Post-decompressive craniectomy functional outcomes in severe traumatic brain injury: Independent predictors,...

Decompressive craniectomy (DC) is a key intervention for severe traumatic brain injury (sTBI), but predictors of long-term functional outcomes remain incompletely defined. This study aimed to identify independent predictors of 6-month functional outcome and develop a validated predictive model.

A retrospective cohort study was conducted on patients with sTBI admitted in our hospital between January 2010 and December 2024. Eligible patients were adults (≥18 years) with sTBI, defined as a Glasgow coma scale score ≤ 8 at admission, and all underwent standardized clinical and neuroimaging assessments upon admission. Patients were excluded if they had pre-existing severe neurological deficits, a history of prior cranial surgery, penetrating brain injury, or incomplete 6-month outcome data. Patients were categorized into 2 groups: those who underwent decompressive craniectomy (DC group) and those managed with conservative medical treatment (non-DC group). Baseline clinical variables, radiologic findings, and complications were collected. Functional outcomes at 6 months were assessed using the Glasgow outcome scale-extended (GOSE). Logistic regression analyses identified independent predictors of poor outcome (GOSE 1 - 3) and developed a predictive equation.

The study included 972 sTBI patients, 624 who underwent DC and 348 who received standard medical management. After 6 months, DC patients had a higher proportion of favorable outcomes (GOSE 4 - 8) than non-DC patients (63.8% vs. 43.4%, p < 0.001), with lower mortality (13.5% vs. 18.1%, p = 0.054). Univariate analysis identified age ≥ 60, Glasgow coma scale ≤ 8, pupillary abnormalities, midline shift > 1 cm, brain herniation, and post-traumatic hydrocephalus as predictors of poor outcome. Multivariate analysis confirmed these factors as independent predictors, while DC was protective (odd ratios (OR): 0.61, 95% confidence interval (CI): 0.42 - 0.89, p = 0.010). Post-traumatic hydrocephalus was strongly associated with poor outcome (OR: 2.18, 95% CI: 1.32 - 3.60, p = 0.002). The predictive model showed good discrimination (c-statistic 0.82) and calibration (Hosmer-Lemeshow p = 0.470).

DC significantly improves 6-month functional outcomes in sTBI. Independent predictors of poor outcome include older age, severe neurologic injury, pupillary abnormality, midline shift, brain herniation, and post-traumatic hydrocephalus. The validated model may guide early risk stratification, clinical decision-making, and management of complications to optimize recovery.