The following is the summary of “Disagreement between pharmacy claims and direct interview to identify patients with non-adherence to chronic cardiometabolic medications” published in the February 2023 issue of Heart Journal by Haff, et al.
There is a need for reliable strategies for identifying individuals who are not adhering as well as they should to their cardiometabolic drugs. In this study, researchers analyzed data from a large trial of patients with poorly managed cardiometabolic disease and evidence of prescription non-adherence determined using pharmacy claims data, and whose adherence was subsequently assessed during a telephone consultation with a clinical pharmacist. Investigators then checked to see if the pharmacist’s evaluation matched the level of non-adherence as determined by the claims data. They utilized multivariable modified Poisson regression to identify patient factors associated with disagreement when pharmacist and claims evaluations were at odds.
The pharmacist interviews validated the non-adherence of 646 (60.4% of the total) of the 1,069 patients who were identified as non-adherent using claims (percentage of days covered [PDC] 80%). Among the 423 patients (39.6%) whose interviews yielded discrepancies between their claims and reality, the most common explanations were: the patient paying out of pocket or using a different insurance plan; the patient no longer taking the medication, and the patient only becoming strictly adherent to their treatment plan within the past few months.
Patients whose interview results did not match their claims were less likely to miss outpatient office visits (RR:0.91, 95% CI:0.85-0.97) and more likely to have a baseline PDC above the median (RR:1.35, 95% CI:1.10-1.64). When pharmacists evaluated patients flagged as non-adherent by claims, they found that 39.6 % were actually following their treatment regimens. Many factors contributed to this disparity, including patients’ decisions to forego insurance coverage, switch plans, or stop taking certain medications. These results have substantial bearing on the utility of pharmacy claims as a tool for detecting and addressing medication non-adherence.