Comparative Effectiveness of Early Neurosurgical Intervention in Civilian Penetrating Brain Injury Management.

Abstract

To compare the outcomes of early vs no-neurosurgical intervention in civilians with penetrating brain injury (PBI). We collected data from the National Trauma Data Bank for PBI between 2017 and 2019. A total of 10 607 cases were identified; 1276 cases met the following criteria: age 16-60 years, an intensive care unit (ICU) length of stay (LOS) of >2 days, a Glasgow Coma Scale of 3-12, and at least one reactive pupil on presentation. Patients with withdrawal of life-sustaining treatments within 72 hours were excluded, leaving 1231 patients for analysis. Neurosurgical intervention was defined as an open-approach cranial procedure involving release, drainage, or extirpation of brain matter performed within 24 hours. Outcomes of interest were mortality, withdrawal of life-sustaining treatments, ICU LOS, and dispositional outcome. The target population was 1231 patients (84.4% male; median [IQR] age, 29 [18] years); 267 (21.7%) died, and 364 (29.6%) had a neurosurgical intervention within the first 24 hours. 1:1 matching yielded 704 patients (352 in each arm). In the matched cohort (mortality 22.6%), 64 patients who received surgery (18.2%) died compared with 95 (27%) in the nonsurgical group. Survival was more likely in the surgical group (odds ratio [OR] 1.66, CI 1.16-2.38, P < .01; number needed to treat 11). Dispositional outcome was not different. Overlap propensity score-weighted analysis (1231 patients) resulted in higher odds of survival in the surgical group (OR 1.8, CI 1.16-2.80, P < .01). The E-value for the OR calculated from the matched data set was 2.83. Early neurosurgical intervention was associated with longer ICU LOS (median 12 days [7.0, 19.0 IQR] vs 8 days [4.0, 15.0 IQR], P < .05). Management including early neurosurgical intervention is associated with decreased mortality and increased ICU LOS in matched cohorts of PBI.