Evaluation of studies investigating the effectiveness of pharmacists' clinical services.

A quantitative evaluation of randomized trials of counseling, education, and other clinical services provided by pharmacists was performed. Data sources were MEDLINE and the bibliographies of published articles. Pharmacists' services were categorized as counseling of patients, counseling of physicians, counseling of both patients and physicians, and patient care. The outcomes extracted were measures of patient behavior, disease, symptoms, and patient knowledge. Thirty-two trials met the inclusion criteria. The pharmacists were specified as clinical pharmacists in 24 trials and as community pharmacists in 2. In six unblinded trials of patient counseling, the outcomes favored the counseled patients over control patients in every trial, and the effects were statistically significant in five trials (the outcome was medication adherence in these five trials). In seven trials of counseling of both patients and their physicians, patient outcomes were significantly better in the intervention group in six trials, four of which were single blind. Two trials in which patients were randomized to either physician counseling or control groups yielded inconsistent results. In one trial in which physicians were randomized to receive counseling from pharmacists, the proportion of prescriptions meeting guidelines was higher in the counseling group than in the control group. Four trials of patient care by pharmacists were inconclusive. These trials demonstrated that counseling of patients and their physicians by pharmacists can improve patient outcomes. The evidence that counseling of patients alone improved patient outcomes was good, though weaker because of suboptimal trial design.