Sixteen studies (n=2,150) were included: 3 RCTs (n=445) and 13 OCS (n=1,705).
RCTs.
One study reported a sample size calculation. Two studies performed intra-operative randomisation but only one described the methods used. One study reported blinding and one defined outcome measures. The RCTs were graded as level 1b evidence.
There was no significant difference between PG and PJ in overall post-operative complications (OR 0.93, 95% CI: 0.63, 1.38, p=0.71), mortality (OR 1.10, 95% CI: 0.42, 2.93, p=0.51) or pancreatic fistula (OR 0.85, 95% CI: 0.50, 1.44, p=0.54). No statistical heterogeneity was found for these analyses.
There was no significant difference between PG and PJ in intra-abdominal fluid collection, delayed gastric emptying or bile leakage, but the results were statistically heterogeneous.
Length of hospital stay was significantly shorter in the PG group than in the PJ group (WMD -0.73, 95% CI: -1.42, -0.05, p=0.04; based on 2 studies).
OCS.
Six studies were classified as level 2b evidence; the other seven were graded as level 4 evidence.
The funnel plot based on pancreatic fistula suggested the absence of smaller studies showing no benefits for PG.
The studies showed a statistically significant reduction in post-operative complications (OR 0.53, 95% CI: 0.34, 0.82, p=0.005; based on 2 studies), pancreatic fistula (OR 0.22, 95% CI: 0.12, 0.40, p<0.00001; based on 12 studies), mortality (OR 0.36, 95% CI: 0.22, 0.58, p<0.0001; based on 12 studies), intra-abdominal fluid collection (OR 0.47, 95% CI: 0.24, 0.92, p=0.03; based on 5 studies) and bile leak (OR 0.32, 95% CI: 0.13, 0.79, p=0.01; based on 7 studies) in patients who had undergone PG compared with PJ. There was evidence of heterogeneity when comparing pancreatic fistulae.
There was no significant difference between PG and PJ in delayed gastric emptying or length of hospital stay. The results for length of hospital stay were heterogeneous.
Analysis of only higher quality OCS papers revealed no statistically significant differences.