Accounting for noncompliance in pharmacoeconomic evaluations

Noncompliance with prescribed drug regimens is a widespread phenomenon which results in decreased efficacy and is often associated with increased medical expenditures. Despite this, economic evaluations based on decision-analytic models rarely incorporate noncompliance to allow for the differences in compliance observed between controlled clinical trials and routine clinical practice. This review examines the issues relating to the measurement of noncompliance, and the clinical and economic consequences of noncompliant drug taking behaviour. In order to fully appreciate the clinical (and therefore the economic) consequences of noncompliance, a detailed understanding of the type of noncompliance, the pharmacokinetic and pharmacodynamic properties of the drug and the pathophysiological processes of the diseases being treated is required. These are described in detail, and a classification of drug-disease combinations according to the potential economic impact of the varying forms of noncompliance is set out. Issues are raised to highlight the need for improved modelling of the impact of noncompliance, and to this end, recommendations are made for future analyses. The main points are that compliance should be defined clearly, distinguishing between the various forms of noncompliance, that the assumptions relating to the health status of noncompliers should be explicit and robust, and that sensitivity analysis should be applied appropriately to ascertain the impact of noncompliance on the cost-effectiveness of drug therapies.