Five RCTs (n=1,762 patients), two of which were cross over studies (n=762 patients), were included in the review. Four studies had adequate generation of allocation. Allocation concealment was reported in only two studies. No studies were blinded. Four studies conducted sample size calculations and reported adequate power for the outcomes measured.
Wired cannulation was associated with a statistically significant greater cannulation success rate compared to the standard method (OR 2.05, 95% CI 1.27 to 3.3; five studies). However, there was evidence of statistically significant heterogeneity (Cochran's Q test 12.65, p=0.01, I2=68.4%). Sensitivity analysis revealed that one study was the source of heterogeneity. Wire cannulation was still associated with a statistically significant greater cannulation success rate compared to standard method when this study was removed from the meta-analysis (OR 1.66, 95% CI 1.14 to 2.43). There was no evidence of statistically significant heterogeneity.
Wired cannulations was associated with statistically significantly lower post-ERCP pancreatic rates (defined as pancreatic pain for more than 24 hours and serum amylase more than three times the upper normal limit) compared to the standard method (OR 0.23, 95% CI 0.13 to 0.41; three parallel studies only). There was no evidence of statistically significant heterogeneity.
There was no evidence of publication bias although, as the authors stated, the number of trials was not large enough to analyse this properly.