Detection of medication nonadherence through review of pharmacy claims data.

Abstract

The utility of pharmacy claims data in detecting improper medication use, medication changes, and formulary adherence was assessed. Pharmacists provided six months of pharmacy claims data to primary care physicians for patients on the day of their scheduled clinic appointments. Similar data, not supplied to physicians, were generated for a matched population of patients on the day of their scheduled clinic appointments. A blinded pharmacist abstracter reviewed dictated office notes for both control and treatment groups. Medication adherence and medication changes were assessed by the abstracter as well as the difference in total medication costs for each of the groups before and after the pharmacy claims data were supplied. Surveys were distributed to physicians to determine whether pharmaceutical care was improved by the pharmacy claims data. In the treatment group, physicians detected medication nonadherence in 30.5% of their patients, while the abstracter noted nonadherence in 58.1% of patients. Physicians failed to detect any nonadherence in the control group, but the abstractor detected nonadherence in 57.1% of these patients. Changes in medication regimens occurred more often in the treatment group (p < 0.001). The mean percentage of patients switched to formulary agents significantly differed between the treatment and control groups (27.7% versus 0.0%, respectively) (p < 0.001). There were no differences in median drug costs for either group before or after the pharmacy claims data were provided. Provision of pharmacy claims data to physicians helped them detect medication nonadherence, evaluate therapeutic duplication or omissions, increase formulary use, and reduce the time required to obtain an accurate medication history.