Exploring Patterns of Differences between EQ-5d and SF-6d: A Systematic Review
Year: 2007
Type: preprint
Study design: Systematic Review
Abstract: Objective Previous studies have found that the EQ-5D and SF-6D provide different estimates of the same person's level of utility. Moreover, it has been suggested that the observed differences between EQ-5D and SF-6D utilities follow a pattern: healthier states tend to have a higher score according to the EQ-5D, whereas poorer health states tend to have a higher score according to the SF-6D (Sach et al, 2006). The objective of this study was to further investigate this phenomenon. We performed a systematic review of the literature on previous studies regarding EQ-5D and SF-6D utilities. Methods A literature search was performed in Medline, Pubmed, Embase, CINAHL, and Cochrane Database. The following key words were used: (EQ-5D or Euro-Qol or EQ5D) and (SF-6D or SF6D). Studies were included if mean EQ5D utility, and SF36 derived SF6D utility, based on a person's own health state description, were presented. References of selected articles were checked. Results In total 37 studies were identified. Two studies did not present empirical data, and in three studies hypothetical health state descriptions were used. Nine studies were excluded because no mean EQ-5D and SF36 derived SF-6D utilities were reported. Four references were a double citation of the same study. The 19 articles that met the inclusion criteria contained data of 43 populations, ranging from the general public to patients with very severe multiple sclerosis. Overall, EQ-5D utilities ranged from 0.10 to 0.93, while the SF-6D utility ranged from 0.58 to 0.93. For health states with EQ-5D utilities below 0.64, the SF-6D utility was equal (1 population) or higher (17 populations). The difference ranged from 0.00 in severe rheumatoid arthritis to 0.48 in very severe multiple sclerosis. For health states with a EQ-5D utility of 0.64 or higher (25 populations), in 14 populations the SF-6D utility was lower than the EQ-5D utility. The difference ranged from 0.01 (asthma) to 0.19 (moderate multiple sclerosis). In 11 populations with a EQ-5D health state of 0.64 or higher, the SF-6D utility was higher (range from 0.01 (irritable bowel syndrome) to 0.16 (outpatients)). Conclusions The suggestion that healthier states tend to have a higher utility according to the EQ-5D, whereas poorer health states tend to have a higher utility according to the SF-6D is partly confirmed by this review. For moderate and poor health states (0.64 on EQ-5D) no clear systematical difference was found. Our results indicate that SF-6D utilities may overestimate the utility of moderate to poor health states. An important next step is to investigate whether this phenomenon translates into larger utility gain for interventions for patients in poor health according to the EQ-5D. For now, researchers and policy makers should bear in mind that the comparison of health state utilities derived with the EQ-5D and SF-6D is systematically biased.
Source: SSRN Electronic Journal
Authors: Manuela A. Joore, Daniëlle Brunenberg
Institution: Maastricht University Medical Centre
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W2265689346
