- Email: [email protected]
Utilities for Type 2 Diabetes Treatment-Related Attributes in a South Korean and Taiwanese Population
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Utilities for Type 2 Diabetes Treatment-Related Attributes in a South Korean and Taiwanese Population Narayan Rajan, MA, MSc1,*, Kristina S. Boye, PhD, MS, MPH, RPh1, Meaghan Gibbs, BSc, MSc2, Yoon Ji Lee, BSc3, Peter Davey, BA, MA4, Mark Ball, BCom (Hons)4, Steve M. Babineaux, MA, MSc1 1 Eli Lilly, Indianapolis, IN, USA; 2Information Fulfillment Ltd., Hong Kong, China; 3Eli Lilly Korea, Seoul, Korea; 4PRIMA Consulting Group, Sydney, New South Wales, Australia
AB STR A CT
Objectives: To elicit utilities associated with type 2 diabetes medication-related attributes from South Korean and Taiwanese populations and to identify key drivers of preferences. Methods: Data from 59 respondents from the general population in South Korea and Taiwan were analyzed. Respondents’ preferences were elicited using a paper-based standard gamble questionnaire. Health states were designed to identify the utility or disutility of type 2 diabetes medication-related attributes, including dose frequency, nausea/vomiting (hereafter referred to as nausea), and weight change. Results: The mean utility for the basic health state (encompassing current body weight and no nausea) was 0.754 ⫾ 0.155 with weekly dose administration. Respondents showed a preference for weekly over daily administration (average increase in utility of 0.043 across all health states with weekly, vs. daily, administration). Nausea was associated with a decrease in utility (average decrease of 0.034 across all health states with, vs. without, nausea). Weight gain had
little effect on utility (average decrease of 0.000 and 0.001 across all health states with, vs. without, 3% and 5% gain, respectively), although weight loss was associated with a small increase in utility (average increase of 0.028 and 0.029 across all health states with, vs. without, 3% and 5% loss, respectively). Conclusions: Utilities associated with type 2 diabetes medication-related attributes were elicited from a general population sample from South Korea and Taiwan. Treatment-related attributes, in particular dose frequency and nausea, had a measurable effect on utility and should be considered when selecting treatment regimens for South Korean or Taiwanese patients with type 2 diabetes. Keywords: Asia, Korea, preference, Taiwan, treatment-related attributes, type 2 diabetes mellitus, utility.
Introduction
treatment preferences in type 2 diabetes, including glycemic control, weight control, the risk of hypoglycemia and gastrointestinal adverse effects, and the frequency of dose administration [9–13]. Dose frequency is one of the key attributes of antihyperglycemic treatments that affect patient preference [13,14]. An increasing frequency of administration is associated with a greater perceived treatment burden and reduced HRQOL [15,16]. A study of patients’ attitudes toward a once-weekly injectable glucose-lowering medication option found that perceived benefits included convenience, improved adherence, improved quality of life, and a reduced sense of treatment burden, whereas concerns included dosage consistency over time and potential forgetfulness [12]. In addition, many patients with type 2 diabetes are overweight or obese, and some antihyperglycemic medications are associated with changes in weight. For example, insulin can cause weight gain, whereas glucagon-like peptide-1 receptor agonists are usually associated with weight loss [17]. Compared with white populations, Asians have a higher percentage of body
The prevalence of type 2 diabetes mellitus in Asian populations is increasing rapidly, and it has been estimated that by 2025 the disease will affect approximately 180 million people in Asia [1]. Among East Asian countries, the prevalence increased from 7.7% to 11.8% in Korea between 2001 and 2009 [2], and from 5.8% in 2000 to 8.3% in 2007 in Taiwan [3]. The associated increase in disease burden will have an impact on health care costs and clinical outcomes in the region, and there will be an increasing need for studies that evaluate the pharmacoeconomics, including cost-utility, of treatments for diabetes [4]. In this context, there are several points that need to be considered, including the most relevant health-related quality-of-life (HRQOL) parameters to assess and the most appropriate methods of evaluation to use within the East Asian region. Diabetes and its complications are known to have an adverse influence on HRQOL [5–8], and consequently it is also important to understand the effect that treatments for diabetes have on patients’ well-being. Several factors are known to affect
Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article. * Address correspondence to: Narayan Rajan, Eli Lilly Australia Pty Ltd., 112 Wharf Road, West Ryde, New South Wales 2114, Australia. E-mail: [email protected] 2212-1099$36.00 – see front matter Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. http://dx.doi.org/10.1016/j.vhri.2015.11.006
68
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
fat [18] and visceral fat [19] for any given body mass index, and are at risk of developing diabetes at a lower body mass index [20]. It is known that body weight can affect quality of life [21–24], and also that ethnicity can influence HRQOL in patients with diabetes [25,26]; it is, however, not clear whether the differences in percentage body fat and associated health risks seen between Asian and white populations result in different utility values for these populations. Consequently, it would be appropriate to include dose frequency and weight change in utility studies evaluating preferences related to diabetes medications. Various methods can be used to elicit utilities for subsequent use in cost-utility analyses, and the preferred method can vary between different countries and regions. The Health Insurance Review and Assessment Service of South Korea guidelines on economic evaluations indicate a preference for administering a multiattribute utility instrument to a patient population or administering a direct method, such as the standard gamble (SG), to a general population sample [27]. There is no consensus among experts as to whether it is preferable for valuations to be made by patients or by the general public [28]. Among arguments in favor of deriving utilities from the general population for hypothetical health states, there are two key points. First, the general public will tend to value health states in an unbiased manner, because they have no vested interest in a particular disease or health state [28]. Second, in publicly funded health care systems, society bears the costs of health care decisions, and therefore the general public’s perspective may be the most relevant [28]. Given the increasing importance of pharmacoeconomic evaluations within East Asia, it is important to understand the implications of adapting studies conducted elsewhere to specific Asian settings and to identify values that are appropriate for calculating cost-utility ratios in countries within this region. The aims of the present study were to elicit utility values associated with type 2 diabetes medication-related attributes from South Korean and Taiwanese populations and to identify the key drivers of preferences within these populations.
Methods Survey Population Study participants came from the general population in South Korea and Taiwan. People who were diagnosed with type 2 diabetes were excluded to avoid potentially perceived bias due to disease and treatment experience. The sample included both men and women and incorporated a mix of age and income groups. All participants were older than 18 years and were able to read a survey in their local language and complete a relatively complicated questionnaire. People with health-related occupations were excluded. All participants provided written informed consent. Ethics approval was not required for the study. The study sample was recruited from random households in the areas of Seoul and Taipei. The sample was then balanced to the South Korean and Taiwanese populations according to local demographic profiles. Participants were contacted via phone and followed up by post through a letter introducing the survey.
