Can We Ever Predict Nonadherence with Drug Therapy?

Despite Much Effort, It is Well Known that adherence with drug therapy, especially for those with chronic conditions, varies and is often poor. Some of the most common indications for long-term drug therapy include cardiovascular risk reduction and treatment with antihypertensives (AH) and lipid-lowering (LL) therapy. Nonadherence with these therapies can start from the point of receiving the prescription through to long-term adherence. Studies have indicated that approximately 15% of patients never have their prescriptions filled and of those who have them filled, another 15% never take the medication. ¹ In order to understand the factors affecting adherence with these medications, Chapman and colleagues conducted a study to identify predictors of adherence with AH and LL therapy. The results of their work were published in Archives of Internal Medicine and although directed to physicians, the material is highly relevant to pharmacists.

The study used a retrospective cohort design that included 8406 patients in a US managed care population — approximately half of the patients were women and half were younger than 65 years of age. Patients were followed for an average of 12.9 months. The authors aimed to identify patient and regimen characteristics that predicted optimal adherence with concomitant therapy. The outcomes measured were adherence with antihypertensives alone, lipid-lowering therapy alone, and the two together. Adherence was measured at 91-day intervals and was defined as “the number of days covered by a given drug class in each time period, based on the number of days supplied and quantity of medication dispensed for each filled prescription.” Patients were considered adherent if they reached the 80% level. One potential limitation of this work is that the actual adherence may have been overestimated.

The results indicated that 44.7% of patients were adherent with both AH and LL therapies three months after the initiation of treatment; at 12 months, this number had decreased to 35.8%. Of interest, at each point of measure, an additional 25%–59% were adherent with either AH or LL therapy. With respect to predictors, it appeared that adherence with AH therapy was on average 10%–15% higher than with LL therapy. It was found that adherence was greatest among those aged 55–64, followed by those aged 65–74 and 45–54 years. Women were also less likely than men to adhere to therapy. Among other predictors, it was apparent that the strongest predictor of adherence was the number of prescriptions taken in the year before starting these therapies. Also of importance was the time between initiation of AH and LL therapy — patients who started the two therapies together had the highest rates of adherence. Finally, patients who had a higher risk of CVD at baseline were also more likely to be adherent to therapy.