A Systematic Review of Adherence With Medications for Diabetes

Cramer JA. Diabetes Care. 2004;27:1218-1224.

Twenty quantitative studies of adherence and persistence for prescribed diabetes medications were reviewed. Retrospective surveys, prospective clinical trials, and prospective studies of adherence interventions for populations and individuals with type 1 or type 2 diabetes were included. Adherence rates were defined as the proportion of doses taken as prescribed. Persistence was defined as either the proportion of patients who remained on prescribed treatment for a specified period, or the mean time to treatment discontinuation. Variations in experimental methodologies limited certain comparisons between studies. Review of the 11 retrospective studies revealed adherence rates ranging from 36% to 93%. Reasons for poor adherence included depression (85% vs 93%), twice-daily dosing (52% vs 61%), polypharmacy (36% vs 49%), and low socioeconomic status in patients starting oral agents (<50% for a cohort of California Medicaid patients). Treatment persistence ranged from 16% to 80% over intervals of 6-24 months. Review of the 3 small prospective studies of electronic dose monitoring revealed adherence rates of 61% to 85% for up to 6 months’ duration. One of these studies showed dramatically poorer adherence for 3-times daily regimens (38.1±36%) vs one- (79.1±19%) or two-times (65.6±30%) daily regimens, and that unsuspected inconsistencies in dosing intervals could lead to inappropriate prescribing decisions. Electronic dose monitoring revealed erratic dosing in 47% of all patients, and consistent underreporting of missed doses. Previous studies have shown even lower adherence rates for insulin use than for oral agents. Missed insulin doses were associated with higher A1C levels and hospitalizations for diabetic ketoacidosis and other complications in 36% of patients with the poorest adherence rates. The review concurs with other studies of prescription compliance that adherence rates are not related to regimen simplicity, disease severity, or the possible consequences of missed doses. Clinicians should not assume that patients with either type 1 or type 2 diabetes are fully adherent to insulin regimens. High glucose or A1C levels may be related to missed doses, not underprescribing. Electronic dose monitoring, once-daily dosing, dosing schedules that permit taking multiple medications simultaneously, and specific patient education regarding missed doses and adverse effects of medications are recommended to improve individual and population-based health outcomes.