Abstract
PURPOSE: Shunt infection is a major problem in paediatric neurosurgery. Our institution introduced a mandatory shunt protocol with the aim of reducing infection rate. METHODS: A retrospective cohort study including consecutive patients undergoing permanent shunt operations (primary insertion and revision) across two study periods: 3 years immediately prior (2009-2012) and 3 years immediately after (2012-2015) protocol introduction. Absolute and relative risk reductions (ARR/RRR) and Chi-square statistical analysis was used alongside logistic regression, where any single factor with p ≤ 0.20 included in the multivariate model, producing an odds ratio (OR). RESULTS: Eight hundred nine operations in 504 children were identified (442 pre-protocol, 367 post). Overall infection rate decreased from 5.43% (24/442) pre-protocol to 3.27% (12/367) post-protocol (ARR = 2.16%, RRR = 39.8%, NNT = 46.3, p = 0.138), which did not reach statistical significance. For primary shunt insertions, infection rate reduced from 3.63 to 2.55% (ARR = 1.08%, RRR = 29.8%, NNT = 92.6, p = 0.565), whilst for revisions, it reduced from 6.83 to 3.81% (ARR = 3.02%, RRR 44.2%, NNT = 33.1, p = 0.156). Multivariate logistic regression showed that surgeon experience was a statistically significant predictor of infection, whilst responsible pathogens and latency were similar across the pre- and post-protocol groups. CONCLUSION: The protocol reduced overall infection rate in primary and revision shunt operations and we recommend paediatric units consider introducing a similar protocol for these procedures.