Table.
Study [Sponsor] | Nature of intervention | Impact on nonadherence | Findings | Limitations |
---|---|---|---|---|
Antoine (2014)a | Coaching by pharmacist | No impact | US studies identified by the systematic literature review found no impact of pharmacist intervention on adherence • Grant (2003): Very high rates of self-reported adherence with and without pharmacist intervention • Odegard (2005): Also no impact on HbA1c control |
• Limited number of RCT studies available for the review |
O'Connor (2014) [Geisinger] | Diabetes mellitus educator or pharmacist-scripted call | No impact | No improvement in first fill with intervention observed during RCT (10%-20% of patients do not fill first new diabetes mellitus medications) | • No assessment of adherence response beyond first fill • Small sample/limited representativeness (2378 patients, single payer) |
Magee (2014) [Sanofi] | 3-month education program on adherence | Improved likelihood of high adherence | Odds of “high adherence” multiplied by 3.75 at month 3 (patient scale assessment, not based on actual prescription claims) | • No control arm • Nonrandomized study • Self-reported adherence measure (using MMAS-4 screening tool) • Representativeness may be limited (intervention in hospitalized patients with severely uncontrolled diabetes, a majority of African American women) |
Elliot (2013) [CCHS] | Copayment waiver | From 41% to 17.5% after 12 months (cost-related only) | Intervention helped reduce cost-related nonadherence, but patients experienced no change in HbA1c, HCRU, or healthcare costs | • No control arm • Study did not provide estimates of impact on nonadherence due to reasons other than cost • By its nature, copayment waiver will remove perception of non-adherence being cost-related, but overall nonadherence may not have been impacted • Self-reported adherence measure |
Wertz (2012) [Kroger, BCBS OH, Novartis] | Copayment support (>$500) and pharmacist coaching | From 22% to 14% after 12 months (estimated as 1-PDC) | • Intervention helped improve adherence to noninsulin antidiabetic agents, reducing HbA1c from 7.9% to 7.1% • However, medical costs increased more for the intervention versus control cohort (+33.2% vs +20.8%—even without considering cost of intervention) |
• Nonrandomized study (control based on retrospective matching) • Kroger employees may be particularly sensitive to pharmacy intervention |
Schmittdiel (2009) [Kaiser] | Nurse-led integrated CM program | No impact | After 12 months: • % of adherent patients was not significantly different for the CM vs matched control patients • However, HbA1c levels of CM patients were 0.3% lower than that of their controls after multivariate adjustment (P <.01) |
• Nonrandomized study (control based on retrospective propensity score matching) |
Literature review of systematic randomized controlled clinical trials.
CM indicates care management; Hb, hemoglobin; HCRU, healthcare resource utilization; MMAS, Morisky Medication Adherence Scale; PDC, proportion of days covered; RCT, randomized controlled trial.