Nonadherence as a predictor of antidiabetic drug therapy intensification (augmentation)
Date of Original Version
Purpose. To determine if nonadherence with antidiabetic drug therapy is predictive of subsequent antidiabetic drug therapy intensification. Methods. We conducted a retrospective cohort study examining retail pharmacy dispensings of sulfonylureas or metformin to 1067 patients having diabetes. Patients that did not receive a sufficient quantity of medication to cover at least 80% of days during the evaluation period were classified as nonadherent. Outcomes identified were increase in the dose of antidiabetic medication utilized, the addition of a second antidiabetic agent to the regimen or either. Results. Among users of sulfonylurea monotherapy, those classified as nonadherent were 45% more likely to intensify therapy in subsequent months as compared with those classified as adherent (age-adjusted odds ratio (OR) 1.45; 95% confidence interval (CI) 1.06-2.00). This finding was largely driven by observed increases in dosage, which were more likely among patients classified as nonadherent (age-adjusted OR 1.48, 95% CI 1.07-2.05). Nonadherence was not found to be predictive of the subsequent addition of a second antidiabetic agent (OR 1.02; 95% CI 0.64-1.63). Overall findings were similar for the smaller sample of patients receiving metformin monotherapy, though observed differences did not achieve statistical significance. Conclusions. Patients who were poorly adherent to oral antidiabetic drug therapy more frequently experienced an increase in the dose of medication prescribed, as compared to patients that were classified as adherent. This finding underscores the need for prescribers to consider nonadherence as a root cause when patients fail to achieve therapeutic goals. Copyright © 2004 John Wiley & Sons, Ltd.
Pharmacoepidemiology and Drug Safety
Kogut, Stephen J., Susan E. Andrade, Cynthia Willey, and E. Paul Larrat. "Nonadherence as a predictor of antidiabetic drug therapy intensification (augmentation)." Pharmacoepidemiology and Drug Safety 13, 9 (2004): 591-598. doi:10.1002/pds.1005.