We have a serious problem. More than one-half of patients in our clinical practices do not take their medications as prescribed. We wouldn’t prescribe it unless we thought they needed it. Medication nonadherence is associated with $290 billion in otherwise avoidable medical expenditures and 125,000 deaths. One quarter of our patients do not even fill prescriptions we hand them or pick up medications we fax or e-prescribe. Among patients actually picking up medications, reasons cited for stopping them include perceived clinical improvement (59%), lack of improvement (25%), side effect concerns (37%), and cost (24%). In the U.S., nonadherence for antihypertensives is approximately 93% because improving adherence is challenging when “biofeedback” for medication use is not available. Contrast this with narcotics for which nonadherence is low, or disturbingly nonexistent.
Last week, results were published from an impressive randomized controlled clinical trial assessing the effectiveness of telephone call counseling to improve osteoporosis medication adherence (Arch Intern Med. 2012;172:477-83 doi:10.1001/archinternmed.2011.1977). In this study, participants were recruited from a pharmacy benefits program and randomized to telephone-based counseling (n = 1,046) or to a control group (n = 1,041). The telephone-counseling group used motivational interviewing techniques (“roll with resistance” & “support self-efficacy”) and the control group received mailed educational materials.
Adherence was calculated as the number of days with filled prescriptions during the observation period times divided by the number of days in the observation period x 100. The mean age of enrolled patients was 79 years and 94% were female.
Median medication adherence was 49% in the telephone group and 41% in the control group (p-value = non significant). In a subgroup analysis, the intervention appeared to be more effective for subjects aged 65 to 74 years compared with those ≥ 75 years (P = 0.45).
There was no statistically significant difference between the groups in terms of the primary outcome. Tragically, this may mean that stakeholders may not pay attention to what may be a very promising and innovative strategy to improve medication adherence. Importantly, the authors mention that a median of 113 days passed by the time a patient received an intervention, which may be well after the period of greatest susceptibility to an adherence intervention.
Other data would suggest that simplification of drug regimens, patient education and information, reminders by mail, and group sessions may enhance medication adherence. In our practices, widespread use of telephone-based medication adherence programs is limited by resource constraints. But there are ways to piggyback such efforts onto existing telephone-based programs, such as appointment reminder systems. Our challenge will be to determine which medications we want to target.
Dr. Ebbert reported having no relevant conflict of interest.
