Prognostic Significance of Concomitant Pulmonary Insult in a Large Cohort of Operative Traumatic Brain Injury Patients

Introduction: A growing body of literature has suggested that traumatic-brain-injury (TBI) is associated with an increased risk for developing pulmonary injury. However, there is limited literature exploring the clinical characteristics and prognostic significance of preoperative pulmonary insults. The aim of this study was to characterize the impact of preoperative pulmonary injury on long-term post-TBI recovery.

Methods: This single-institutional-retrospective study examined a large cohort of 139 patients with, and 160 patients without preoperative pulmonary injury who had undergone surgery between 2007 and 2020. To assess recovery from TBI, Modified Rankin Scale (mRS) and Glasgow Outcome Scale-Extended (GOS-E) scores were calculated at discharge and the latest follow-up.

Results: Among those with preoperative pulmonary insult, 118 (84.9%) patients presented with structural abnormalities (effusion, nodules, atelectasis, and trauma-induced changes), 7 (5%) had infections, and 14 (10.1%) had preoperative respiratory failure. There were no differences in rates of preoperative comorbidities or sociodemographic factors. Poorer preoperative mRS, GCS, and injury-severity scores were observed in those with lung injuries (p=0.04, p< 0.001, and p< 0.001 respectively). These patients tended to present with higher rates of intraparenchymal-hemorrhage (p=0.05) and involvement in motor vehicle accidents (p=0.01). Interestingly rates of serum sodium (p < 0.001) and preoperative hyperosmolar therapy (28.1% vs 17%, p=0.04) were higher among those with lung injuries. Among operation type, craniectomies were more common in this group (23.7% vs 44%, p< 0.001). Postoperatively and at latest follow-up, lower mRS and GOSE scores were observed in patients with lung injuries preoperatively (p < 0.001 for all), alongside longer ICU (p=0.002) and hospital length of stays (p < 0.001). A multivariable logistic-regression model demonstrated that age (OR: 1.02; 95% CI: 1.01-1.04) and severe (3-8) GCS scores (OR 6.6; 95% CI: 3.2-13.4) were factors predictive of poor (≤2 GOS-E) outcomes at the latest follow up.

Conclusion : Preoperative lung injuries are associated with worse outcomes postoperatively as measured by mRS and GOS-E.