Neurosurg Rev. 2026 Jul 20;49(1):484. doi: 10.1007/s10143-026-04405-8.
ABSTRACT
Chronic subdural hematoma (CSDH) is increasingly common in older adults and in patients receiving antithrombotic therapy. Although burr-hole drainage is generally safe and effective, perioperative mortality remains a concern, and reliable preoperative predictors are incompletely defined. We aimed to identify independent preoperative predictors of in-hospital mortality after burr-hole drainage for CSDH, with explicit characterization of causes of death, comorbidity burden, and the discriminative performance of candidate predictors. This single-center retrospective cohort study included 121 consecutive adult patients surgically treated for CSDH between January 2015 and December 2024. Preoperative variables included demographics, Glasgow Coma Scale (GCS) score, hematoma thickness, midline shift (MLS), cerebral edema on CT, antithrombotic use, and a comprehensive set of comorbidities (chronic kidney disease, chronic obstructive pulmonary disease, congestive heart failure, coronary artery disease, atrial fibrillation, diabetes mellitus, dementia, malignancy). Intensive care unit admission and postoperative complications were also recorded. The primary outcome was in-hospital mortality. Causes of death were systematically categorized. Associations were evaluated using Firth penalized logistic regression. Discriminative performance was assessed using receiver operating characteristic (ROC) analysis with bootstrap confidence intervals. A prespecified exploratory interaction analysis between cerebral edema and significant midline shift was performed. In-hospital mortality was 12.4% (15/121). The principal cause of death was cerebral herniation (9/15, 60.0%); extracranial complications (respiratory, septic, cardiac) accounted for 6/15 (40.0%). In the original multivariable Firth model, GCS ≤ 13 (adjusted OR 11.71, 95% CI 2.82-48.61, p < 0.001) and cerebral edema (adjusted OR 8.90, 95% CI 1.93-40.98, p = 0.005) were independently associated with mortality. In an extended model incorporating comorbidities, chronic kidney disease (OR 23.78, p = 0.025) and congestive heart failure (OR 42.39, p = 0.043) emerged as additional independent predictors, while cerebral edema (OR 61.89, p = 0.008) and GCS ≤ 13 (OR 5.80, p = 0.032) retained their associations. Combined model discrimination was excellent (Model 1: AUC 0.920; Model 2 with comorbidities: AUC 0.958). Subgroup analysis demonstrated a marked mortality gradient: 0% in patients with neither cerebral edema nor significant midline shift (n = 59), versus 72.2% in patients with both findings (n = 18). The exploratory interaction term (cerebral edema × midline shift ≥ 5 mm) was directionally consistent with synergy but did not reach statistical significance (OR 39.19, 95% CI 0.26-∞, p = 0.152), reflecting limited statistical power. Preoperative neurological impairment and cerebral edema are independently associated with in-hospital mortality after burr-hole drainage for CSDH. The coexistence of cerebral edema and significant midline shift identifies a clinically recognizable high-risk phenotype that may warrant heightened perioperative attention. Renal and cardiac comorbidities further contribute to mortality risk. These findings are hypothesis-generating and require validation in larger, prospective cohorts before incorporation into clinical risk stratification.Clinical trial number: not applicable.
PMID:42472980 | DOI:10.1007/s10143-026-04405-8
