The international generalisability of evidence for health policy: A cross country comparison of medication adherence following policy change

- How generalisable is evidence for health policy making across health systems, in particular evidence for prescription copayments?
- We addressed this question by studying two similar copayment policies in two health systems. We assessed the impact of each policy on adherence.
- These two similar copayment policies did not invoke similar responses in adherence to medications.
- Nuanced differences between the health systems and the patients within them may affect differences in impact of policies.
- Before applying evidence from one health system to another, critical questions about local applicability are key to maximising its' utility.

Copayments for prescriptions may increase morbidity and mortality via reductions in adherence to medications. Relevant data can inform policy to minimise such unintended effects. We explored the generalisability of evidence for copayments by comparing two international copayment polices, one in Massachusetts and one in Ireland, to assess whether effects on medication adherence were comparable. We used national prescription data for public health insurance programmes in Ireland and Medicaid data in the U.S. New users of oral anti-hypertensive, anti-hyperlipidaemic and diabetic drugs were included (total n = 14,259 in U.S. and n = 43,843 in Ireland). We examined changes in adherence in intervention and comparator groups in each setting using segmented linear regression with generalised estimating equations.In Massachusetts, a gradual decrease in adherence to anti-hypertensive medications of −1% per month following the policy occurred. In contrast, the response in Ireland was confined to a −2.9% decrease in adherence immediately following the policy, with no further decrease over the 8 month follow-up. Reductions in adherence to oral diabetes drugs were larger in the U.S. group in comparison to the Irish group. No difference in adherence changes between the two settings for anti-hyperlipidaemic drugs occurred.Evidence on cost-sharing for prescription medicines is not 'one size fits all'. Time since policy implementation and structural differences between health systems may influence the differential impact of copayment policies in international settings.