Comparison of measures of health-related quality of life in patients with alcohol misuse: mapping of two alcohol-specific instruments to three preference-based instruments

Problems associated with alcohol use disorder (AUD) can be quantified using generic and specific measures. The aim of this study is to compare different preference-based instruments for measuring health-related quality of life (HRQoL) in patients with AUD and to examine their relationship with alcohol-specific measures used in clinical practice. Patients with AUD were recruited from a Spanish alcoholism unit. We administered the EuroQoL-5Dimension-5levels (EQ-5D), ShortForm-6Dimension (SF-6D), AlcoholQuality-of-life-4Dimension (AlcQ-4D), Diagnostic and Statistical Manual of Mental Disorders (DSM-5) and Alcohol Use Disorders Identification Test (AUDIT) instruments at baseline and 12 months later. Differences in HRQoL instrument scores were estimated and their ability to discriminate between known clinical severity groups was analysed. Several mapping functions were tested to transform clinical scores (AUDIT or DSM-5) into HRQoL scores (EQ-5D, SF-6D or AlcQ-4D). The results show that HRQoL scores are sensitive to the instrument used. Mean utility scores are always highest with EQ-5D, followed by SF-6D and AlcQ-4D. All HRQoL instruments discriminate between clinical severity groups defined by DSM-5 or AUDIT. Although several mapping functions were estimated for each pair of clinical vs preference-based instruments, the model using the total score of the clinical measures as the independent variable was selected for all of them. The results suggest that clinical measures used in the field of AUD could be adapted for use in economic evaluation. However, the incremental cost-utility ratio of AUD programmes, and hence the policy decisions derived from it, may depend on the HRQoL instrument used.