Thirty RCTs with 11,765 participants were included. Methodological quality was variable: few trials reported on allocation concealment or blinding of outcome assessors, but most were free of selective outcome reporting.
Pharmacist care was associated with statistically significant reductions in systolic (weighted mean difference (WMD) -8.1mmHg, 95% CI -10.2 to -5.9; 19 trials) and diastolic (WMD -3.8mmHg, 95% CI -5.3 to -2.3; 19 trials) blood pressure, total cholesterol (WMD -17.4mg/L, 95% CI -25.5 to -9.2; nine trials) and LDL-C (WMD -13.4mg/L, 95% CI -23.0 to -3.8; seven trials) compared with usual care. There was a significant reduction in the risk of smoking in the pharmacist care group (RR 0.77, 95% CI 0.67 to 0.89; two trials). Substantial heterogeneity was present for all outcomes except smoking.
For blood pressure, no differences in effect were found between pharmacist-directed and pharmacist-collaborative interventions or most of the other subgroups that were examined. The only exception was that interventions that did not involve patient education did not significantly reduce diastolic blood pressure. Sensitivity analyses did not substantially alter the findings. Funnel plots and Egger tests indicated potential publication bias.