Five studies reported in the review included two randomised controlled trials (RCT) (n=941) and three non-randomised controlled studies (n=750). Study quality was described as variable. All of the studies included comparable study groups, followed up over 95% of participants and used an intention-to-treat analysis. Allocation concealment was reported in only one of the two RCTs, as was study blinding. Sample size ranged from 56 to 600. Two of the non-randomised studies assessed fewer than 100 participants.
Outpatient hysteroscopy was associated with a statistically significant increase in the pregnancy rate in RCTs (RR 1.57, 95% CI 1.29 to 1.92), non-randomised studies (RR 2.01, 95% CI 1.60 to 2.52) and overall (RR 1.75, 95% CI 1.51 to 2.03). There was no evidence of significant statistical heterogeneity in any of the analyses. The significant difference in favour of outpatient hysteroscopy remained after patients with abnormal outpatient hysteroscopy findings were excluded from the analysis (RR 1.63, 95% CI 1.35 to 1.98; two RCTs and two non-randomised controlled trials).
No evidence of publication bias was found.