Aim: To review economic evaluations of weight loss drugs and compare reported incremental cost-effectiveness ratios (ICERs).
Methods: A literature search was conducted for cost-effectiveness (CEAs) and cost-utility analyses (CUAs) of sibutramine, orlistat and rimonabant.
Results: Fourteen unique articles were identified (11 CUAs and 3 CEAs; 9 orlistat, 4 sibutramine and 1 rimonabant). All used diet and exercise as comparator, whereas none included indirect costs. Time horizons varied from treatment period only (1-4 years) to 80 years (median 7.5 years). Longer studies modeled effects on diabetes, micro- and macrovascular complications, coronary heart disease and death. Of the CUAs, the median ICER was 16,000 euro(2007)/QALY (quality-adjusted life-year; range 10,000-88,000), with the worst cost-effectiveness when recommended stop rules for non-responding patients were not applied. All studies but three were funded by the manufacturing company, and the median ICER was considerably higher for independent than for sponsored analyses (62,000 euro vs 15,000 euro/QALY). However, two of the three independent CUAs did not use recommended stop rules, as compared with one of eight manufacturer-sponsored analyses. The results were most sensitive to assumptions regarding weight loss sustainability and utility per kilogram lost. Side effects and dropout because of reasons other than lack of efficacy were generally not incorporated.
Conclusion: Published economic evaluations indicate that orlistat, sibutramine and rimonabant are within the range of what is generally regarded as cost-effective. Uncertainty remains about weight loss sustainability, utility gain associated with weight loss and extrapolations from transient weight loss to long-term health benefits. Modeling of head-to-head comparisons and attrition is needed, as are analyses conducted independently of manufacturing companies.