Background: Non-adherence to antihypertensive drugs is high, and the economic consequences of non-adherence may be substantial. The Medication Events Monitoring System (MEMS), which is a method to improve adherence, has been shown to be a useful tool for the management of adherence problems.
Objective: To assess the cost effectiveness of the MEMS compared with usual care in a population of hypertensive patients with poor adherence. The MEMS programme consisted of provision of containers fitted with electronic caps together with adherence training if indicated.
Methods: In a randomised controlled trial, 164 hypertensive patients in the experimental strategy and 89 patients in the usual care strategy were followed for 5 months. Patients who had a systolic blood pressure (SBP) > or = 160 mm Hg and/or diastolic BP (DBP) > or = 95 mm Hg despite the use of antihypertensive drugs were eligible. Patients were recruited by a GP, and treatment took place in general practice. In the experimental strategy, electronic monitoring of the intake of antihypertensive drugs was introduced without change of medication. Unsatisfactory adherence was defined as < 85% of days with the number of doses taken as prescribed. In the usual care strategy, antihypertensive treatment was intensified by the addition or change of antihypertensive drugs, if necessary, without provision of an electronic monitor. Outcome parameters included the proportion of patients with normalised blood pressure (NBP) at 5 months and QALYs. Costs were quantified from the healthcare and societal perspective. Non-parametric bootstrap simulations were performed to quantify the uncertainty around the mean estimates and cost-effectiveness acceptability curves were presented. In addition, a number of univariate sensitivity analyses were performed on deterministic variables.
Results: At 5 months, 3.1% (95% UI [uncertainty interval] -9.7%, +15.8%) more patients had NBP, and 0.003 (95% UI -0.005, +0.010) more QALYs were generated in the experimental strategy. A statistically significant lower percentage of patients had a dose escalation in the experimental strategy. Irrespective of the ceiling ratio for cost effectiveness, the cost-effectiveness probability was between 75% and 80% for the analysis from the healthcare perspective using proportion of patients with NBP as the outcome parameter. For the analysis from the societal perspective using QALYs as the outcome parameter, this probability was between 45% and 51%.
Conclusion: For a time horizon of 5 months, a difference in both cost and effect could not be detected between an adherence-improving programme compared with usual care for hypertensive patients. The probability that the adherence-improving programme is cost effective is at best moderate. Moreover, the cost-effectiveness result is surrounded with considerable uncertainty and large-scale implementation warrants additional research into the economic consequences of this intervention. Patients may benefit from the use of a MEMS monitor in situations where BP targets are not reached because of suspected non-adherence and both patient and GP are reluctant to increase the dose or number of antihypertensive drugs.