Characterizing Interhospital Variability in Neurosurgical Interventions for Patients with Mild Traumatic Brain Injury and Intracranial Hemorrhage

The objective of this study was to quantify nation-wide interhospital variation in neurosurgical intervention risk by intracranial hemorrhage (ICH) type in the setting of mild traumatic brain injury (mTBI). This was a retrospective cohort study of adult (≥18 years) trauma patients included in the National Trauma Data Bank from 2007 to 2019 with an emergency department Glasgow Coma Scale score 13-15, diagnosed ICH, no skull fracture. The primary outcome was neurosurgical intervention. Interhospital variation was assessed by examining the best linear unbiased predictors (BLUPs) obtained from mixed-effects logistic regression with random slopes and intercepts for hospitals and covariates for time and 14 demographic, injury, and hospital characteristics; one model per ICH type. Intercept BLUPs are estimates of how different each hospital is from the average hospital (after covariate adjustment). The study population included 49,220 (7%) neurosurgical interventions among 666,842 patients in 1060 hospitals. In 2019, after adjusting for patient case-mix and hospital characteristics, the percentage of hospitals with hemorrhage-specific neurosurgical intervention risk significantly different from the average hospital was as follows: isolated unspecified hemorrhage (0% of 995 hospitals); isolated contusion/laceration (0.54% of 929); isolated epidural hemorrhage (0.39% of 778); isolated subarachnoid hemorrhage (0.10% of 1002); multiple hemorrhages (2.49% of 963); and isolated subdural hemorrhage (16.25% of 1028). In the setting of mTBI, isolated subdural hemorrhages were the only ICH type to have considerable interhospital variability. Causes for this significant variation should be elucidated and might include changing hemorrhage characteristics and practice patterns over time.