Factors associated with early shunt revision within 30 days: analyses from the National Surgical Quality Improvement Program.

Abstract

CSF shunt insertion is the most commonly performed neurosurgical procedure for pediatric patients with hydrocephalus, and complications including infections and catheter obstruction are common. The rate of readmission in the first 30 days after surgery has been used across surgical disciplines to determine healthcare quality. In the current study, the authors sought to assess factors associated with early shunt revision within 30 days using real-world data. Targeted shunt data set participant user files of the National Surgical Quality Improvement Program (NSQIP) from 2016 to 2019 were queried for patients undergoing a shunt procedure. A multivariable logistic regression model was performed to assess the impact of demographics, etiologies, comorbidities, congenital malformations, and shunt adjuncts on shunt revision within 30 days, as well as shunt revision due to infection within 30 days. A total of 3919 primary pediatric shunt insertions were identified in the NSQIP database, with a mean (± SD) patient age of 26.3 ± 51.6 months. There were a total of 285 (7.3%) unplanned shunt revisions within 30 days, with a mean duration of 14.9 ± 8.5 days to first intervention. The most common reason for intervention was mechanical shunt failure (32.6% of revision, 2.4% overall, n = 93), followed by infection (31.2% of all interventions, 2.3% overall, n = 89) and wound disruption or CSF leak (22.1% of all interventions, 1.6% overall, n = 63). Patients younger than 6 months of age had the highest overall unplanned 30-day revision rate (8.5%, 203/2402) as well as the highest 30-day shunt infection rate (3%, 72/2402). Patients who required a revision were also more likely to have a cardiac risk factor (34.7%, n = 99, vs 29.2%, n = 1061; p = 0.048). Multivariable logistic regression revealed that compared to patients 9-18 years old, those aged 2-9 years had significantly lower odds of repeat shunt intervention (p = 0.047), while certain etiologies including congenital hydrocephalus (p = 0.0127), intraventricular hemorrhage (IVH) of prematurity (p = 0.0173), neoplasm (p = 0.0005), infection (p = 0.0004), and syndromic etiology (p = 0.0136), as well as presence of ostomy (p = 0.0095), were associated with higher odds of repeat intervention. For shunt infection, IVH of prematurity was found to be associated with significantly higher odds (p = 0.0427) of shunt infection within 30 days, while use of intraventricular antibiotics was associated with significantly lower odds (p = 0.0085). In this study of outcomes after pediatric shunt placement using a nationally derived cohort, early shunt failure and infection within 30 days were found to remain as considerable risks. The analysis of this national surgical quality registry confirms that, in accordance with other multicenter studies, hydrocephalus etiology, age, and presence of ostomy are important predictors of the need for early shunt revision. IVH of prematurity is associated with early infections while intraventricular antibiotics may be protective. These findings could be used for benchmarking in hospital efforts to improve quality of care for pediatric patients with hydrocephalus.