs per day do you use? • 1 - 2 puffs / day • 3 - 4 puffs / day • 5 - 8 puffs / day • >8 puffs / day Approximately how many bronchodilator (blue) inhalers (canisters) in total have you used in the past month? • <1 • 2 - 3 • 4 - 5 • 6 - 8 • >8 - 142-M7. Approximately how many bronchodilator (blue) inhalers (canisters) in total have you have used in the past year? • <4 • 5 - 8 • 9 - 1 2 • 1 3 - 2 0 • >20 M8. Do you keep extra bronchodilator (blue) inhalers at various locations, such as the office or in the car, just in case you forget to take your inhaler with you? • Y e s • No • [Go to Question M10] • I don't know • I prefer not to answer this question M9. Approximately how many extra bronchodilator (blue) inhalers do you have (that you are using) right now? • 1 - 2 • 3 - 4 • 5 - 7 • >8 That is all the questions specifically about your bronchodilator (blue) inhaler. The next set of questions asks about a different type of inhaler. M10. In the past year, have you used a steroid, anti-inflammator (brown/orange) inhaler, such as Beclovent®, Pulmicort®, Qvar®, or Flovenf for the management of your asthma? • Y e s • No • [Go to Question M17] • I don't know • I prefer not to answer this question - 143 -You have indicated that you currently use, or have used, a steroid, anti-inflammatory (brown/orange) inhaler for the treatment of your asthma. The next set of questions asks you specifically about this/these inhaler(s). Please try and answer all of the next few questions as they pertain to your steroid, anti-inflammatory (brown/orange) inhaler only. M11. C a n you identify which steroid, anti-inflammatory (brown/orange) inhaler(s) you have used over the previous year? (Please check all that apply) Drug name Currently use Used in past year beclomethasone (e.g. Beclovent®, Becloforte®, Beclodisk®, Vanceril®) • • budesonide (e.g. Pulmicort®) • • flunisolide (e.g. AeroBid®) • fluticasone (e.g. Flovent®) • • triamcinolone (e.g. Azmacort®) • • Other: please specify: • • M12. Is your steroid, anti-inflammatory (brown/orange) inhaler a muti-dose inhaler, like a puffer, or a dry powder inhaler such as rotohaler capsules, or do you use both types? • Multi-dose inhaler (Puffer) • Dry powder inhaler (capsule or disc) • I use both types M13. How often do you use your steroid, anti-inflammatory (brown/orange) inhaler? U l Every day • Less than daily, but more than once a w e e k ^ • Approximately once / week • Less than once a week • Only when I exercise • Only when I have an asthma attack • Other s . G o to Question M15 If OTHER, please specify: • - 144-M14. If you use your steroid, anti-inflammatory (brown/orange) inhaler every day, how many puffs per day do you use? • 1 - 2 puffs / day • 3 - 4 puffs / day • 5 - 8 puffs / day • >8 puffs / day M15. Approximately how many steroid, anti-inflammatory (brown/orange) inhalers (canisters) in total have you used in the past month? • <1 • 2 - 3 • 4 - 5 • 6 - 8 • >8 M16. Approximately how many steroid, anti-inflammatory (brown/orange) inhalers (canisters) in total have you used in the past year? • <4 • 5 - 8 • 9 - 1 2 • 1 3 - 2 0 • >20 M17. In the past year, have you used a combination inhaler, such as Advair®, Symbicort®, or C o m b i v e n t ® for the management of your asthma? • Y e s • No • [Go to Question M24] • I don't know • I prefer not to answer this question - 145 -You have indicated that you currently use, or have used, a combination inhaler for the treatment of your asthma. The next set of questions asks you specifically about this/these inhaler(s). P lease try and answer all of the next few questions as they pertain to your combinat ion inhaler only. M18. Can you identify which combination inhaler(s) you have used over the previous year? (Please check all that apply) Drug name Currently use Used in past year Advair® (salmeterol/fluticasone Combination) • • Symbicort® (formoterol/budesonide Combination) • • Combivent® (ipratropium/salbutamol Combination) • • Other: please specify: • • M19. Is your combinat ion inhaler a multi-dose inhaler, like