New study examines incidence, potential reasons for primary medication nonadherence in primary care

Primary medication nonadherence is common in primary care and may be improved by lower costs to patients and increased follow-up, according to a new study.

Researchers in Canada performed a prospective study of primary care patients to determine the incidence of primary medication nonadherence as well as the drug, patient, and physician characteristics associated with it. Patients' incident prescriptions from 2006 and 2009 were obtained from primary care electronic health records and were linked to drug insurer data on drugs dispensed from community pharmacies. The researchers defined primary nonadherence as failure to fill an incident prescription within 9 months and used multivariate alternating logistic regression to estimate predictive factors associated with nonadherence. Results appeared in the April 1 Annals of Internal Medicine.

Overall, 15,961 patients treated by 131 physicians were included in the study. The study population's mean age was 61.55 years, and 62.3% were women. Of the 37,506 incident prescriptions written for these patients, 31.3% were not filled. Prescriptions for higher-cost drugs, skin agents, gastrointestinal drugs, and autonomic drugs were the least likely to be filled, while anti-infective drugs had the lowest incidence of nonadherence by pharmacologic class. Lower odds of nonadherence were associated with older age (odds ratio per 10 years, 0.89; 95% CI, 0.85 to 0.92), eliminated copayments in low-income groups (odds ratio, 0.37; 95% CI, 0.32 to 0.41), and a larger proportion of physician visits with the prescribing physician (odds ratio per 0.5 increase, 0.77; 95% CI, 0.70 to 0.85).

The authors noted that they did not assess the need for the prescribed therapies (e.g., whether they were prescribed to be filled “as needed”) or patients' attitudes and beliefs about medication. However, they concluded that primary nonadherence is common in primary care and may be addressed through lower prescription costs and more patient contact with the prescribing physician. “Future research should estimate the contribution of medication attitudes and beliefs to the likelihood of primary nonadherence as well as the effect of nonadherence on subsequent illness, death, and health care use,” they wrote. “If primary nonadherence is an important contributor to avoidable illness, then policy interventions to minimize risk for primary nonadherence for the most vulnerable groups … should be evaluated.”