Abstracts of search results were reviewed to eliminate articles that clearly did not pertain to CE methods, cost-effectiveness, or educational theory and to categorize the studies as informative, equivocal, or not informative of CE effectiveness. A wide range of designs were classified as informative, including randomized controlled trials, prospective cohort studies, observational studies, and studies with pre- and post-intervention assessment methodologies. Quantitative and qualitative approaches were included, and inclusion was not limited to studies with positive results. The most common reasons articles were classified as not informative were absence of a trial design, small sample size, and high likelihood of confounding factors in the design that could affect outcomes. The two reviewers independently classified abstracts and full texts of the articles and then compared their classification results. Interreviewer reliability was greater than 80 percent, and discrepancies were resolved by a consensus process. A third reviewer verified the results classified as informative or equivocal in a final round of detailed assessment of the study design, populations, intervention, type of outcome, and conclusions for each article. Systematic reviews and metaanalyses are included in Table A-1; studies and articles are included in Table A-2.