Value of the periodic health evaluation.

Objectives: To systematically review evidence on definitions of the periodic health evaluation (PHE), its associated benefits and harms, and system-level interventions to improve its delivery.

Data sources: Electronic searches in MEDLINE, and other databases; hand searching of 24 journals and bibliographies through February 2006.

Review methods: Paired investigators abstracted data and judged study quality using standard criteria. We reported effect sizes for mean differences and proportions in randomized controlled trials (RCTs). We adapted GRADE Working Group criteria to assess quantity, quality and consistency of the best evidence pertaining to each outcome, assigning grades of "high," "medium," "low," or "very low."

Results: Among 36 identified studies (11 RCTs), definitions of the PHE varied widely. In studies assessing benefits, the PHE consistently improved (over usual care) the delivery/receipt of the gynecological exam/Pap smear (2 RCTs, small effect (Cohen's d (95% confidence interval (CI)):0.07 (0.07,0.07)) to large effect (Cohen's d (CI):1.71 (1.69, 1.73)), strength and consistency graded "high"); cholesterol screening (1 RCT, small effect (Cohen's d (CI):0.02 (0.00,0.04)) with large associations in 4 observational studies, graded "medium"); fecal occult blood testing (2 RCTs, large effects (Cohen's d (CI): 1.19 (1.17, 1.21) and 1.07 (1.05, 1.08)), graded "high"). Effects of the PHE were mixed among studies assessing delivery/receipt of counseling (graded "low"), immunizations (graded "medium"), and mammography (graded "low"). In one RCT, the PHE led to a smaller increase in patient "worry" (13%) compared to usual care (23%) (graded "medium"). The PHE had mixed effects on serum cholesterol (graded "low"), blood pressure, body mass index, disease detection, health habits and health status (graded "medium"), hospitalization (graded "high"), and costs, disability, and mortality (graded "medium"). No studies assessed harms. Delivery of the PHE was improved by scheduling of appointments for PHE (1 RCT, medium effects (Cohen's d (CI): 0.69 (0.68, 0.70)) and offering a free PHE (1 non-RCT, 22% increase) (graded "medium").

Conclusions: The evidence suggests delivery of some recommended preventive services are improved by the PHE and may be more directly affected by the PHE than intermediate or long-term clinical outcomes and costs. Descriptions of the PHE and outcomes were heterogeneous, and some trials were performed before dissemination of recommendations by the U.S. Preventive Services Task Force, limiting interpretations of findings. Efforts are needed to clarify the long-term benefits of receiving multiple preventive services in the context of the PHE. Future studies assessing the PHE should incorporate diverse populations, carefully define comparisons to "usual care," and comprehensively assess intermediate outcomes, harms, and costs.