a puffer, or a dry powder inhaler such as rotohaler capsules, or do you use both types? • Multi-dose inhaler (Puffer) • Dry powder inhaler (capsule or disc) • I use both types M20. How often do you use your combination inhaler? • Every day • Less than daily, but more than once a week • Approximately once / week • Less than once a week ^ G o to Question M22 Q Only when I exercise • Only when I have an asthma attack • Other If OTHER, please specify: • - 146-M21. If you use your combination inhaler every day, how many puffs per day do you use? • 1 - 2 puffs / day • 3 - 4 puffs / day • 5 - 8 puffs / day • >8 puffs / day M22. Approximately how many combination inhalers (canisters) in total have you used in the past month? • <1 • 2 - 3 • 4 - 5 • 6 - 8 • >8 M23. Approximately how many combination inhalers (canisters) in total have you used in the past year? • 4 • 5 - 8 • 9 - 1 2 • 1 3 - 2 0 • >20 M24. In the past year, have you used a long-acting bronchodilator (light blue) inhaler, such as Serevent® (salmeterol) or Oxeze® (formoterol), for the treatment of your asthma? • Y e s • No • [Go to Question M27] • I don't know • I prefer not to answer this question - 147-M25. How regularly do you use your Serevent® or Oxeze® (green) inhaler? d Every day • Less than daily, but more than once a week • Approximately once / week Q Less than once a week Q Only when I exercise • Only when I have an asthma attack • Other > G o to Question M27 M26. If you use your Serevent® or Oxeze® (green) inhaler every day, how many puffs per day do you use? • 1 - 2 puffs / day • 3 - 4 puffs / day • 5 - 8 puffs / day • >8 puffs / day M27. Do you ever have prescriptions for your inhalers filled and then share the inhalers with someone else, such as another member of your family or a friend? • Y e s • No • I don't know • I prefer not to answer this question M28. Are you currently taking a steroid (such as prednisone tablets) by mouth to control your asthma? • Y e s ; How long have you been taking it? ^ m o n t h s / y e a r s • No ' ' • I don't know • I prefer not to answer this question - 148-M29. If you are not currently taking a steroid tablet by mouth, have you had to take any at any time over the previous year? • Y e s ; How many times? ^ • No 1 1 • I don't know • I prefer not to answer this question M30. Over the past year have you used, or are you currently using, any other medications for the treatment of your asthma? • Y e s • No • [Go to Question M32] • I don't know Q I prefer not to answer this question M31. C a n you identify these medications? (Please check all that apply) Drug name Currently use Used in past year cromolyn sodium (e.g. Intal®, Rynacrom®, Cromolyn®) • • ipratropium bromide (e.g. Atrovent,®) • • montelukast (e.g. Singulair®) or zafirlukast (e.g. Accolate®) • • nedocromil (e.g. Tilade®) • • theophylline (e.g. Theo-Dur®, Quibron®, Slo-Bid®, Uniphyl®) • • other: please specify: • • M32. Do you use a spacer with any of your inhalers? • Y e s • No • I don't know D I prefer not to answer this question -149-M33. Other than your asthma, do you have any other chronic d iseases (such as high blood pressure, arthritis, diabetes, angina, depression) that have been diagnosed by your doctor? • Y e s • No • [Go to Question M35] I don't know • I prefer not to answer this question M34. If you have other chronic diseases in addition to your asthma, can you list these d iseases? 1. 4. 2. 5. 3. 6. M35. Overall, how would you describe the severity of your asthma? [Please circle one] 1 2 3 4 5 Very Mild Mild Moderate Severe Very Severe M36. Overall, how would you classify the control of your asthma? [Please circle one] 1 2 3 4 5 Very Well Well Adequately Not Well Not Controlled Controlled Controlled Controlled Controlled At All M37. Below is a line with '0' at the left-hand end and a '1' at the right-hand end The '0' represents "death", the T represents "perfect health", and the area in between represents a state of health somewhere in between. Make a mark on the line at the point that you feel represents