OBJECTIVES

: The objective of this meta-analysis was to synthesize the effect of value based insurance designs (VBIDs) on adherence in patients with chronic diseases. Furthermore, the associations of effects with implementation periods, concurrent disease management programs (DMPs), and medicines were analyzed.

METHODS

: A systematic review of the literature is done on Business Source Complete, EconLit, PsycINFO, Medline, Scopus, Web of Science Core Collection, Cochrane Library, and ClinicalTrials.gov. Interventional studies, which evaluated the effect of reduced cost sharing on proportion of days covered (PDCs) were included. For synthesis of effects, random effects meta-analyses of continuous were used. Between population difference was estimated based on the DerSimonian-Laird method. Risk of bias was assessed by the risk of bias tool of the Cochrane Collaboration for controlled studies.

RESULTS

: After screening 2421 abstracts, 17 studies (2 RCT, 1 interrupted time series analysis, 14 retrospective cohort studies), with 22 independent populations and 73 subgroup analyses were included. Overall risk of bias was low in most risk dimensions. VBIDs significantly increased PDCs in the indication areas diabetes, chronic heart diseases and asthma within up to 1 year. Overall, the effect on adherence was highest in diabetes (5.94, 3.03-8.85), followed by chronic heart diseases (3.85, 3.01-4.7), and asthma (1.9, 0.63-3.27). Within heart diseases, the VBIDs’ effect on adherence to lipid-lowering medications was higher than to antihypertensives (5.00 vs. 3.44). Longer implementation periods were significantly associated with lower effects (p<0.0001). The effects did not significantly differ between VBIDs with and without additional DMP (p=0.92).

CONCLUSIONS

: In the included studies VBIDs increased adherence. This effect decreased over time and differed by indication area and, within heart diseases, by drug class. No difference was found between VBIDs with and without concurrent DMP. This study is the first to provide explanations for differences between effects in VBID, which has implications for health insurance and policy.