Thirteen studies (n = 347) were included in the review.
Successful dilatation: 203 (58%) of patients were surgery free at the end of follow up. When technical failures for procedural reasons were excluded the figure was 68%. Where failure occurred, the mean interval between dilatation and surgery was 15 months (range one month to 70 months). When studies with fewer than 10 patients were excluded the success rate was 64%.
Complications: Major complications occurred in 14 cases (2%); 13 of these cases were bowel perforation. In 11 of the 13 studies the complication rate was less than 5%. In the other studies it was 11% and 18%.
Meta-regression showed an association between steroid injection and efficacy, as it was performed in 40% of patients above the efficacy cut off (median of 64%) and in only 3% of those below it (p<0.0001).
Results were reported for individual patient data analysis for the nine studies (n = 167) for which this was possible. The success rate in this analysis was 67% (112 patients). The meta-regression showed that a stricture length of 4cm or less was associated with a surgery-free outcome (odds ration was 4.01, 95% CI: 1.16 to 13.8, p<0.028). No other independent variable was found to be significantly associated with successful outcome.