Selection of Health States and Utility Instruments Health states for the present study were derived from two studies performed in the United Kingdom by Matza et al. [29] and Boye et al. [13]. The basic health state represented a patient with type 2 diabetes with glucose levels under control and with no complications (at current weight and without nausea/vomiting [hereafter referred to as nausea]). Other health states were designed to identify the utility or disutility of type 2 diabetes medication-
Table 1 – Attributes included in the health state descriptions. Basic health state Dose frequency Basic health state plus one of the following: In addition to taking oral medication (pills or tablets), you give yourself injections every day (once or twice per day) In addition to taking oral medication (pills or tablets), you give yourself injections once a week Nausea/vomiting Basic health state plus one of the following: You sometimes experience nausea/vomiting You do not experience nausea/vomiting Weight Basic health state plus one of the following: You weigh ___ kg (which is 5% more than you weigh now) You weigh ___ kg (which is 3% more than you weigh now) You weigh ___ kg (which is 3% less than you weigh now) You weigh ___ kg (which is 5% less than you weigh now) Current health state Worst health state You have had type 2 diabetes for several years You take oral medications (pills or tablets) Your blood sugar levels are not in control (sometimes resulting in excessive thirst, frequent urination, fatigue, irritability, and/or blurry vision) You have tingling or prickling sensations in your hands or feet You have shortness of breath during physical activity You have pain in your legs In addition to taking oral medications (pills or tablets), you give yourself injections every day (once or twice per day) You sometimes experience nausea/vomiting You weigh ___ kg (which is 5% more than you weigh now)
related attributes, including weight change and nausea, and were adapted to include attributes related to dose frequency for the present study. The attributes included in the health states for this study are presented in Table 1. Given that the participants were from the general population, weight change was based on the average baseline weight of patients in a large clinical trial that evaluated exenatide in Asian patients (69 kg) rather than on a participant’s own weight [30]. The utility instrument selected for use in the study was the SG. The paper-based SG questionnaire is a reliable means of measuring respondents’ preferences [31,32] and has been shown to be feasible and reliable for use in a Korean population [33].
Utility Sessions Participants attended one of six utility sessions (three in Taipei, Taiwan, and three in Seoul, South Korea), all of which delivered the same presentation in the local language of that country. Each session involved a maximum of 12 participants. Overseers monitored the sessions to assist with questions and verify responders’ logic. All participants completed a form covering demographic information. At the start of the session, participants were given a presentation on type 2 diabetes that contained information taken from published literature and did not reference any specific treatment therapy. This was followed by a 10-question comprehension test to confirm they had understood the information. Participants were familiarized with the various health states by completing a visual analogue scale under the supervision of a trained interviewer; each health state was rated on a preference assessment
69
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
rating scale ranging from 10 (best possible health state) to 0 (worst possible health state). Participants then completed a paper-based SG questionnaire for each health state. To help ensure consistency and reliability, the same lead investigator conducted all utility sessions in South Korea and Taiwan. A local support team that was present for all sessions in that country assisted the lead investigator, and all the members of the support team received the same training from the lead investigator before the utility sessions.
Scaling Methods The SG questionnaire elicits patients’ preferences for health outcomes by determining how much risk they would be willing to take to avoid remaining in a poor health state [34,35]. The paper-based questionnaire presented the worst health state to the respondent, who was offered two alternatives. Alternative 1 was a treatment with two possible outcomes: either the respondent was returned to normal health and lived for an additional T years (probability P) or the patient died immediately (probability 1 P). Alternative 2 had a certain outcome of chronic state i for life (T years). The probability P was varied until the respondent was indifferent between the two alternatives, at which point the required preference value for health state i was simply P [36]. The probabilities were varied using a ping-pong method. When using immediate death as the treatment-failure option, a possible problem is that respondents may not be willing to accept any chance of treatment failure when non–life-threatening or temporary states of poor health are being valued. In this situation, it may appear that non–life-threatening and temporary health states are valued as highly as full health, simply because the basic reference SG questionnaire is not sensitive enough to capture the true underlying preferences. To overcome this problem, health states were indirectly linked to death, a technique referred to as “chaining” [37,38]. When valuing the remaining, less severe, health states, a nonfatal health outcome (the worst health state, i.e., the most severe nonfatal state) was used instead of immediate death as the treatment-failure outcome. Participant ranking of all heath states confirmed the worst health state.
Statistical Methods The primary results of this study (i.e., mean utility values) were descriptive in nature and therefore there was no key statistical comparison involving significance testing on which a sample size should be based. Hence, a power analysis was not applicable for determining the target sample size of this study. Descriptive statistics were used to summarize utilities in terms of mean and SD. Respondents who failed the logic test (i.e., gave four or more illogical responses) or who were unable to complete the survey were excluded from the calculation. An SG chaining method was used to detect differences associated with weight changes, nausea, and dose frequency. The chained SG was derived by calculating the raw scores. Participants rated the worst health state in comparison to death, thereby allowing the raw score of less severe health states to be adjusted using the formula r (1 v) þ v, where r was the raw score of the worst health state and v was the raw score of the valued health state. Results were presented descriptively. Disutility or increased utility was calculated as the difference between health state values with or without specific attributes. An example of the calculation of the disutility associated with nausea would be as follows: value for basic health state at current weight with nausea (weekly administration) minus value for basic health state at current weight without nausea (weekly administration). Average values for disutility or increased utility were the average
Table 2 – Populations’ demographic characteristics. Characteristic
Sex: male/female (n) Mean age (y) Marital status (%) Single Married Divorced Widowed Highest educational level (%) Did not complete high school High school diploma Degree/equivalent qualification Master’s degree/doctorate
General population (N ¼ 59) 28/31 43 39 58 2 1 0 12 75 13
differences across all health states with, versus without, the attribute of interest.
Results The study recruited 67 participants from the general population of South Korea (n ¼ 32) and Taiwan (n ¼ 35). Participants who did not complete the questionnaire (n ¼ 5) or failed the logic or comprehension test (n ¼ 3) were excluded. The demographic characteristics of the final study population (n ¼ 59) are summarized in Table 2. The mean age of the group was 43 years, just over half were married (58%), and most had received tertiarylevel education or higher. The mean time taken to complete the survey was 1.5 hours. Utilities derived using the SG in general population respondents from South Korea and Taiwan who completed the study (n ¼ 59) are presented in Table 3. The mean utility associated with the basic health state with weekly dose administration was 0.754 ⫾ 0.155. Respondents indicated that nausea was associated with disutility (average decrease of 0.034 in utility across all health states with, vs. without, nausea). They showed a preference for weekly dose administration, with an average increase in utility, compared with daily administration, of 0.043 (average across all health states with weekly, vs. daily, administration). Respondents had no clear preference with respect to weight gain (average change in utility of 0.000 for a 3% weight gain and 0.001 for a 5% weight gain), although weight loss appeared to be associated with increased utility (average increase of 0.028 for a 3% weight loss and 0.029 for a 5% weight loss).
