Study designs of evaluations included in the review
Controlled trials with concurrent arms were eligible if they included asymptomatic women and were based in the United States. Uncontrolled trials, descriptive studies, pre- and post-intervention studies, and those with historical controls were excluded.
Specific interventions included in the review
All interventions to improve physicians' use of breast cancer screening methods were eligible. The interventions included physician reminder systems, other office systems, audit with feedback, and physician and patient education. Most of the interventions were conducted in academic settings; two were community-based.
Participants included in the review
No inclusion criteria were specified with respect to the participants in the primary studies. The review focused on physicians in academic or community settings who conducted clinical breast examinations or referred asymptomatic women for mammography. A range of participant characteristics were tabulated in the article.
Outcomes assessed in the review
No inclusion criteria were specified relating to the outcomes in the primary studies. The main outcome assessed was increased uptake of mammography or clinical breast examination. Effect sizes and 95% confidence intervals (CIs) were calculated from the extracted data. An intervention effect size was calculated for the intervention employed in each study. For studies of multiple interventions, where possible, separate effect sizes were calculated for each intervention.
For randomised controlled trials, the intervention effect size was calculated as the difference in screening rates between the intervention and control groups at the end of the intervention period. For non-randomised trials, the intervention effect size was calculated as the difference in screening rates between the intervention and control groups of the difference between post- and pre-intervention levels.
How were decisions on the relevance of primary studies made?
Two reviewers assessed the studies for relevance based on predefined criteria. Inter-rater agreement on the inclusion criteria was 90%. Any disagreements were resolved by concurrent review and consensus.