Meta-analysis of laparoscopic versus open pyloromyotomy

Abstract

Objective: To perform a meta-analysis of studies comparing open pyloromyotomy (OP) and laparoscopic pyloromyotomy (LP) in the treatment of infantile hypertrophic pyloric stenosis.

Background: LP has become increasingly popular for the management of pyloric stenosis. Despite a decade of experience, the real benefit of LP over the open procedure remains unclear.

Methods: Using a defined search strategy, studies directly comparing OP with LP were identified (n = 8). Data for infants treated by both approaches were extracted and used in our meta-analysis. OP and LP were compared in terms of complications, efficacy, operating time, and recovery time. Weighted mean difference (WMD) between continuous variables and 95% confidence intervals (95% CI) were calculated. For dichotomous data, relative risk (RR) and 95% CI were determined.

Results: Only 3 studies were prospective, and just 1 study was a prospective randomized controlled trial. Mucosal perforations and incomplete pyloromyotomy were both more common with LP. Compared with OP, LP is associated with higher complication rate (RR 0.81 [0.5, 1.29], P = 0.4), similar operating time (WMD 1.52 minutes [-0.26, 3.29], P = 0.09), shorter time to full feeds (WMD 8.66 hours [7.25, 10.07], P < 0.00001), and shorter postoperative length of stay (WMD 7.03 hours [3.74, 10.32], P = 0.00003).

Conclusions: OP is associated with fewer complications and higher efficacy. Recovery time appears significantly shorter following LP. A prospective randomized controlled trial is warranted to fully investigate these and other outcome measures.

FIGURE 1. Forest plot comparing rates of mucosal perforation (A), incomplete pyloromyotomy (B) and wound infection (C) between infants treated with OP and LP. The blocks indicate the estimate of RR and their size relates to the size of the individual study. Blocks to the right of the line of no effect (RR >1.0) favor LP and those to the left of the line of no effect favor OP. The diamond indicates the overall estimate from the meta-analysis. χ² Test for heterogeneity between studies and the significance of the overall effect are indicated.

FIGURE 2. Forest plot comparing total complication rates between infants treated with OP and LP. For explanation, see legend to Figure 1.

FIGURE 3. Forest plot comparing operation time for infants treated with OP and LP. The weighted mean difference (WMD) in minutes and 95% confidence interval between OP and LP was calculated. The blocks indicate the estimate of the WMD and their size relates to the size of the study. Blocks to the right of the line of no effect (WMD >0) favor LP and to the left of the line of no effect (WMD <0) favor OP. The diamond indicates the overall estimate from the meta-analysis. χ² Test for heterogeneity between studies and the significance of the overall effect are indicated.

FIGURE 4. Forest plot comparing time to full feeds (A) and postoperative length of stay (B) between infants treated with OP and LP. Times are in hours. For explanation, see legend to Figure 3.

FIGURE 5. Forest plot of subgroup analysis of prospective studies only comparing operation time (A), time to full feeds (B), and postoperative length of stay (C) between infants treated with OP and LP. For explanation, see legend to Figure 3.