Discussion Localized utility studies provide insight into geographical and cultural differences in preferences related to health states, and can be designed to use the most appropriate methodology for a particular country or region. The present study used a method in keeping with the recommendations of the Health Insurance Review and Assessment Service of South Korea guidelines on economic evaluations [27]. The study obtained local utility values associated with key attributes related to type 2 diabetes treatments in South Korean and Taiwanese people, using the SG method and a general population sample. The results showed that treatment-related attributes, in particular dose frequency and nausea, would be expected to have a measurable effect on patients’ HRQOL. Although additional research may be needed to further generalize these findings, because of the limited number of subjects
70
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
Table 3 – Utility values derived using the SG method in the general population (N ¼ 59). Health state
Daily dose administration Basic health state (current weight) þ no nausea Basic health state (current weight) þ nausea Basic health state þ 3% higher weight þ nausea Basic health state þ 5% higher weight þ nausea Basic health state þ 3% higher weight þ no nausea Basic health state þ 5% higher weight þ no nausea Basic health state þ 3% lower weight þ nausea Basic health state þ 5% lower weight þ nausea Basic health state þ 3% lower weight þ no nausea Basic health state þ 5% lower weight þ no nausea Weekly dose administration Basic health state (current weight) þ no nausea Basic health state (current weight) þ nausea Basic health state þ 3% higher weight þ nausea Basic health state þ 5% higher weight þ nausea Basic health state þ 3% higher weight þ no nausea Basic health state þ 5% higher weight þ no nausea Basic health state þ 3% lower weight þ nausea Basic health state þ 5% lower weight þ nausea Basic health state þ 3% lower weight þ no nausea Basic health state þ 5% lower weight þ no nausea
Utility value (mean ⫾ SD) 0.705 ⫾ 0.165 0.663 ⫾ 0.193 0.676 ⫾ 0.170 0.674 ⫾ 0.173 0.707 ⫾ 0.164 0.703 ⫾ 0.160 0.700 ⫾ 0.162 0.704 ⫾ 0.161 0.736 ⫾ 0.151 0.737 ⫾ 0.155
0.754 ⫾ 0.155 0.721 ⫾ 0.160 0.717 ⫾ 0.165 0.711 ⫾ 0.160
difference in findings might reflect a greater awareness among patients with type 2 diabetes of the dangers of weight gain and the weight-related problems associated with some treatments for diabetes. It is notable that utility values obtained from respondents in our study for changes in weight were similar whether the change was 3% or 5%. The study had several potential limitations. The first limitation was the use of a general population sample to derive utilities. The general public lack experience of the disease under consideration and may overemphasize negative aspects of health states [28], whereas patients with type 2 diabetes have knowledge and experience of the disorder and associated treatments. However, there is a risk that patients could provide biased judgments because they have vested interests, or alternatively may assign higher values to health states because of adaptation to their condition or lowered expectations, whereas the general public is more likely to provide an unbiased judgment [28]. There is no consensus as to which population is the most appropriate to use, and both approaches are valid [28]. The present study reported utility values elicited from the general population using the SG method, an approach acceptable to many health care payers [27,45,46]. Respondents were informed about type 2 diabetes, and their understanding was tested, to help ensure that the utilities obtained were relevant [47]. Second, although the fairly small sample size used for the study could be considered a potential limitation, it was large enough to show differences in the parameters of interest. Finally, the study combined data from two countries: statistical comparison of baseline characteristics between the two subgroups was not performed and, in addition, it is possible that cultural differences may have affected responses to the questionnaire. Nevertheless, it has been shown that the relationship between income, education, or class identification and self-reported health is similar in Taiwan and South Korea [48], and we therefore believe that it is reasonable to combine the results for these two East Asian countries.
0.747 ⫾ 0.151 0.748 ⫾ 0.153 0.740 ⫾ 0.156 0.740 ⫾ 0.151 0.777 ⫾ 0.152
Conclusions This study elicited utility values associated with type 2 diabetes medication-related attributes from South Korean and Taiwanese populations. The results showed that treatment-related attributes, in particular dose frequency and nausea, would be expected to have a measurable effect on HRQOL, suggesting that they should be taken into consideration when selecting treatments for South Korean or Taiwanese patients with type 2 diabetes.
0.776 ⫾ 0.153
SG, standard gamble.
enrolled in the study, the results were consistent with other published data [13,29]. Increased utility associated with reduced dose frequency and disutility associated with nausea reflect the general fear associated with injections and nausea. Studies specifically in patients with diabetes have also found that nausea is associated with a reduction in utility [29] and that weekly injections are associated with increased utility compared with daily administration [13]. In the present study, which used a general population sample, weight loss was associated with an increase in utility; weight gain, however, did not have a detrimental effect on HRQOL. This is in contrast with studies involving patients with type 2 diabetes [24,29,39–43], which have reported that an increase in body weight is associated with a decrease in utility. It is recognized that health state values elicited from the general public can differ from those elicited from patients [28,44]. In this case, the
Acknowledgments We thank Dr. Katherine Croom and Caroline Spencer (Rx Communications, Mold, UK) for their medical writing assistance during the preparation of this manuscript. Source of financial support: The study was funded by Eli Lilly (Indianapolis, IN, USA). R EF E R EN C ES
[1] Chan JC, Malik V, Jia W, et al. Diabetes in Asia: epidemiology, risk factors, and pathophysiology. JAMA 2009;301:2129–40. [2] Kim YJ, Lim MN, Lee DS. Trend analysis in the prevalence of type 2 diabetes according to risk factors among Korean adults: based on the 2001–2009 Korean National Health and Nutrition Examination Survey data. J Korean Acad Nurs 2014;44:743–50. [3] Lin CC, Li CI, Hsiao CY, et al. Time trend analysis of the prevalence and incidence of diagnosed type 2 diabetes among adults in Taiwan from 2000 to 2007: a population-based study. BMC Public Health 2013;13:318.
