Using Meta-Analysis to Explore the Factors Affecting the Potency of Pharmacists’ Patient Interventions

Abstract

Specific Aims: To identify the factors that affects the potency of pharmacists’ interventions. Methods: Literature search was based on keywords and Mesh terms in eight different databases. The inclusion criteria were evidence of pharmacist involvement in direct patient care, patient-related therapeutic outcomes, studies done in the United States, randomized controlled trials, studies with reported number of subjects in the intervention and control group and reported means and standard deviations of therapeutic outcomes. For the study selection and data extraction, two students independently reviewed each study and met to resolve any discrepancies. In addition, each study was assigned a potency score using the potency tool. Data extraction included: pharmacists’ interventions (technical, behavioral, educational, and affective), patient characteristics, and therapeutic outcomes. The standardized mean difference (SMD) was calculated; studies with SMD ≥ -0.3 formed the low impact group (controls) and studies with SMD  -0.8 formed the high impact group (cases). Main Results: The included randomized control trials (N=11) were conducted in a variety of settings from ambulatory clinics to hospital. The high impact group was favored in the educational category (ES=0.88, p=0.18) while the low impact group was favored in the behavioral category (ES=-0.19, p=0.81). In general, there was a difference between the high impact and low impact (ES=0.82, p=0.37) groups with the high impact group being favored. Conclusion: There is a difference between the low impact and high impact groups, but it is unclear which pharmacist interventions have a significant impact on therapeutic outcomes.