The impact of cost-sharing schemes on drug compliance in Italy: evidence based on quantile regression

Objectives In this article we investigate the causal effect of cost-sharing schemes on compliance with statins in a quantile regression framework. Methods We use the health search CSD-LPD data, a longitudinal observational dataset containing computerbased patient records collected by Italian general practitioners. We exploit a series of natural experiments referring to several introductions of co-payment schemes in some of the Italian regions between 2000 and 2009. We adopt an extended difference-in-differences approach to provide quantile estimates of the impact of co-payments on compliance. Results We find that (i) introduction of co-payments hurts residents of regions with worse quality and provision of health care; (ii) within these regions, co-payments were particularly harmful for high compliers; (iii) gender, clinical history and geographic residence are important determinants of compliance among poor compliers; (iv) compliance decreases with the potency and dosage of statins, particularly for poor compliers. Conclusions In the presence of inefficient health-care provision, co-payments are harmful for drug compliance, and this is especially true for patients who are originally good compliers.