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
[4] Kennedy-Martin T, Mitchell BD, Boye KS, et al. The health technology assessment environment in mainland China, Japan, South Korea and Taiwan—implications for the evaluation of diabetes mellitus therapies. Value Health Reg Issues 2014;3:108–16. [5] Choi YJ, Lee MS, An SY, et al. The relationship between diabetes mellitus and health-related quality of life in Korean adults: the Fourth Korea National Health and Nutrition Examination Survey (2007–2009). Diabetes Metab J 2011;35:587–94. [6] Solli O, Stavem K, Kristiansen IS. Health-related quality of life in diabetes: the associations of complications with EQ-5D scores. Health Qual Life Outcomes 2010;8:18. [7] Grandy S, Fox KM, SHIELD Study Group. Change in health status (EQ5D) over 5 years among individuals with and without type 2 diabetes mellitus in the SHIELD longitudinal study. Health Qual Life Outcomes 2012;10:99. [8] Zhang P, Brown MB, Bilik D, et al. Health utility scores for people with type 2 diabetes in U.S. managed care health plans: results from Translating Research Into Action for Diabetes (TRIAD). Diabetes Care 2012;35:2250–6. [9] Purnell TS, Joy S, Little E, et al. Patient preferences for noninsulin diabetes medications: a systematic review. Diabetes Care 2014;37:2055–62. [10] Bøgelund M, Vilsbøll T, Faber J, et al. Patient preferences for diabetes management among people with type 2 diabetes in Denmark—a discrete choice experiment. Curr Med Res Opin 2011;27:2175–83. [11] Jendle J, Torffvit O, Ridderstråle M, et al. Willingness to pay for health improvements associated with anti-diabetes treatments for people with type 2 diabetes. Curr Med Res Opin 2010;26:917–23. [12] Polonsky WH, Fisher L, Hessler D, et al. Patient perspectives on onceweekly medications for diabetes. Diabetes Obes Metab 2011;13:144–9. [13] Boye KS, Matza LS, Walter KN, et al. Utilities and disutilities for attributes of injectable treatments for type 2 diabetes. Eur J Health Econ 2011;12:219–30. [14] Hauber AB, Tunceli K, Yang JC, et al. A survey of patient preferences for oral antihyperglycemic therapy in patients with type 2 diabetes mellitus. Diabetes Ther 2015;6:75–84. [15] Vijan S, Hayward RA, Ronis DL, Hofer TP. Brief report: the burden of diabetes therapy: implications for the design of effective patientcentered treatment regimens. J Gen Intern Med 2005;20:479–82. [16] Evans M, Jensen HH, Bøgelund M, et al. Flexible insulin dosing improves health-related quality-of-life (HRQoL): a time trade-off survey. J Med Econ 2013;16:1357–65. [17] Heine RJ, Van Gaal LF, Johns D, et al. Exenatide versus insulin glargine in patients with suboptimally controlled type 2 diabetes: a randomized trial. Ann Intern Med 2005;143:559–69. [18] Deurenberg P, Yap M, van Staveren WA. Body mass index and percent body fat: a meta analysis among different ethnic groups. Int J Obes Relat Metab Disord 1998;22:1164–71. [19] Lear SA, Humphries KH, Kohli S, et al. Visceral adipose tissue accumulation differs according to ethnic background: results of the Multicultural Community Health Assessment Trial (M-CHAT). Am J Clin Nutr 2007;86:353–9. [20] Huxley R, James WP, Barzi F, et al. Ethnic comparisons of the crosssectional relationships between measures of body size with diabetes and hypertension. Obes Rev 2008;9(Suppl. 1):53–61. [21] Sach TH, Barton GR, Doherty M, et al. The relationship between body mass index and health-related quality of life: comparing the EQ-5D, EuroQol VAS and SF-6D. Int J Obes (Lond) 2007;31:189–96. [22] Wee HL, Cheung YB, Loke WC, et al. The association of body mass index with health-related quality of life: an exploratory study in a multiethnic Asian population. Value Health 2008;11(Suppl. 1):S105–14. [23] Søltoft F, Hammer M, Kragh N. The association of body mass index and health-related quality of life in the general population: data from the 2003 Health Survey of England. Qual Life Res 2009;18:1293–9. [24] Hunger M, Schunk M, Meisinger C, et al. Estimation of the relationship between body mass index and EQ-5D health utilities in individuals with type 2 diabetes: evidence from the population-based KORA studies. J Diabetes Complications 2012;26:413–8. [25] Wee HL, Li SC, Cheung YB, et al. The influence of ethnicity on healthrelated quality of life in diabetes mellitus: a population-based, multiethnic study. J Diabetes Complications 2006;20:170–8.
71
[26] Jhita T, Petrou S, Gumber A, et al. Ethnic differences in health related quality of life for patients with type 2 diabetes. Health Qual Life Outcomes 2014;12:83. [27] Health Insurance Review and Assessment Service. Guidelines for economic evaluation of pharmaceuticals in Korea, Seoul: Health Insurance Review and Assessment Service, 2005. Available from: 〈http: www.hira.or.kr〉. [28] Stamuli E. Health outcomes in economic evaluation: who should value health? Br Med Bull 2011;97:197–210. [29] Matza LS, Boye KS, Yurgin N, et al. Utilities and disutilities for type 2 diabetes treatment-related attributes. Qual Life Res 2007;16:1251–65. [30] Gao Y, Yoon KH, Chuang LM, et al. Efficacy and safety of exenatide in patients of Asian descent with type 2 diabetes inadequately controlled with metformin or metformin and a sulphonylurea. Diabetes Res Clin Pract 2009;83:69–76. [31] Ross PL, Littenberg B, Fearn P, et al. Paper standard gamble: a paperbased measure of standard gamble utility for current health. Int J Technol Assess Health Care 2003;19:135–47. [32] Littenberg B, Partilo S, Licata A, Kattan MW. Paper Standard Gamble: the reliability of a paper questionnaire to assess utility. Med Decis Making 2003;23:480–8. [33] Kim SH, Lee SI, Jo MW. Feasibility, comparability, and reliability of the standard gamble compared with the rating scale and time trade-off techniques in the EQ-5D-5 L valuation study. Value Health 2012;15: A650. [34] Drummond MF, Sculpher MJ, Torrance GW, et al. Methods for the Economic Evaluation of Health Care Programmes (3rd ed.) New York, NY: Oxford University Press, 2005. [35] Wang Y, Xie F, Kong MC, et al. Patient-reported health preferences of anticoagulant-related outcomes. J Thromb Thrombolysis 2015;40:268–73. [36] Torrance GW. Measurement of health state utilities for economic appraisal. J Health Econ 1986;5:1–30. [37] Rutten-van Mölken MP, Bakker CH, van Doorslaer EK, van der Linden S. Methodological issues of patient utility measurement: experience from two clinical trials. Med Care 1995;33:922–37. [38] Spencer A. The implications of linking questions within the SG and TTO methods. Health Econ 2004;13:807–18. [39] Redekop WK, Koopmanschap MA, Stolk RP, et al. Health-related quality of life and treatment satisfaction in Dutch patients with type 2 diabetes. Diabetes Care 2002;25:458–63. [40] Lee AJ, Morgan CL, Morrissey M, et al. Evaluation of the association between the EQ-5D (health-related utility) and body mass index (obesity) in hospital-treated people with type 1 diabetes, type 2 diabetes and with no diagnosed diabetes. Diabet Med 2005;22:1482–6. [41] Lane S, Levy AR, Mukherjee J, et al. The impact on utilities of differences in body weight among Canadian patients with type 2 diabetes. Curr Med Res Opin 2014;30:1267–73. [42] Dennett SL, Boye KS, Yurgin NR. The impact of body weight on patient utilities with or without type 2 diabetes: a review of the medical literature. Value Health 2008;11:478–86. [43] Lee WJ, Song KH, Noh JH, et al. Health-related quality of life using the EuroQol 5D questionnaire in Korean patients with type 2 diabetes. J Korean Med Sci 2012;27:255–60. [44] Peeters Y, Stiggelbout AM. Health state valuations of patients and the general public analytically compared: a meta-analytical comparison of patient and population health state utilities. Value Health 2010;13:306–9. [45] Thavorncharoensap M. Measurement of utility. J Med Assoc Thai 2014;97(Suppl. 5):S43–9. [46] Weinstein MC, Siegel JE, Gold MR, et al. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA 1996;276:1253–8. [47] McTaggart-Cowan H. Elicitation of informed general population health state utility values: a review of the literature. Value Health 2011;14:1153–7. [48] Hanibuchi T, Nakaya T, Murata C. Socio-economic status and self-rated health in East Asia: a comparison of China, Japan, South Korea and Taiwan. Eur J Public Health 2010;22:47–52.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/vhri
