Risk factors for pulmonary infection in patients with acute-on-chronic liver failure and establishment of a predictive model / 临床肝胆病杂志

ABSTRACT

ObjectiveTo investigate the risk factors for pulmonary infection in patients with acute-on-chronic liver failure （ACLF）， and to establish a predictive model. MethodsA retrospective analysis was performed for 585 ACLF patients who were admitted to Department of Infectious Diseases， The Second Affiliated Hospital of Air Force Medical University， from January 2009 to September 2022， and according to the condition of pulmonary infection after admission， they were divided into infection group with 213 patients and non-infection group with 372 patients. The independent-samples t test or the Mann-Whitney U test was used for comparison of continuous data between two groups， and the chi-square test was used for comparison of categorical data between groups. The clinical data of these patients were collected. Univariate and multivariate Logistic regression analyses were used to investigate the risk factors for pulmonary infection in ACLF patients and establish a predictive model， and the receiver operating characteristic （ROC） curve was plotted to assess the predictive value of the model. The Hosmer-Lemeshow test was used to evaluate the degree of fitting of the model， and the ROC curve and the area under the ROC curve （AUC） were used to assess the predictive performance of the model. ResultsAmong the 585 patients with ACLF， 213 experienced pulmonary infection， with an infection rate of 36.41%. The multivariate logistic analysis showed that upper gastrointestinal bleeding （odds ratio ［OR］=2.463， P=0.047）， infection at other sites （OR=2.218， P=0.004）， femoral vein catheterization （OR=2.520， P<0.001）， and combined use of two or more antibiotics （OR=2.969， P<0.001） were risk factors for pulmonary infection in ACLF patients. These factors were included in the risk factor predictive model of Logit （P）=-1.869+0.901×upper gastrointestinal bleeding+0.755×infection at other sites+0.924×femoral vein catheterization+1.088×combined use of two or more antibiotics. The ROC curve analysis showed that the model had a good predictive value （Hosmer-Lemeshow χ2=3.839， P=0.698）， with an AUC of 0.753 （95% confidence interval： 0.700 — 0.772）. ConclusionThere is a relatively high incidence rate of pulmonary infection in patients with ACLF， and upper gastrointestinal bleeding， spontaneous peritonitis， femoral vein catheterization， and combined use of two or more antibiotics are related risk factors. The model established based on these factors can effectively predict the onset of pulmonary infection in ACLF patients.