What question did this study set out to answer?

To identify early postoperative predictors of post-traumatic hydrocephalus in adults after decompressive craniectomy.

March 24, 2026Open Access

Early risk prediction to guide surveillance for post-traumatic hydrocephalus after decompressive craniectomy in adult TBI: a single-center retrospective study

Puntos clave

To identify early postoperative predictors of post-traumatic hydrocephalus in adults after decompressive craniectomy.
Conducted a single-center retrospective cohort study
Included adult patients with severe traumatic brain injury who underwent decompressive craniectomy
Used multivariable Firth penalized logistic regression for analysis
Measured performance with AUC and Brier score
Considered only early perioperative variables as predictors
26.7% of patients developed post-traumatic hydrocephalus during follow-up
Intraventricular hemorrhage and high-energy injury mechanisms are strong independent predictors
A prediction model achieved an AUC of 0.82 with a Brier score of 0.14 in the training cohort
The model showed moderate discriminative performance after internal validation with a Harrell C-index of 0.69
Traumatic subarachnoid hemorrhage was inversely associated with post-traumatic hydrocephalus occurrence

Resumen

Decompressive craniectomy (DC) is a life-saving procedure for patients with severe traumatic brain injury (TBI). However, post-traumatic hydrocephalus (PTH) remains a common and clinically significant complication that negatively impacts neurological recovery and often requires secondary cerebrospinal fluid diversion. Reliable tools for early postoperative risk stratification are still lacking. We conducted a single-center retrospective cohort study involving adult patients with severe traumatic brain injury (TBI) who underwent unilateral decompressive craniectomy (DC) between 2014 and 2023. Only variables available at admission or within the first seven postoperative days were considered as candidate predictors. Post-traumatic hydrocephalus (PTH) was diagnosed based on radiological progression, clinical deterioration, and the subsequent need for cerebrospinal fluid shunting. A multivariable Firth penalized logistic regression model was developed to identify independent predictors. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), Brier score, and internal validation with 1000 bootstrap resamples. Among 277 eligible patients, 74 (26.7%) developed post-traumatic hydrocephalus (PTH) during follow-up. Multivariable analysis revealed that intraventricular hemorrhage (IVH; adjusted odds ratio aOR = 7.64, 95% confidence interval CI 1.80–32.46, P = 0.006), high-energy injury mechanism (P < 0.001), fall from height (aOR = 4.46, P = 0.027), subdural hematoma (aOR = 26.23, P = 0.09), complex intracranial injury patterns (aOR = 51.26, P = 0.003), and longer time from injury to surgery (per hour aOR = 1.011, P = 0.039) were independently associated with an increased risk of PTH. Traumatic subarachnoid hemorrhage was inversely associated with PTH occurrence (aOR = 0.45, P = 0.043). The prediction model achieved an area under the curve (AUC) of 0.82 with a Brier score of 0.14 in the training cohort. After bootstrap internal validation, the optimism-corrected Harrell C-index was 0.69, suggesting moderate discriminative performance. A prediction model based exclusively on early perioperative variables can identify patients at high risk for post-traumatic hydrocephalus (PTH) following decompressive craniectomy. This tool may facilitate targeted surveillance and timely intervention; however, external validation in prospective, multicenter cohorts is necessary before routine clinical implementation.

Me gusta

Guardar

Ver artículo completo