how you feel today. o 1 Death Perfect Health - 150-APPENDIX V STANDARDIZED VERSION OF THE ASTHMA QUALITY OF LIFE QUESTIONNAIRE (SELF-ADMINISTERED) - 151 -A S T H M A Q U A L I T Y O F LIFE Q U E S T I O N N A I R E (S) P A T I E N T ID S E L F - A D M I N I S T E R E D D A T E Page 1 of 5 P l e a s e c o m p l e t e all q u e s t i o n s b y c i r c l i ng the n u m b e r tha t b e s t d e s c r i b e s h o w y o u h a v e been du r ing the last 2 weeks as a result of your asthma. H O W LIMITED H A V E Y O U B E E N DURING THE LAST 2 WEEKS IN T H E S E ACT IV IT IES AS A RESULT OF YOUR ASTHMA? Totally Extremely Very Moderate Some A Little Not at all Limited Limited Limited Limitation Limitation Limitation Limited 1. S T R E N U O U S ACTIV IT IES (such as hurry ing, exerc is ing , running up stairs, sports) 2. M O D E R A T E ACT IV IT IES (such as wa lk ing , housework , gardening, shopping, c l imbing stairs) 3. S O C I A L ACTIV IT IES (such as ta lk ing, p lay ing w i th pets /ch i ldren, v is i t ing fr iends/relat ives) 4 . W O R K - R E L A T E D A C T I V I T I E S * ( tasks you have to do at work) 5. S L E E P I N G H O W M U C H DISCOMFORT OR DISTRESS H A V E Y O U FELT DURING THE LAST 2 WEEKS? A Very A Great A Good Moderate Some Very None Great Deal Deal Deal Amount Little 6. H o w much d iscomfor t or d is t ress have you felt over the last 2 weeks as a result of C H E S T 1 2 3 4 5 6 7 T I G H T N E S S ? 1 - 152-A S T H M A Q U A L I T Y O F L IFE Q U E S T I O N N A I R E (S) P A T I E N T ID S E L F - A D M I N I S T E R E D D A T E Page 2 of 5 IN G E N E R A L , HOW MUCH OF THE TIME DURING THE LAST 2 WEEKS DID Y O U : 7. Feel C O N C E R N E D A B O U T H A V I N G A S T H M A ? 8. Feel S H O R T OF B R E A T H as a result of your as thma? 9. Exper ience a s t h m a s y m p t o m s as a R E S U L T O F BEING E X P O S E D TO C I G A R E T T E S M O K E ? 10 . Exper ience a W H E E Z E in your ches t? All of the Time Most of the Time A Good Bit of the Time Some of the Time A Little ol the Time Hardly Any of the Time None ol the Time 1 1. Feel you had to A V O I D A S I T U A T I O N OR E N V I R O N M E N T B E C A U S E OF C I G A R E T T E S M O K E ? H O W M U C H DISCOMFORT OR DISTRESS H A V E Y O U R FELT DURING THE LAST 2 WEEKS? A Very A Great A Good Moderate Some Very None Great Deal Deal Deal Amount Little 12. H o w m u c h d i scomfo r t or d is t ress have you felt over the last 2 w e e k s as a result of 2 3 4 5 6 7 C O U G H I N G ? IN G E N E R A L , HOW M U C H OF THE TIME DURING THE LAST 2 WEEKS DID Y O U : 1 3. Feel F R U S T R A T E D as a result of your a s t h m a ? 14. Exper ience a feel ing of C H E S T H E A V I N E S S ? All of Most A Good Some A Little Hardly None the of the Bit of the o l the of the Any of of the Time Time Time Time Time the Time Time 2 - 153 -A S T H M A QUAL ITY OF LIFE QUESTIONNAIRE (S) PATIENT ID SELF-ADMINISTER ED DATE Page 3 of 5 IN G E N E R A L . HOW M U C H OF THE TIME DURING T H E LAST 2 WEEKS DID Y O U : AM of Most A Good Some A Little Hardly None the of the Bit.of the of the of the Any of of the Time Time Time Time Time the Ttme Time 15. Feel C O N C E R N E D A B O U T THE NEED TO U S E M E D I C A T I O N for 1 2 3 4 5 6 7 your as thma? 16 . Feel the need to C L E A R Y O U R T H R O A T ? 1 2 3 4 5 6 7 17. Exper ience a s t h m a s y m p t o m s as a R E S U L T O F BE ING E X P O S E D T O 1 2 3 4 5 6 7 D U S T ? 18. Exper ience D I F F I C U L T Y B R E A T H I N G O U T as a result of your 1 2 3 4 5 6 7 as thma? 19 . Feel you had to A V O I D A S I T U A T I O N OR E N V I R O N M E N T 1 2 3 4 5 6 7 B E C A U S E OF D U S T ? 2 0 . W A K E U P IN T H E M O R N I N G WITH A S T H M A S Y M P T O M S ? 1 2 3 4 5 6 7 2 1 . Feel A F R A I D O F N O T H A V I N G Y O U R A S T H M A M E D I C A T I O N 1 2 3 4 5 6 7 A V A I L A B L E ? 2 2 . Feel bothered by H E A V Y B R E A T H I N G ? 1 2 3 4 5 6 7 2 3 . Exper ience a s t h m a s y m p t o m s as a R E S U L T OF T H E W E A T H E R OR AIR 1 2 3 4 5 6 7 P O L L U T I O N O U T S I D E ? 