Résumé
Objective:To construct a prediction model for prognosis of severe pneumonia patients combined with sepsis.Methods:Clinical data of 318 severe pneumonia patients combined with sepsis admitted at Taizhou People’s Hospital affiliated to Nanjing Medical University from March 2019 to March 2022 were retrospectively analyzed. Patients were randomized into a modeling set ( n=233) and a validation set ( n=85) with a 3∶1 ratio. In the modeling set there were 180 survival cases and 53 fatal cases according to the clinical outcomes within 30 days of admission. Multivariate Cox regression analysis was used to evaluate the independent prognostic factors for patients in the modeling set. A nomogram prediction model was constructed by R based on these prognostic factors and further verified using the data of the validation set with receiver operating curve (ROC), decision curve analysis (DCA), and calibrated with calibration curve analyses. Results:Multivariate Cox regression analysis suggested that septic shock ( HR=2.32, 95% CI 1.37-3.89, P=0.013) and neutrophil/lymphocyte ratio (NLR) ( HR=2.52, 95% CI 1.23-5.61, P=0.017) were independent risk factors for mortality in severe pneumonia patients combined with sepsis within 30 days of admission, while albumin/fibrinogen ratio (AFR) ( HR=0.64, 95% CI 0.41-0.83, P=0.011) and prognostic nutritional index (PNI) ( HR=0.68, 95% CI 0.57-0.83, P=0.009) were independent protective factors. The area under ROC curve (AUC) of the nomogram model based on these four indicators in the modeling and validation sets were 0.875 and 0.880, respectively. The DCA curve analysis indicated that the clinical benefit of this model was better than "All" or "None" curves in both the modeling and verification sets.The calibrate curve analysis indicated that the actual and corrected curves fitted well and were close to the ideal curve. Conclusion:The constructed nomogram model based on septic shock, AFR, NLR and PNI has a well prognostic value in severe pneumonia patients combined with sepsis.