Study selection and trial characteristics
According to the search strategy, we identified 1563 studies. Of these, 420 duplicate articles were excluded, and we retrieved the remaining 1143 studies based on their titles and abstracts. After excluding irrelevant articles for various reasons, we included a final 14 trials involving a total of 2043 patients [6, 9-12, 15-23]. A flowchart of the literature search process is shown in Figure 1, and the characteristics and quality evaluation of the included studies are shown in Table 1. There were four randomized controlled trials (RCTs), one prospective study, and eight retrospective studies. The isolated pancreaticojejunostomy group comprised 482 patients, the isolated gastrojejunostomy group comprised 92 patients, and the isolated pancreaticojejunostomy + gastrojejunostomy group comprised 112 patients. The sample sizes among the studies ranged from 40 to 700 patients.
Meta-analysis
POPF
Twelve studies provided data regarding POPF, with 908 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group, 259 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group, and 700 patients in the isolated pancreaticojejunostomy + gastrojejunostomy group versus the conventional pancreaticojejunostomy group. Regarding POPF, we found no significant difference between isolated pancreaticojejunostomy and conventional pancreaticojejunostomy (OR = 0.83, 95% CI: 0.58–1.18; P=0.29)(Fig.2a) or between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 0.78, 95% CI: 0.42–1.42; P=0.41) (Fig.2a). One study evaluating isolated pancreaticojejunostomy + gastrojejunostomy performed by Grobmyer et al. showed that conventional pancreaticojejunostomy was associated with lower rates of POPF vs isolated pancreaticojejunostomy + gastrojejunostomy (OR = 2.90, 95% CI: 1.53–5.48; P=0.001) (Fig.2a).
CR-POPF
Twelve studies provided data for the incidence of CR-POPF, with 983 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group, and 259 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. Our meta-analysis revealed no significant difference regarding the incidence of CR-POPF between isolated pancreaticojejunostomy and conventional pancreaticojejunostomy (OR = 0.70, 95% CI: 0.35–1.42; P=0.32) or between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 0.81, 95% CI: 0.23–2.90; P=0.74) (Fig.2b).
DGE
Thirteen studies provided data for the incidence of DGE, with 908 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group, 259 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group, and 80 patients in the isolated pancreaticojejunostomy + gastrojejunostomy versus conventional pancreaticojejunostomy group. We found no significant differences regarding DGE when comparing isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy (OR = 1.10, 95% CI: 0.76–1.58; P=0.62), isolated gastrojejunostomy versus conventional pancreaticojejunostomy (OR = 035, 95% CI: 0.11–5.20; P=0.78), or pancreaticojejunostomy + gastrojejunostomy versus conventional pancreaticojejunostomy (OR = 0.99, 95% CI: 0.50–1.95; P=0.96) (Fig.2c).
CR-DGE
Four studies provided data for the incidence of CR-DGE, with 239 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group and 259 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. We found no significant differences for CR-DGE between isolated pancreaticojejunostomy and conventional pancreaticojejunostomy (OR = 2.18, 95% CI: 0.79–6.06; P=0.13) or between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 0.36, 95% CI: 0.10–1.31; P=0.12) (Fig.2d).
Bile leakage
Eleven studies provided data for the incidence of bile leakage, with 945 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group and 259 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. We found no significant difference for the incidence of bile leakage between isolated pancreaticojejunostomy and conventional pancreaticojejunostomy (OR = 0.68, 95% CI: 0.27–1.69; P=0.40) or between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 0.67, 95% CI: 0.28–1.63; P=0.38) (Fig.3a).
Hemorrhage
Nine studies provided data for the incidence of hemorrhage, with 228 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group and 80 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. We found no significant difference regarding the incidence of hemorrhage between isolated pancreaticojejunostomy and conventional pancreaticojejunostomy (OR = 0.77, 95% CI: 0.38–1.57; P=0.47) or between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 0.57, 95% CI: 0.13–2.55; P=0.46) (Fig.3b).
Reoperation
Nine studies provided data for the reoperation rate, with 659 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group and 80 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. We found that isolated pancreaticojejunostomy was associated with a lower reoperation rate versus conventional pancreaticojejunostomy (OR = 0.36, 95% CI: 0.15–0.86; p=0.02), but there was no significant difference between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 0.65, 95% CI: 0.10–4.11; P=0.65) (Fig.3c).
Operation time
Four studies provided data for operation time, with 436 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group and 80 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. Our results showed that conventional pancreaticojejunostomy was associated with shorter operation times versus isolated pancreaticojejunostomy (WMD = 43.61, 95% CI: 21.64–65.58; P=0.00); however, there was no significant difference between isolated gastrojejunostomy and conventional pancreaticojejunostomy (WMD = 23.00, 95% CI: −14.92–60.92; P=0.23) (Fig.3d).
Postoperative hospital stay
Six studies provided data for the length of postoperative hospital stay, with 659 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group and 80 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. We found no significant difference for postoperative hospital stay between isolated pancreaticojejunostomy and conventional pancreaticojejunostomy (WMD = −2.01, 95% CI: −5.66–1.65; P=0.53) or between isolated gastrojejunostomy and conventional pancreaticojejunostomy (WMD = 3.67, 95% CI: −7.89–15.22; P=0.28) (Fig.4a).
Major complications
Four studies provided data describing major complications, with 228 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group and 80 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. Our results showed that isolated pancreaticojejunostomy was associated with fewer major complications versus conventional pancreaticojejunostomy (OR = 0.35, 95% CI: 0.13–0.96; P=0.04), but there was no significant difference between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 1.00, 95% CI: 0.35–2.86; P=1.00) (Fig.4b).
Overall complications
Twelve studies provided data describing the overall complications rate, with 983 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group and 259 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group. Our meta-analysis revealed no significant difference between isolated pancreaticojejunostomy and conventional pancreaticojejunostomy (OR = 1.08, 95% CI: 0.83–1.40; P=0.56) or between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 1.04, 95% CI: 0.54–2.01; P=0.91) (Fig.4c).
Mortality
Thirteen studies provided data for mortality rates, with 983 patients in the isolated pancreaticojejunostomy versus conventional pancreaticojejunostomy group, 259 patients in the isolated gastrojejunostomy versus conventional pancreaticojejunostomy group, and 700 patients in the isolated pancreaticojejunostomy + gastrojejunostomy versus conventional pancreaticojejunostomy group. We found no significant difference in mortality rates between isolated pancreaticojejunostomy and conventional pancreaticojejunostomy (OR = 0.87, 95% CI: 0.41–1.83; P=0.71), between isolated gastrojejunostomy and conventional pancreaticojejunostomy (OR = 1.23, 95% CI: 0.20–7.64; P=0.82), or between isolated pancreaticojejunostomy + gastrojejunostomy versus conventional pancreaticojejunostomy (OR = 0.65, 95% CI: 0.08–5.27; p=0.69) (Fig.4d).
Publication bias and sensitivity analysis
The funnel plots for the parameters were symmetrical, and Egger’s test revealed no significant publication bias. The sensitivity analyses confirmed the robustness of the results.