Utilities for Type 2 Diabetes Treatment-Related Attributes in a South Korean and Taiwanese Population Narayan Rajan, MA, MSc1,*, Kristina S. Boye, PhD, MS, MPH, RPh1, Meaghan Gibbs, BSc, MSc2, Yoon Ji Lee, BSc3, Peter Davey, BA, MA4, Mark Ball, BCom (Hons)4, Steve M. Babineaux, MA, MSc1 1 Eli Lilly, Indianapolis, IN, USA; 2Information Fulfillment Ltd., Hong Kong, China; 3Eli Lilly Korea, Seoul, Korea; 4PRIMA Consulting Group, Sydney, New South Wales, Australia
AB STR A CT
Objectives: To elicit utilities associated with type 2 diabetes medication-related attributes from South Korean and Taiwanese populations and to identify key drivers of preferences. Methods: Data from 59 respondents from the general population in South Korea and Taiwan were analyzed. Respondents’ preferences were elicited using a paper-based standard gamble questionnaire. Health states were designed to identify the utility or disutility of type 2 diabetes medication-related attributes, including dose frequency, nausea/vomiting (hereafter referred to as nausea), and weight change. Results: The mean utility for the basic health state (encompassing current body weight and no nausea) was 0.754 ⫾ 0.155 with weekly dose administration. Respondents showed a preference for weekly over daily administration (average increase in utility of 0.043 across all health states with weekly, vs. daily, administration). Nausea was associated with a decrease in utility (average decrease of 0.034 across all health states with, vs. without, nausea). Weight gain had
little effect on utility (average decrease of 0.000 and 0.001 across all health states with, vs. without, 3% and 5% gain, respectively), although weight loss was associated with a small increase in utility (average increase of 0.028 and 0.029 across all health states with, vs. without, 3% and 5% loss, respectively). Conclusions: Utilities associated with type 2 diabetes medication-related attributes were elicited from a general population sample from South Korea and Taiwan. Treatment-related attributes, in particular dose frequency and nausea, had a measurable effect on utility and should be considered when selecting treatment regimens for South Korean or Taiwanese patients with type 2 diabetes. Keywords: Asia, Korea, preference, Taiwan, treatment-related attributes, type 2 diabetes mellitus, utility.
Introduction
treatment preferences in type 2 diabetes, including glycemic control, weight control, the risk of hypoglycemia and gastrointestinal adverse effects, and the frequency of dose administration [9–13]. Dose frequency is one of the key attributes of antihyperglycemic treatments that affect patient preference [13,14]. An increasing frequency of administration is associated with a greater perceived treatment burden and reduced HRQOL [15,16]. A study of patients’ attitudes toward a once-weekly injectable glucose-lowering medication option found that perceived benefits included convenience, improved adherence, improved quality of life, and a reduced sense of treatment burden, whereas concerns included dosage consistency over time and potential forgetfulness [12]. In addition, many patients with type 2 diabetes are overweight or obese, and some antihyperglycemic medications are associated with changes in weight. For example, insulin can cause weight gain, whereas glucagon-like peptide-1 receptor agonists are usually associated with weight loss [17]. Compared with white populations, Asians have a higher percentage of body
The prevalence of type 2 diabetes mellitus in Asian populations is increasing rapidly, and it has been estimated that by 2025 the disease will affect approximately 180 million people in Asia [1]. Among East Asian countries, the prevalence increased from 7.7% to 11.8% in Korea between 2001 and 2009 [2], and from 5.8% in 2000 to 8.3% in 2007 in Taiwan [3]. The associated increase in disease burden will have an impact on health care costs and clinical outcomes in the region, and there will be an increasing need for studies that evaluate the pharmacoeconomics, including cost-utility, of treatments for diabetes [4]. In this context, there are several points that need to be considered, including the most relevant health-related quality-of-life (HRQOL) parameters to assess and the most appropriate methods of evaluation to use within the East Asian region. Diabetes and its complications are known to have an adverse influence on HRQOL [5–8], and consequently it is also important to understand the effect that treatments for diabetes have on patients’ well-being. Several factors are known to affect
Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc.
Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this article. * Address correspondence to: Narayan Rajan, Eli Lilly Australia Pty Ltd., 112 Wharf Road, West Ryde, New South Wales 2114, Australia. E-mail: [email protected] 2212-1099$36.00 – see front matter Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. http://dx.doi.org/10.1016/j.vhri.2015.11.006
68
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
fat [18] and visceral fat [19] for any given body mass index, and are at risk of developing diabetes at a lower body mass index [20]. It is known that body weight can affect quality of life [21–24], and also that ethnicity can influence HRQOL in patients with diabetes [25,26]; it is, however, not clear whether the differences in percentage body fat and associated health risks seen between Asian and white populations result in different utility values for these populations. Consequently, it would be appropriate to include dose frequency and weight change in utility studies evaluating preferences related to diabetes medications. Various methods can be used to elicit utilities for subsequent use in cost-utility analyses, and the preferred method can vary between different countries and regions. The Health Insurance Review and Assessment Service of South Korea guidelines on economic evaluations indicate a preference for administering a multiattribute utility instrument to a patient population or administering a direct method, such as the standard gamble (SG), to a general population sample [27]. There is no consensus among experts as to whether it is preferable for valuations to be made by patients or by the general public [28]. Among arguments in favor of deriving utilities from the general population for hypothetical health states, there are two key points. First, the general public will tend to value health states in an unbiased manner, because they have no vested interest in a particular disease or health state [28]. Second, in publicly funded health care systems, society bears the costs of health care decisions, and therefore the general public’s perspective may be the most relevant [28]. Given the increasing importance of pharmacoeconomic evaluations within East Asia, it is important to understand the implications of adapting studies conducted elsewhere to specific Asian settings and to identify values that are appropriate for calculating cost-utility ratios in countries within this region. The aims of the present study were to elicit utility values associated with type 2 diabetes medication-related attributes from South Korean and Taiwanese populations and to identify the key drivers of preferences within these populations.
Methods Survey Population Study participants came from the general population in South Korea and Taiwan. People who were diagnosed with type 2 diabetes were excluded to avoid potentially perceived bias due to disease and treatment experience. The sample included both men and women and incorporated a mix of age and income groups. All participants were older than 18 years and were able to read a survey in their local language and complete a relatively complicated questionnaire. People with health-related occupations were excluded. All participants provided written informed consent. Ethics approval was not required for the study. The study sample was recruited from random households in the areas of Seoul and Taipei. The sample was then balanced to the South Korean and Taiwanese populations according to local demographic profiles. Participants were contacted via phone and followed up by post through a letter introducing the survey.
