Background: Cochrane reviews originated to examine medical treatments, i.e., drugs and devices. Now the systematic approach is also applied to interventions intended to help people use the medical treatments. However, methodological heterogeneity in ‘umbrella reviews’ limits the ability to conduct meta-analysis. Search strategies are also challenging. We address implications of reviewing such interventions, make comparisons with reviewing a clinical treatment, and offer alternatives for synthesizing results. Objective: To provide alternative methods for synthesizing results of educational intervention reviews. Methods: We used concept mapping to illustrate the factors influencing the ability to conduct meta-analysis. Educational interventions can vary by content, dose (contacts), format (didactic; discussion), or context (clinic; community). Literature searches are complex; key words (education, counseling, evaluation) are not unique to the review topic. We compared the design heterogeneity in our reviews of theory-based interventions for contraceptive use and of steroidal contraceptive effect on carbohydrate metabolism. Results: Of 26 trials in the theory-based review, 8 randomly assigned clusters and 18 assigned individuals. We identified five groupings for the theoretical basis, e.g., Social Cognitive Theory with or without another model. Outcomes included pregnancy and contraceptive use. Design differences in this review precluded any meta-analysis. To synthesize results, we grouped studies by theoretical basis and comparison intervention. The metabolic review had 28 contraceptive comparisons with varying outcomes across studies (e.g., glucose area under the curve; fasting blood glucose). Three comparisons involved meta-analysis. One progestin group had 11 different comparisons across 10 trials. Education-related literature searches produced twice the number of citations found for metabolic issues. Conclusions: Reviews of clinical interventions may have design heterogeneity similar to reviews of educational interventions. Alternative ways to present results from heterogeneous studies, such as grouping them conceptually, aid interpretation. Such systematic reviews can then provide useful information about effective methods and research
needs.