BackgroundAt Red Cross War Memorial Children’s Hospital (RCCH), it is the preferred practice to use non-ventriculoperitoneal (non-VP) shunts when the peritoneum is ineffective or contraindicated for cerebrospinal fluid (CSF) diversion and when endoscopy is not an option. The objective of this study is to evaluate the clinical course of patients having undergone these procedures.MethodA single-centre retrospective review at RCCH wherein 43 children with a total of 59 episodes of non-VP shunt placement over a 12-year period were identified for inclusion.ResultsTwenty-five ventriculoatrial (VA) and 32 ventriculopleural (VPL) shunts were analysed with a median age at insertion of 2.9 (0.3–14.9) and 5.3 years (0.5–13.4), respectively. The median number of previous shunt procedures prior to VA or VPL shunt insertion was 6.0 (2–28) versus 4.5 (2–17), respectively. Three VA (12.0%) and three VPL (9.4%) shunt patients were lost to follow-up. Of those remaining, 10 VA shunts (45.5%) compared to 19 (65,5%) VPL shunts required revision. One ventriculovesical shunt and one ventriculocholecystic shunt were placed in the same patient after 21 and 25 shunt-related procedures, respectively, and both were revised within 3 weeks of insertion. Median shunt survival was 8 months longer for the VA compared to the VPL shunts, being 13.5 (0–67) and 5 months (0–118), respectively. Complications for VA shunts were low, with the overall shunt sepsis rate in the VA group at 4% (n = 1) compared to 15.6% (n = 5) in the VPL group.ConclusionOur findings support that VA and VPL shunts are acceptable second-line options in an already compromised group of patients where safe treatment options are limited, provided attention is paid to the technical details specific to their placement.