Selection of Health States and Utility Instruments Health states for the present study were derived from two studies performed in the United Kingdom by Matza et al. [29] and Boye et al. [13]. The basic health state represented a patient with type 2 diabetes with glucose levels under control and with no complications (at current weight and without nausea/vomiting [hereafter referred to as nausea]). Other health states were designed to identify the utility or disutility of type 2 diabetes medication-
Table 1 – Attributes included in the health state descriptions. Basic health state Dose frequency Basic health state plus one of the following: In addition to taking oral medication (pills or tablets), you give yourself injections every day (once or twice per day) In addition to taking oral medication (pills or tablets), you give yourself injections once a week Nausea/vomiting Basic health state plus one of the following: You sometimes experience nausea/vomiting You do not experience nausea/vomiting Weight Basic health state plus one of the following: You weigh ___ kg (which is 5% more than you weigh now) You weigh ___ kg (which is 3% more than you weigh now) You weigh ___ kg (which is 3% less than you weigh now) You weigh ___ kg (which is 5% less than you weigh now) Current health state Worst health state You have had type 2 diabetes for several years You take oral medications (pills or tablets) Your blood sugar levels are not in control (sometimes resulting in excessive thirst, frequent urination, fatigue, irritability, and/or blurry vision) You have tingling or prickling sensations in your hands or feet You have shortness of breath during physical activity You have pain in your legs In addition to taking oral medications (pills or tablets), you give yourself injections every day (once or twice per day) You sometimes experience nausea/vomiting You weigh ___ kg (which is 5% more than you weigh now)
related attributes, including weight change and nausea, and were adapted to include attributes related to dose frequency for the present study. The attributes included in the health states for this study are presented in Table 1. Given that the participants were from the general population, weight change was based on the average baseline weight of patients in a large clinical trial that evaluated exenatide in Asian patients (69 kg) rather than on a participant’s own weight [30]. The utility instrument selected for use in the study was the SG. The paper-based SG questionnaire is a reliable means of measuring respondents’ preferences [31,32] and has been shown to be feasible and reliable for use in a Korean population [33].
Utility Sessions Participants attended one of six utility sessions (three in Taipei, Taiwan, and three in Seoul, South Korea), all of which delivered the same presentation in the local language of that country. Each session involved a maximum of 12 participants. Overseers monitored the sessions to assist with questions and verify responders’ logic. All participants completed a form covering demographic information. At the start of the session, participants were given a presentation on type 2 diabetes that contained information taken from published literature and did not reference any specific treatment therapy. This was followed by a 10-question comprehension test to confirm they had understood the information. Participants were familiarized with the various health states by completing a visual analogue scale under the supervision of a trained interviewer; each health state was rated on a preference assessment
69
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
rating scale ranging from 10 (best possible health state) to 0 (worst possible health state). Participants then completed a paper-based SG questionnaire for each health state. To help ensure consistency and reliability, the same lead investigator conducted all utility sessions in South Korea and Taiwan. A local support team that was present for all sessions in that country assisted the lead investigator, and all the members of the support team received the same training from the lead investigator before the utility sessions.
Scaling Methods The SG questionnaire elicits patients’ preferences for health outcomes by determining how much risk they would be willing to take to avoid remaining in a poor health state [34,35]. The paper-based questionnaire presented the worst health state to the respondent, who was offered two alternatives. Alternative 1 was a treatment with two possible outcomes: either the respondent was returned to normal health and lived for an additional T years (probability P) or the patient died immediately (probability 1 P). Alternative 2 had a certain outcome of chronic state i for life (T years). The probability P was varied until the respondent was indifferent between the two alternatives, at which point the required preference value for health state i was simply P [36]. The probabilities were varied using a ping-pong method. When using immediate death as the treatment-failure option, a possible problem is that respondents may not be willing to accept any chance of treatment failure when non–life-threatening or temporary states of poor health are being valued. In this situation, it may appear that non–life-threatening and temporary health states are valued as highly as full health, simply because the basic reference SG questionnaire is not sensitive enough to capture the true underlying preferences. To overcome this problem, health states were indirectly linked to death, a technique referred to as “chaining” [37,38]. When valuing the remaining, less severe, health states, a nonfatal health outcome (the worst health state, i.e., the most severe nonfatal state) was used instead of immediate death as the treatment-failure outcome. Participant ranking of all heath states confirmed the worst health state.
Statistical Methods The primary results of this study (i.e., mean utility values) were descriptive in nature and therefore there was no key statistical comparison involving significance testing on which a sample size should be based. Hence, a power analysis was not applicable for determining the target sample size of this study. Descriptive statistics were used to summarize utilities in terms of mean and SD. Respondents who failed the logic test (i.e., gave four or more illogical responses) or who were unable to complete the survey were excluded from the calculation. An SG chaining method was used to detect differences associated with weight changes, nausea, and dose frequency. The chained SG was derived by calculating the raw scores. Participants rated the worst health state in comparison to death, thereby allowing the raw score of less severe health states to be adjusted using the formula r (1 v) þ v, where r was the raw score of the worst health state and v was the raw score of the valued health state. Results were presented descriptively. Disutility or increased utility was calculated as the difference between health state values with or without specific attributes. An example of the calculation of the disutility associated with nausea would be as follows: value for basic health state at current weight with nausea (weekly administration) minus value for basic health state at current weight without nausea (weekly administration). Average values for disutility or increased utility were the average
Table 2 – Populations’ demographic characteristics. Characteristic
Sex: male/female (n) Mean age (y) Marital status (%) Single Married Divorced Widowed Highest educational level (%) Did not complete high school High school diploma Degree/equivalent qualification Master’s degree/doctorate
General population (N ¼ 59) 28/31 43 39 58 2 1 0 12 75 13
differences across all health states with, versus without, the attribute of interest.
Results The study recruited 67 participants from the general population of South Korea (n ¼ 32) and Taiwan (n ¼ 35). Participants who did not complete the questionnaire (n ¼ 5) or failed the logic or comprehension test (n ¼ 3) were excluded. The demographic characteristics of the final study population (n ¼ 59) are summarized in Table 2. The mean age of the group was 43 years, just over half were married (58%), and most had received tertiarylevel education or higher. The mean time taken to complete the survey was 1.5 hours. Utilities derived using the SG in general population respondents from South Korea and Taiwan who completed the study (n ¼ 59) are presented in Table 3. The mean utility associated with the basic health state with weekly dose administration was 0.754 ⫾ 0.155. Respondents indicated that nausea was associated with disutility (average decrease of 0.034 in utility across all health states with, vs. without, nausea). They showed a preference for weekly dose administration, with an average increase in utility, compared with daily administration, of 0.043 (average across all health states with weekly, vs. daily, administration). Respondents had no clear preference with respect to weight gain (average change in utility of 0.000 for a 3% weight gain and 0.001 for a 5% weight gain), although weight loss appeared to be associated with increased utility (average increase of 0.028 for a 3% weight loss and 0.029 for a 5% weight loss).
