Six studies (2,504 participants, range 48 to 1,570) were included in the review: two RCTs (353 participants) and four quasi-experimental studies. Both RCTs were considered to be of moderate quality (scored 3 on the Jadad Scale). Both RCTs reported appropriate randomisation and description of withdrawals and drop-outs but neither was double blind. The other studies were considered to be low quality (scored 3 or 4 on the Newcastle-Ottawa Scale).
Length of stay (three studies, one RCT): Women stayed significantly longer in integrated treatment programmes compared to women in non-integrated programmes (ES 0.35, 95% CI 0.28 to 0.47). The authors considered this a small effect. There was no evidence of statistical heterogeneity for this analysis. Rosenthal’s file drawer statistic indicated 31 studies with null results would be required to reduce significance to “just significant”.
Treatment completion (six studies, two RCTs): There was a small non-significant increase in the number of women who completed treatment in integrated programmes compared those in non-integrated programmes (ES 0.38, -0.05 to 0.80). There was evidence of significant heterogeneity for this analysis. Moderator analyses did not significantly alter the results. Capture/re-capture assessment indicated a 90% capture rate.