(1723) Gastrointestinal Bleeding in Trauma ICU Patients

Introduction: Gastrointestinal bleeding (GiB) is a prevalent and potentially lethal problem associated with hypoperfusion, cytokine release, and alterations to the mucosal barrier frequently seen in the critical care population. Risk factors in the population at large have been well-studied but few studies have specifically addressed the unique circumstances surrounding critical trauma patients. We aimed to evaluate incidence and risk factors for GiB in the trauma critical care population.

Methods: A retrospective analysis of patients admitted to the trauma intensive care unit at a level 1 trauma center between 3/2019-7/2023 was performed. Patients mechanically ventilated for > 48h were included. Patients with GiB were then matched (1:3) by age to control patients for case-control analysis. Demographics, management, and outcomes were compared. Multivariable logistic regression (MLR) was performed to identify predictors of GiB.

Results: 2289 patients were reviewed with 64 GiB. After matching, 256 patients were included in the study. An additional 4 GiBs were confirmed upon further chart review, with GiB incidence 3%. The overall population was male (77%), had a median age of 41, an injury severity score of 22, and mortality of 21%. Of 68 patients with GiB, 51 (75%) were clinically significant with increased pressors, systolic BP decrease of >20mmHg, 2 g/dl decrease in hemoglobin, or greater than a 2 unit pRBC transfusion requirement. Patients with a GiB were more likely male, suffered penetrating mechanism, or were coagulopathic. There was no difference in proton pump inhibitor use between cohorts. GiB patients had more ventilator days, ICU days, and had a longer length of stay. MLR identified male sex (AOR 4.11,95%CI 1.49-11.31,p=.006), coagulopathy (AOR 2.10,95%CI 1.03-4.27,p=.040), pressor use (AOR 3.22,95%CI 1.48-6.94, p=.003), renal replacement therapy (AOR 3.43,95%CI 1.37-8.50, p=.008), and intolerance to enteral nutrition (AOR 3.43,95%CI 1.42, 8.26, p=0.006) as predictors for GiB. Median time to GiB diagnosis was 14 days.

Conclusions: GiB remains a significant problem in critically ill populations. Identifying risk factors unique to the critical trauma patient may lead to earlier identification of susceptible patients and allow for more robust preventative measures to reduce incidence.