Accurate risk stratification for development of organ/space surgical site infections after emergent trauma laparotomy

BACKGROUND: Organ/space surgical site infection (OS-SSI) develops frequently after trauma laparotomies and is associated with significant morbidity. No valid model exists to accurately risk-stratify the probability of OS-SSI development after emergent laparotomy. Risk stratification for OS-SSI in these patients could guide promising, but unproven, interventions for OS-SSI prevention, such as more frequent dosing of intraoperative antibiotics or direct peritoneal resuscitation. We hypothesize that in trauma patients who undergo emergent laparotomy, probability of OS-SSI can be accurately estimated using patient data available during the index operation. METHODS: Retrospective review was performed on a prospectively maintained database of emergent trauma laparotomies from 2011 to 2016. Patient demographics and risk factors for OS-SSI were collected. We performed Bayesian multilevel logistic regression to develop the model based on a 70% training sample. Evaluation of model fit using area under the curve was performed on a 30% test sample. The Bayesian approach allowed the model to address clustering of patients within physician while implementing regularization to improve predictive performance on test data. RESULTS: One hundred seventy-two (15%) of 1,171 patients who underwent laparotomy developed OS-SSI. Variables thought to affect development of surgical site infections and were available to the surgeon near the conclusion of the trauma laparotomy were included in the model. Two variables that contributed most to OS-SSIs were damage-control laparotomy and colon resection. The area under the curve of the predictive model validated on the test sample was 0.78 (95% confidence interval, 0.71-0.85). CONCLUSION: Using a combination of factors available to surgeons before the end of an emergent laparotomy, the probability of OS-SSI could be accurately estimated using this retrospective cohort. A Web-based calculator is under design to allow the real-time estimation of probability of OS-SSI intraoperatively. Prospective validation of its generalizability to other trauma cohorts and of its utility at the point of care is required. LEVEL OF EVIDENCE: Prognostic study, level IV.