24 . Were y o u W O K E N A T N I G H T by your as thma? 1 2 3 4 5 6 7 2 5 . A V O I D OR LIMIT G O I N G O U T S I D E B E C A U S E O F T H E W E A T H E R O R 1 2 3 4 5 6 7 AIR P O L L U T I O N ? 3 - 154-A S T H M A Q U A L I T Y O F LIFE Q U E S T I O N N A I R E (S) S E L F - A D M I N I S T E R E D P A T I E N T ID , D A T E Page 4 of 5 IN GENERAL , HOW MUCH OF THE TIME DURING THE LAST 2 WEEKS DID Y O U : 26. Experience asthma symptoms as a RESULT OF BEING E X P O S E D TO S T R O N G S M E L L S OR PERFUME? 27. Feel. AFRAID OF GETTING OUT OF BREATH? 28. Feel you had to AVOID A SITUATION OR ENVIRONMENT B E C A U S E OF S T R O N G S M E L L S OR PERFUME? 29. Has your asthma INTERFERED WITH GETTING A G O O D NIGHT'S SLEEP? -30. Have a feeling of FIGHTING FOR AIR? All of the Time Most of the Time A Good Bit of the Time Some of the Time A Little of the Time Hardly None Any of of the the Time Time 3 4 5 6 7 3 4 5 6 7 3 4 5 6 7 HOW LIMITED H A V E Y O U BEEN DURING THE LAST 2 WEEKS? 31 . Think of the O V E R A L L RANGE OF ACTIVITIES that you would have liked to have done during the last 2 weeks. How much has your range of activit ies been limited by your asthma? Most Not Several Very Few No Done , Not Done Not Done Limitation 4 - 155 -A S T H M A QUAL ITY OF LIFE QUESTIONNAIRE (S) PATIENT ID SELF-ADMINISTERED DATE . Page 5 of 5 H O W LIMITED H A V E Y O U B E E N DURING THE LAST 2 WEEKS? Totally Extremely Very Moderate Some A Little Not at all Limited Limited Limited Limitation Limit ation Limitation Limited 3 2 . Overa l l , among A L L T H E A C T I V I T I E S that you have done during the last 2 1 2 3 4 5 6 7 w e e k s , h o w l imi ted have you been by your a s t h m a ? DOMAIN CODE: Symptoms: 6. 8, 10, 12, 14, 16, 18, 20 , 22 , 24. 29, 30 Activi ty Limitation: 1, 2, 3, 4 , 5, 11, 19, 25 . 28 , 31 . 32 Emotional Function: 7, 13, 15, 21 , 27 Environmental Stimuli: 9, 17, 23 , 26 5 - 156-APPENDIX VI ASTHMA CONTROL QUESTIONNAIRE (SELF-ADMINISTERED) - 157-1.U-'. JUNIPER UT Al.. Appendix ASTHMA CONTROL QUESTIONNAIRE^";) Plcnse answer questions 1-6. Circle .he number of (he response that best describes how you have been during (he pas. we, d°Urin7thc n i g h " 1 " W e C k ' h ° W o f t c n WCrC W 0 k c n *>y »^n« (, On average, during-.he pas. week, how bad were your asthma .symptoms when ,vou woke up m (he morning? ' 3. In general, during the post week, how limited were you In your activities because of your asthma? In general during (he pa.st week, how much shortness of brenth did V im experience because of you asthma? In general, during Hie past week, how much of the lime did you wheeze? On avenige,-during (he past week, how many puffs «r shon-actine bronchodtlntnr (eg. Ventolin) have you used each day' To be completed by a manner of the clinic staff 7. F l iV i prc-broneliodilator: R-V i predicted FEVi % predicted (Record-actual values on (he dotted lines and .score the FEVi % predicted in (he next column) 5 6 0 I 2 3 4 5 6 0 I 2 3 4 5 (i Never Hardly ever A few minutes Several limes Many limes A gieal many times Unable in sleep because of asthma No sympioms Very mild symptoms Mild symptoms Moderate symptoms Quite severe symptoms Severe symptoms Very severe symptoms Not limited al all Very slightly limited Slightly limited Moderately limited Very limiicd Extremely limited Totally limited None A very little A little A moderate amount Qui.e a lol A great deal A very great deal Not a. all Hardly any of the time A little of the time A moderate amount of the lime A lot of the time Most of the time All the lime None I -2 puffs most days 3-4 puffs most days 5-K puffs most days ()-I2 pufls most days 13-16 puffs most days More than 16 puffs most days >95% predicted 95-90% N9 •«()% 79-70% 59-50% <50% predicted srrEs sssr** i s c o p y r i 8 h , e d - " ™ y " o i * * « • « » • or s o . d ( p a p e r „ r s a n w a r c ) w i t h t m l l l c - 158-

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.831.1-0092726/manifest

Comment

Related Items