reports

Adherence to prescription drugs and adverse health events for patients with hypertension

Los Angeles, CA, USA: University of California, Los Angeles (doctoral dissertation).

University of California, Los Angeles, CA, USA

Increased technology in the development of effective and efficacious drugs to treat chronic diseases has stimulated greater interest in understanding how to improve patients’ adherence to prescription drugs and what health benefits can be obtained from greater adherence. Higher patient cost-sharing for drugs has one of the strongest influences in lowering adherence to these drugs. Whether this relationship is consistent for low and high utilizers of prescription drugs, however, is unknown. Results from this study of working-age patients with hypertension show that adherence to antihypertensive drugs was relatively high. Quantile regression results show that at the 10th percentile of adherence, patients with drug copayments greater than $5 or drug coinsurance had drug adherence 8-9 points less as measured by the medication possession ratio (MPR) compared to patients with the lowest cost-sharing, $5 or less copayments. At the median, the effects of cost-sharing were more modest, with differences of 2-3 points in MPR between the lowest cost-sharing group and the higher cost-sharing groups, and costs-haring was no longer significant at the higher percentiles of adherence. Low utilizers of prescription drugs may be much more likely to experience adverse side effects of drugs and be poorer, less educated, and take multiple medications, and, as a result, face greater barriers to drug adherence.

Reduced adherence from higher cost-sharing has serious, potential health consequences for patients with hypertension. This study estimates the effect of adherence to antihypertensive drugs on acute, adverse health events related to cardiovascular disease occurring over a short-term period. These results extend prior research by using a linear measure of drug adherence and econometric methods to address potential omitted variables bias. Results from a two-stage least squares model show considerably higher predicted probabilities of adverse health events for low adherence levels compared to the single-equation model. However, tests of the instruments provide evidence to suggest that drug cost-sharing and prevention orientation are not valid instruments for adherence. Maintaining low out-of-pocket patient payments for drugs is an important policy consideration for improving adherence to drugs and reducing overall health care costs, particularly for patients with poor adherence.