Interaction Between Admission Hemoglobin and Sex on ICU Mortality in Patients With Moderate-to-Severe Traumatic Brain Injury: A Retrospective Bicentric Cohort Study.

Abstract

Recent studies have sought to define an optimal transfusion threshold in patients with traumatic brain injury (TBI). However, baseline hemoglobin levels differ by sex. Given the male predominance in TBI epidemiology, it is essential to determine if this sex-specific physiologic difference influences the prognosis. Bicentric, retrospective cohort study. Two level 1 trauma centers in France from November 2010 to July 2025. Adult critically ill patients with moderate-to-severe TBI. None. We included 5513 TBI patients. After adjustment, every 1 g/dL increase in admission hemoglobin was associated with a 31% decrease in the odds of ICU mortality (odds ratio [OR], 0.69; 97.5% CI, 0.52-0.91; p = 0.0027). The sex × hemoglobin interaction effect was not significantly associated with ICU mortality (OR, 1.04; 97.5% CI, 0.91-1.19; p = 0.433). Predicted ICU mortality risk was lower in females across all hemoglobin levels (Wald tests, all p < 0.05 from 9.0 to 13.0 g/dL). However, causal inference methods found no evidence of a sex-specific difference in standardized ICU mortality risk at equivalent admission hemoglobin levels (adjusted OR, 1.01; 95% CI, 0.85-1.19; p = 0.936). No interaction between admission hemoglobin and sex was found regarding ICU mortality following TBI in this cohort. These results do not support the implementation of sex-differentiated hemoglobin transfusion thresholds based solely at admission hemoglobin levels in the initial management of TBI patients. Further prospective studies are needed to confirm these findings and to evaluate whether sex-related differences may emerge when considering longitudinal hemoglobin trajectories during the ICU stay and transfusion practices.