Discussion Localized utility studies provide insight into geographical and cultural differences in preferences related to health states, and can be designed to use the most appropriate methodology for a particular country or region. The present study used a method in keeping with the recommendations of the Health Insurance Review and Assessment Service of South Korea guidelines on economic evaluations [27]. The study obtained local utility values associated with key attributes related to type 2 diabetes treatments in South Korean and Taiwanese people, using the SG method and a general population sample. The results showed that treatment-related attributes, in particular dose frequency and nausea, would be expected to have a measurable effect on patients’ HRQOL. Although additional research may be needed to further generalize these findings, because of the limited number of subjects
70
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
Table 3 – Utility values derived using the SG method in the general population (N ¼ 59). Health state
Daily dose administration Basic health state (current weight) þ no nausea Basic health state (current weight) þ nausea Basic health state þ 3% higher weight þ nausea Basic health state þ 5% higher weight þ nausea Basic health state þ 3% higher weight þ no nausea Basic health state þ 5% higher weight þ no nausea Basic health state þ 3% lower weight þ nausea Basic health state þ 5% lower weight þ nausea Basic health state þ 3% lower weight þ no nausea Basic health state þ 5% lower weight þ no nausea Weekly dose administration Basic health state (current weight) þ no nausea Basic health state (current weight) þ nausea Basic health state þ 3% higher weight þ nausea Basic health state þ 5% higher weight þ nausea Basic health state þ 3% higher weight þ no nausea Basic health state þ 5% higher weight þ no nausea Basic health state þ 3% lower weight þ nausea Basic health state þ 5% lower weight þ nausea Basic health state þ 3% lower weight þ no nausea Basic health state þ 5% lower weight þ no nausea
Utility value (mean ⫾ SD) 0.705 ⫾ 0.165 0.663 ⫾ 0.193 0.676 ⫾ 0.170 0.674 ⫾ 0.173 0.707 ⫾ 0.164 0.703 ⫾ 0.160 0.700 ⫾ 0.162 0.704 ⫾ 0.161 0.736 ⫾ 0.151 0.737 ⫾ 0.155
0.754 ⫾ 0.155 0.721 ⫾ 0.160 0.717 ⫾ 0.165 0.711 ⫾ 0.160
difference in findings might reflect a greater awareness among patients with type 2 diabetes of the dangers of weight gain and the weight-related problems associated with some treatments for diabetes. It is notable that utility values obtained from respondents in our study for changes in weight were similar whether the change was 3% or 5%. The study had several potential limitations. The first limitation was the use of a general population sample to derive utilities. The general public lack experience of the disease under consideration and may overemphasize negative aspects of health states [28], whereas patients with type 2 diabetes have knowledge and experience of the disorder and associated treatments. However, there is a risk that patients could provide biased judgments because they have vested interests, or alternatively may assign higher values to health states because of adaptation to their condition or lowered expectations, whereas the general public is more likely to provide an unbiased judgment [28]. There is no consensus as to which population is the most appropriate to use, and both approaches are valid [28]. The present study reported utility values elicited from the general population using the SG method, an approach acceptable to many health care payers [27,45,46]. Respondents were informed about type 2 diabetes, and their understanding was tested, to help ensure that the utilities obtained were relevant [47]. Second, although the fairly small sample size used for the study could be considered a potential limitation, it was large enough to show differences in the parameters of interest. Finally, the study combined data from two countries: statistical comparison of baseline characteristics between the two subgroups was not performed and, in addition, it is possible that cultural differences may have affected responses to the questionnaire. Nevertheless, it has been shown that the relationship between income, education, or class identification and self-reported health is similar in Taiwan and South Korea [48], and we therefore believe that it is reasonable to combine the results for these two East Asian countries.
0.747 ⫾ 0.151 0.748 ⫾ 0.153 0.740 ⫾ 0.156 0.740 ⫾ 0.151 0.777 ⫾ 0.152
Conclusions This study elicited utility values associated with type 2 diabetes medication-related attributes from South Korean and Taiwanese populations. The results showed that treatment-related attributes, in particular dose frequency and nausea, would be expected to have a measurable effect on HRQOL, suggesting that they should be taken into consideration when selecting treatments for South Korean or Taiwanese patients with type 2 diabetes.
0.776 ⫾ 0.153
SG, standard gamble.
enrolled in the study, the results were consistent with other published data [13,29]. Increased utility associated with reduced dose frequency and disutility associated with nausea reflect the general fear associated with injections and nausea. Studies specifically in patients with diabetes have also found that nausea is associated with a reduction in utility [29] and that weekly injections are associated with increased utility compared with daily administration [13]. In the present study, which used a general population sample, weight loss was associated with an increase in utility; weight gain, however, did not have a detrimental effect on HRQOL. This is in contrast with studies involving patients with type 2 diabetes [24,29,39–43], which have reported that an increase in body weight is associated with a decrease in utility. It is recognized that health state values elicited from the general public can differ from those elicited from patients [28,44]. In this case, the
Acknowledgments We thank Dr. Katherine Croom and Caroline Spencer (Rx Communications, Mold, UK) for their medical writing assistance during the preparation of this manuscript. Source of financial support: The study was funded by Eli Lilly (Indianapolis, IN, USA). R EF E R EN C ES
[1] Chan JC, Malik V, Jia W, et al. Diabetes in Asia: epidemiology, risk factors, and pathophysiology. JAMA 2009;301:2129–40. [2] Kim YJ, Lim MN, Lee DS. Trend analysis in the prevalence of type 2 diabetes according to risk factors among Korean adults: based on the 2001–2009 Korean National Health and Nutrition Examination Survey data. J Korean Acad Nurs 2014;44:743–50. [3] Lin CC, Li CI, Hsiao CY, et al. Time trend analysis of the prevalence and incidence of diagnosed type 2 diabetes among adults in Taiwan from 2000 to 2007: a population-based study. BMC Public Health 2013;13:318.
