Mass Effect Score Associations in Comatose Trauma Patients With Intracranial Hemorrhage Requiring Mechanical Ventilation: A Retrospective Study.

Objective There is a paucity of studies evaluating computed tomography (CT) scores in patients with Glasgow Coma Scale (GCS) scores of 3-8. This study assessed the associations between risk factors (including admission GCS score, head Abbreviated Injury Scale (AIS) score, brain mass effect score (MES), and age) and unfavorable outcomes (including in-hospital mortality, adverse outcomes, failure to follow commands at hospital discharge, and failure to follow commands at three months) in mechanically ventilated patients with intracranial hemorrhage (ICH) and GCS scores of 3-8. In relation to these outcomes, we also compared the MES with the Marshall (MAR) classification and Rotterdam (ROT) score. Methods The adult blunt trauma cohort had ICH, GCS 3-8, need for mechanical ventilation, and level I trauma center admission (2018-2024). The GCS, age, AIS, ventilator status, intensive care unit (ICU) stay, hospital stay, and hospital mortality were obtained from the trauma registry. Adverse outcomes included hospital death, ICU stay ≥18 days, or hospital stay >20 days. MES indicated the presence (1) or absence (0) for each of the following: basal cistern compression, midline shift, and lateral ventricle compression (0-3). Brain CT images were reviewed to compute MES, MAR, and ROT. Electronic medical records (EMRs) were reviewed to determine following command status at hospital discharge and at three months after injury. Results Among the 148 blunt trauma patients, univariate associations with mortality were MES (p < 0.01), GCS (p = 0.46), age ≥70 (p < 0.01), and AIS (p < 0.01); multivariate associations were MES (p < 0.01), age ≥70 (p < 0.01), and AIS (p > 0.10). For adverse outcomes, univariate associations were MES (p < 0.01), GCS (p = 0.46), age ≥70 (p = 0.04), and AIS (p < 0.01); multivariate associations were MES (p < 0.01), age ≥70 (p > 0.05), and AIS (p > 0.05). For failure to follow commands at discharge, univariate associations were MES (p < 0.01), GCS (p = 0.75), age (p = 0.01), and AIS (p < 0.01); multivariate associations were MES (p < 0.01), age (p = 0.01), and AIS (p > 0.10). For failure to follow commands at three months, univariate associations were MES (p < 0.01), GCS (p = 0.30), age (p < 0.01), and AIS (p < 0.01); multivariate associations were MES (p < 0.01), age (p < 0.01), and AIS (p > 0.10). MES correlated strongly with MAR (p < 0.01; r = 0.83) and ROT (p < 0.01; r = 0.82). In multivariate stepwise continuous score regression (MCR), mortality and adverse outcome were associated with MES (both p < 0.01), but not MAR or ROT (both p > 0.05). For failure to follow commands at discharge and three months, MCR showed an association with ROT (both p < 0.01), but not MES or MAR (both p > 0.05). In multivariate stepwise dichotomous score regression (MDR), mortality and adverse outcomes were associated with MES ≥2 (both p < 0.01), but not MAR ≥4 or ROT ≥4 (both p > 0.05). Failure to follow commands at discharge was associated with ROT ≥4 (p < 0.01), but not MES ≥2 or MAR ≥4 (p > 0.05). Failure to follow commands at three months was associated with MES ≥2 (p < 0.01), but not MAR ≥4 or ROT ≥4 (p > 0.05). Conclusions MES was a better predictor of adverse outcomes than AIS and GCS in comatose patients with ICH. MES was competitive with other established brain CT scores for predicting adverse outcomes. Further studies are needed to confirm these findings.