Gastrostomy and internal cerebrospinal fluid shunt in adults. A systematic review and meta-analysis of the risk of infection

Abstract

BackgroundHydrocephalus is a frequent neurological condition, commonly treated by ventriculoperitoneal shunting (VPS), a neurosurgical procedure with significant risk of infection. Some severely brain-injured hydrocephalic patients with swallowing dysfunction may require percutaneous endoscopic gastrostomy (PEG). There are few data on the safety of PEG in patients with VPS, with contradictory results reported. ObjectiveThe aim of this systematic review and meta-analysis was to determine the rate of VPS infection in the setting of PEG. MethodsSix databases were searched for the period January 1990 to June 2022. Only original articles reporting the rate of shunt infection in the setting of PEG in adults were included. Random-effects meta-analysis was used to assess the rate of infection. ResultsFifteen of the 1,703 identified articles were selected, reporting 701 internal cerebrospinal fluid shunts, with 63 infections. The pooled rate of infection in patients with both PEG and VPS was 7.41% (95% CI [3.67–14.38]). There was a significantly higher risk of VPS infection in the PEG group vs. the control group with VPS without PEG: relative risk (RR)=2.33 (95% CI [1.11–4.89]). On the other hand, the risk of infection was the same whether the PEG was placed before or after the VPS surgery: RR=1.05 (95% CI [0.57–1.92]). ConclusionGastrostomy tube placement is a significant risk factor for VPS infection. However, onset of infection was not related to the sequence of or interval between VPS and PEG. Trial registrationThis meta-analysis is registered in https://www.crd.york.ac.uk/PROSPERO/, PROSPERO ID: CRDCRD42022326774.