Surgical Resident Brookdale University Hospital Medical Center Brooklyn, New York, United States
Introduction: Hospital acquired infection (HAI) after spinal tumor resection surgery contributes to adverse patient outcomes and excess healthcare resource utilization. This study sought to develop a predictive model for HAI occurrence following spinal tumor resection.
Methods: The National Surgical Quality Improvement Program (NSQIP) 2015-2019 database was queried for spinal tumor resections. Baseline demographics and pre-operative clinical characteristics, including frailty, were analyzed. Frailty was measured by modified frailty score 5 (mFI-5) and risk analysis index (RAI). Univariate and multivariable analyses were performed to identify independent risk factors for HAI occurrence. A logit-based predictive model for HAI occurrence was designed and discriminative power assessed via receiver operating characteristic (ROC) analysis.
Results: Of 5,883 patients undergoing spinal tumor surgery, HAI occurred in 574 (9.8%). The HAI (vs. non-HAI) cohort was older and frailer with higher rates of pre-operative functional dependence, chronic steroid use, chronic lung disease, coagulopathy, diabetes, hypertension, tobacco smoking, unintentional weight loss, and hypoalbuminemia (all P< 0.05). In multivariable analysis, independent predictors of HAI occurrence included severe frailty (mFI-5, OR: 2.3, 95% CI: 1.1-5.2, P=0.035), non-elective surgery (OR: 1.7, 95% CI: 1.1-2.4, P=0.007), and hypoalbuminemia (OR: 1.5, 95% CI: 1.1-2.2, P=0.027). A logistic regression model with frailty score alongside age, race, BMI, elective vs. non-elective surgery, and pre-operative labs robustly predicted HAI occurrence with a C-statistic of 0.68 (95% CI: 0.64-0.72).
Conclusion : HAI occurrence after spinal tumor surgery can be robustly predicted by standardized frailty metrics, mFI-5 and RAI-A, alongside routinely measured pre-operative characteristics (demographics, comorbidities, pre-operative labs).
