Outcome Differences Between Individual and Group Formats When Identical and Nonidentical Treatments, Patients, and Doses Are Compared: A 25-Year Meta-Analytic Perspective

There are mixed findings regarding the differential efficacy of the group and individual format. One explanation of these mixed findings is that nearly all-recent meta-analyses use between-study effect sizes to test format equivalence introducing uncontrolled differences in patients, treatments, and outcome measures. Only 3 meta-analyses were located from the past 20 years that directly tested format differences in the same study using within-study effect sizes; mixed findings were reported with a primary limitation being the small number of studies. However, we located 67 studies that compared both formats in the same study. Format equivalence (g = -0.01) with low effect size heterogeneity (variability) was found in 46 studies that compared identical treatments, patients, and doses on primary outcome measures. Format equivalence (g = -0.06) with moderate effect size heterogeneity was found for 21 studies that compared nonidentical treatments; however, allegiance to a specific format moderated differences in effect sizes. There were no differences between formats for rates of treatment acceptance, dropout, remission, and improvement. Additionally, there were no differences in outcome between formats by patient diagnosis; however, differences in pre-to-post improvement were explained by diagnosis with depression, anxiety, and substance disorder posting the highest outcomes and medical and childhood disorders the lowest. Findings are discussed with reference to the practical challenges of implementing groups in clinical practice from an agency, clinician, and reimbursement perspective.