Systematic review of ventricular peritoneal shunt and percutaneous endoscopic gastrostomy: a safe combination

OBJECTIVE Various international and national gastrointestinal guidelines take different positions on whether ventriculoperitoneal shunt (VPS) insertion is a contraindication to percutaneous endoscopic gastrostomy (PEG). The objective of this meta-analysis was to try to answer the question of whether VPS insertion is a contraindication to PEG. METHODS A systematic review of the literature was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) criteria. Electronic databases PubMed and Embase were searched using variations of the terms "ventriculo-peritoneal shunt" and "percutaneous (endoscopic) gastrostomy." This search resulted in 70 studies, 9 of which were relevant. These were cross-referenced, and 1 additional study was found, resulting in 10 studies in this systematic review. RESULTS The 10 relevant studies in adult cohorts included 208 patients. All studies save one were retrospective and, in general, poor quality. Among the studies with relevant data, there were 26 (12.5% of 208 cases) VPS infections and 4 (4.4% of 90 cases) VPSs that malfunctioned. In 137 patients the VPS had been placed before the PEG tube, with a VPS infection rate of 4.4%. More VPS infections occurred among the 55 patients who first had a PEG and a subsequent VPS (21.8%) and in the 16 patients who had simultaneous PEG tube and VPS placement (50%). The heterogeneity of the studies in this analysis prohibited statistical comparisons of the timing of VPS and PEG tube placement. CONCLUSIONS This systematic review indicated that VPS placement in combination with a PEG has a high but acceptable VPS complication rate. Therefore, VPS insertion should not be considered a contraindication to the placement of a PEG tube. Preferably, a PEG tube should be placed after the VPS. Waiting 7-10 days between VPS insertion and a PEG seems reasonable, but this could not be corroborated in this review.