Author:
Lin Qingwei,Peng Enlan,Deng Xingping,Song Xiaomin,Zhong Lincui,He Longping,Zeng Qingbo,Song Jingchun
Abstract
ObjectiveTo investigate the pivotal determinants contributing to the adverse prognosis in patients afflicted with traumatic lung injury (TLI), with an aim to mitigate the elevated mortality rate associated with this condition.MethodsA retrospective analysis was carried out on 106 TLI patients who were admitted to the intensive care unit of a comprehensive hospital from March 2018 to November 2022. The patients were categorized into two groups based on their 28-day outcome: the survival group (n = 88) and the death group (n = 18). Random forest model, least absolute shrinkage and selection operator (LASSO) regression and support vector machine recursive feature elimination (SVM-RFE) were utilized to pinpoint the primary factors linked to poor prognosis in TLI patients. The Receiver Operating Characteristic (ROC) curve analysis was utilized to ascertain the predictive value of INR in forecasting the prognosis of TLI patients. Based on the cut-off value of INR, patients were categorized into two groups: INR ≥ 1.36 group (n = 35) and INR < 1.36 group (n = 71). The 28-day survival rate was then compared using Kaplan–Meier analysis.ResultsRandom forest model, LASSO, and SVM-RFE jointly identified International standardization ratio (INR) as a risk factor for TLI patients. The area under the ROC curve for INR in predicting the 28-day mortality of TLI patients was 0.826 (95% CI 0.733–0.938), with a cut-off value of 1.36. The 28-day mortality risk for TLI patients with an INR ≥ 1.36 was 8.5 times higher than those with an INR < 1.36.ConclusionTraumatic lung injury patients with elevated INR have a poor prognosis. An INR of ≥1.36 can be used as an early warning indicator for patients with traumatic lung injury.