Adherence with pharmacotherapy for type 2 diabetes: a retrospective cohort study of adults with employer-sponsored health insurance

Clin Ther. 2005 Jul;27(7):1064-73. doi: 10.1016/j.clinthera.2005.07.009.


Background: Trials have shown that intensive therapy to control blood glucose levels results in lower rates of microvascular complications and myocardial infarction among patients with type 2 diabetes. They have also demonstrated the economic and quality-of-life benefits of improved glycemic control among this patient population. Glycemic control achievable in controlled settings, however, may differ from that observed in actual practice settings, in part due to the patient's autonomy in determining whether, or to what extent, adherence to the prescribed regimen is acceptable.

Objective: The goal of this study was to determine adherence with pharmacotherapy for type 2 diabetes among newly treated working-aged adults (ie, those aged 18-64 years) who had employer-sponsored health insurance. Adherence was defined as the regular refilling of prescriptions as indicated, such that an appropriate supply of medication is available over time.

Methods: A retrospective cohort study of newly treated patients (aged 18-64 years) was conducted using an administrative claims database with coverage from 1997 through 2000. Eligibility required at least 12 months of history before and after the index prescription date. Early nonpersistence (failure to fill a second prescription for the index drug or any other antihyperglycemic medication) and 12-month non-persistence rates were calculated, as was nonadherence based on a medication possession ratio (MPR) <80%. Survival and logistic regression models were used to examine adherence rates and behavior predictors.

Results: A total of 6090 patients (median age, 51.0 years; 3263 men, 2827 women) were included. After the first prescription, 10.5% of patients (95% CI, 9.8-11.3) failed to fill a second prescription for the initial or any other antihyperglycemic medication. At 12 months after the initial prescription date, 37.0% of patients (95% CI, 35.8-38.2) had discontinued pharmacotherapy. During the period of persistence (the time interval during which prescriptions were being filled), 46.2% of patients (95% CI, 44.7-47.7) were nonadherent according to the MPR-based analysis. After adjustment for covariates, younger age (ie, 18-24 years) and female gender were found to be risk factors for early nonpersistence (odds ratio [OR], 1.77 [95% CI, 1.07-2.94] and OR, 1.47 [95% CI, 1.25-1.73], respectively) and for discontinuation over time (hazard ratio [HR], 2.44 [95% CI, 1.89-3.15] and HR, 1.18 [95% CI, 1.09-1.28], respectively). Another risk factor for early nonpersistence and discontinuation over time was initial treatment using insulin (OR, 3.00 [95% CI, 2.30-3.91]; HR, 2.68 [95% CI, 2.31-3.10]) or an alpha-glucosidase inhibitor (OR, 2.07 [95% CI, 1.11-3.84]; HR, 1.57 [95% CI, 1.11-2.22]).

Conclusions: Adherence with antihyperglycemic pharmacotherapy was poor among working-aged patients newly treated for type 2 diabetes. Patients prescribed insulin as initial pharmacotherapy were less likely to persist on medication than those initially prescribed oral agents.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Cohort Studies
  • Diabetes Mellitus, Type 2 / drug therapy*
  • Female
  • Health Benefit Plans, Employee
  • Humans
  • Hypoglycemic Agents / administration & dosage*
  • Male
  • Middle Aged
  • Patient Compliance*
  • Retrospective Studies


  • Hypoglycemic Agents