Cost-effectiveness of integrated care in frail elderly using the ICECAP-O and EQ-5D: does choice of instrument matter?

The European journal of health economics : HEPAC : health economics in prevention and care

PubMedID: 24760405

Makai P, Looman W, Adang E, Melis R, Stolk E, Fabbricotti I. Cost-effectiveness of integrated care in frail elderly using the ICECAP-O and EQ-5D: does choice of instrument matter?. Eur J Health Econ. 2014;.
Economic evaluations likely undervalue the benefits of interventions in populations receiving both health and social services, such as frail elderly, by measuring only health-related quality of life. For this reason, alternative preference-based instruments have been developed for economic evaluations in the elderly, such as the ICECAP-O. The aim of this paper is twofold: (1) to evaluate the cost-effectiveness using a short run time frame for an integrated care model for frail elderly, and (2) to investigate whether using a broader measure of (capability) wellbeing in an economic evaluation leads to a different outcome in terms of cost-effectiveness. We performed univariate and multivariate analyses on costs and outcomes separately. We also performed incremental net monetary benefit regressions using quality adjusted life years (QALYs) based on the ICECAP-O and EQ-5D. In terms of QALYs as measured with the EQ-5D and the ICECAP-O, there were small and insignificant differences between the instruments, due to negligible effect size. Therefore, widespread implementation of the Walcheren integrated care model would be premature based on these results. All results suggest that, using the ICECAP-O, the intervention has a higher probability of cost-effectiveness than with the EQ-5D at the same level of WTP. In case an intervention's health and wellbeing effects are not significant, as in this study, using the ICECAP-O will not lead to a false claim of cost-effectiveness of the intervention. On the other hand, if differences in capability QALYs are meaningful and significant, the ICECAP-O may have the potential to measure broader outcomes and be more sensitive to differences between intervention and comparators.