Table 3. Logistic Regression Models Adjusting for Age, Sex, and Primary Care Provider Participation in the Primary Care Information Project
AllPCMH Pilot Practices Removed
OR*95% CIOR*95% CI
Diabetes (only)
    Diabetes-specific medications
        Non-PCIP1.140.81–1.601.120.79–1.58
        PCIP2.031.08–3.831.760.96–3.24
    Lipid-controlling medications§
        Non-PCIP0.850.55–1.320.830.53–1.29
        PCIP1.640.73–3.651.590.73–3.44
Hypertension (only)
    Hypertension-specific medications
        Non-PCIP1.261.15–1.381.261.15–1.38
        PCIP1.241.03–1.491.241.03–1.48
    Lipid-controlling medications§
        Non-PCIP1.281.13–1.441.281.13–1.49
        PCIP1.110.86–1.441.160.90–1.50
Comorbid diabetes and hypertension
    Diabetes-specific medications
        Non-PCIP1.030.90–1.181.050.92–1.21
        PCIP1.060.81–1.401.140.88–1.49
    Hypertension-specific medications
        Non-PCIP1.191.04–1.361.191.04–1.36
        PCIP1.100.84–1.441.100.85–1.43
    Lipid-controlling medications§
        Non-PCIP1.140.99–1.321.161.00–1.33
        PCIP1.260.93–1.721.250.94–1.68
  • * Odd ratios (OR) shown for changes in the odds of the proportion of members who are adherent (i.e., medication possession ratio ≥80%) from baseline to the end of the study.

  • Diabetes therapeutic drug classes include biguanides, sulfonylureas, thiazolidinediones, insulins, dipeptidyl peptidase IV inhibitors.

  • P values <.05.

  • § Lipid-controlling therapeutic drug classes include HMG-CoA reductase inhibitors and fibric acid derivatives.

  • Hypertension therapeutic drug classes include angiotensin-converting enzyme inhibitors, angiotensin II receptor antagonists, beta-adrenergic blocking agents, calcium channel blocking agents (dihydropyridines), phosphodiesterase inhibitors, thiazide diruetics.

  • CI, confidence interval; PCMH, patient-centered medical home; OR, odds ratio; PCIP, Primary Care Information Project.