VALUE IN HEALTH REGIONAL ISSUES 9C (2016) 67–71
[4] Kennedy-Martin T, Mitchell BD, Boye KS, et al. The health technology assessment environment in mainland China, Japan, South Korea and Taiwan—implications for the evaluation of diabetes mellitus therapies. Value Health Reg Issues 2014;3:108–16. [5] Choi YJ, Lee MS, An SY, et al. The relationship between diabetes mellitus and health-related quality of life in Korean adults: the Fourth Korea National Health and Nutrition Examination Survey (2007–2009). Diabetes Metab J 2011;35:587–94. [6] Solli O, Stavem K, Kristiansen IS. Health-related quality of life in diabetes: the associations of complications with EQ-5D scores. Health Qual Life Outcomes 2010;8:18. [7] Grandy S, Fox KM, SHIELD Study Group. Change in health status (EQ5D) over 5 years among individuals with and without type 2 diabetes mellitus in the SHIELD longitudinal study. Health Qual Life Outcomes 2012;10:99. [8] Zhang P, Brown MB, Bilik D, et al. Health utility scores for people with type 2 diabetes in U.S. managed care health plans: results from Translating Research Into Action for Diabetes (TRIAD). Diabetes Care 2012;35:2250–6. [9] Purnell TS, Joy S, Little E, et al. Patient preferences for noninsulin diabetes medications: a systematic review. Diabetes Care 2014;37:2055–62. [10] Bøgelund M, Vilsbøll T, Faber J, et al. Patient preferences for diabetes management among people with type 2 diabetes in Denmark—a discrete choice experiment. Curr Med Res Opin 2011;27:2175–83. [11] Jendle J, Torffvit O, Ridderstråle M, et al. Willingness to pay for health improvements associated with anti-diabetes treatments for people with type 2 diabetes. Curr Med Res Opin 2010;26:917–23. [12] Polonsky WH, Fisher L, Hessler D, et al. Patient perspectives on onceweekly medications for diabetes. Diabetes Obes Metab 2011;13:144–9. [13] Boye KS, Matza LS, Walter KN, et al. Utilities and disutilities for attributes of injectable treatments for type 2 diabetes. Eur J Health Econ 2011;12:219–30. [14] Hauber AB, Tunceli K, Yang JC, et al. A survey of patient preferences for oral antihyperglycemic therapy in patients with type 2 diabetes mellitus. Diabetes Ther 2015;6:75–84. [15] Vijan S, Hayward RA, Ronis DL, Hofer TP. Brief report: the burden of diabetes therapy: implications for the design of effective patientcentered treatment regimens. J Gen Intern Med 2005;20:479–82. [16] Evans M, Jensen HH, Bøgelund M, et al. Flexible insulin dosing improves health-related quality-of-life (HRQoL): a time trade-off survey. J Med Econ 2013;16:1357–65. [17] Heine RJ, Van Gaal LF, Johns D, et al. Exenatide versus insulin glargine in patients with suboptimally controlled type 2 diabetes: a randomized trial. Ann Intern Med 2005;143:559–69. [18] Deurenberg P, Yap M, van Staveren WA. Body mass index and percent body fat: a meta analysis among different ethnic groups. Int J Obes Relat Metab Disord 1998;22:1164–71. [19] Lear SA, Humphries KH, Kohli S, et al. Visceral adipose tissue accumulation differs according to ethnic background: results of the Multicultural Community Health Assessment Trial (M-CHAT). Am J Clin Nutr 2007;86:353–9. [20] Huxley R, James WP, Barzi F, et al. Ethnic comparisons of the crosssectional relationships between measures of body size with diabetes and hypertension. Obes Rev 2008;9(Suppl. 1):53–61. [21] Sach TH, Barton GR, Doherty M, et al. The relationship between body mass index and health-related quality of life: comparing the EQ-5D, EuroQol VAS and SF-6D. Int J Obes (Lond) 2007;31:189–96. [22] Wee HL, Cheung YB, Loke WC, et al. The association of body mass index with health-related quality of life: an exploratory study in a multiethnic Asian population. Value Health 2008;11(Suppl. 1):S105–14. [23] Søltoft F, Hammer M, Kragh N. The association of body mass index and health-related quality of life in the general population: data from the 2003 Health Survey of England. Qual Life Res 2009;18:1293–9. [24] Hunger M, Schunk M, Meisinger C, et al. Estimation of the relationship between body mass index and EQ-5D health utilities in individuals with type 2 diabetes: evidence from the population-based KORA studies. J Diabetes Complications 2012;26:413–8. [25] Wee HL, Li SC, Cheung YB, et al. The influence of ethnicity on healthrelated quality of life in diabetes mellitus: a population-based, multiethnic study. J Diabetes Complications 2006;20:170–8.
71
[26] Jhita T, Petrou S, Gumber A, et al. Ethnic differences in health related quality of life for patients with type 2 diabetes. Health Qual Life Outcomes 2014;12:83. [27] Health Insurance Review and Assessment Service. Guidelines for economic evaluation of pharmaceuticals in Korea, Seoul: Health Insurance Review and Assessment Service, 2005. Available from: 〈http: www.hira.or.kr〉. [28] Stamuli E. Health outcomes in economic evaluation: who should value health? Br Med Bull 2011;97:197–210. [29] Matza LS, Boye KS, Yurgin N, et al. Utilities and disutilities for type 2 diabetes treatment-related attributes. Qual Life Res 2007;16:1251–65. [30] Gao Y, Yoon KH, Chuang LM, et al. Efficacy and safety of exenatide in patients of Asian descent with type 2 diabetes inadequately controlled with metformin or metformin and a sulphonylurea. Diabetes Res Clin Pract 2009;83:69–76. [31] Ross PL, Littenberg B, Fearn P, et al. Paper standard gamble: a paperbased measure of standard gamble utility for current health. Int J Technol Assess Health Care 2003;19:135–47. [32] Littenberg B, Partilo S, Licata A, Kattan MW. Paper Standard Gamble: the reliability of a paper questionnaire to assess utility. Med Decis Making 2003;23:480–8. [33] Kim SH, Lee SI, Jo MW. Feasibility, comparability, and reliability of the standard gamble compared with the rating scale and time trade-off techniques in the EQ-5D-5 L valuation study. Value Health 2012;15: A650. [34] Drummond MF, Sculpher MJ, Torrance GW, et al. Methods for the Economic Evaluation of Health Care Programmes (3rd ed.) New York, NY: Oxford University Press, 2005. [35] Wang Y, Xie F, Kong MC, et al. Patient-reported health preferences of anticoagulant-related outcomes. J Thromb Thrombolysis 2015;40:268–73. [36] Torrance GW. Measurement of health state utilities for economic appraisal. J Health Econ 1986;5:1–30. [37] Rutten-van Mölken MP, Bakker CH, van Doorslaer EK, van der Linden S. Methodological issues of patient utility measurement: experience from two clinical trials. Med Care 1995;33:922–37. [38] Spencer A. The implications of linking questions within the SG and TTO methods. Health Econ 2004;13:807–18. [39] Redekop WK, Koopmanschap MA, Stolk RP, et al. Health-related quality of life and treatment satisfaction in Dutch patients with type 2 diabetes. Diabetes Care 2002;25:458–63. [40] Lee AJ, Morgan CL, Morrissey M, et al. Evaluation of the association between the EQ-5D (health-related utility) and body mass index (obesity) in hospital-treated people with type 1 diabetes, type 2 diabetes and with no diagnosed diabetes. Diabet Med 2005;22:1482–6. [41] Lane S, Levy AR, Mukherjee J, et al. The impact on utilities of differences in body weight among Canadian patients with type 2 diabetes. Curr Med Res Opin 2014;30:1267–73. [42] Dennett SL, Boye KS, Yurgin NR. The impact of body weight on patient utilities with or without type 2 diabetes: a review of the medical literature. Value Health 2008;11:478–86. [43] Lee WJ, Song KH, Noh JH, et al. Health-related quality of life using the EuroQol 5D questionnaire in Korean patients with type 2 diabetes. J Korean Med Sci 2012;27:255–60. [44] Peeters Y, Stiggelbout AM. Health state valuations of patients and the general public analytically compared: a meta-analytical comparison of patient and population health state utilities. Value Health 2010;13:306–9. [45] Thavorncharoensap M. Measurement of utility. J Med Assoc Thai 2014;97(Suppl. 5):S43–9. [46] Weinstein MC, Siegel JE, Gold MR, et al. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA 1996;276:1253–8. [47] McTaggart-Cowan H. Elicitation of informed general population health state utility values: a review of the literature. Value Health 2011;14:1153–7. [48] Hanibuchi T, Nakaya T, Murata C. Socio-economic status and self-rated health in East Asia: a comparison of China, Japan, South Korea and Taiwan. Eur J Public Health 2010;22:47–52.