AUTHOR=Li Xiaoliang , Lei Xiaoli , Kou Lijie , Wang Junli , Yang Zhigang 

TITLE=Development and validation of mortality risk prediction model for sepsis secondary to pneumonia at intensive care unit admission: a retrospective case-cohort study

JOURNAL=Frontiers in Medicine

VOLUME=Volume 12 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2025.1592325

DOI=10.3389/fmed.2025.1592325

ISSN=2296-858X

ABSTRACT=BackgroundTo establish a noninvasive mortality risk prediction model for sepsis secondary to pneumonia (SSP) and validate the model in the prediction of mortality risk in SSP patients at hospital admission.MethodsA retrospective cohort of SSP patients were recruited from January 2017 to December 2020 at the Henan Provincial People’s Hospital. Clinical data were collected at admission. Least absolute shrinkage and selection operator and logistic regression were used to construct a prognosis prediction model. The predictive performance of the model was evaluated by receiver operating characteristic (ROC) curve. Another retrospective cohort with SSP was recruited from January 2021 to July 2022 at the same hospital to validate the model.ResultsA total of 1,337 patients were screened, including 941 patients in the derivation cohort and 396 patients in the validation cohort. The model included age, white blood cell count, neutrophil-to-lymphocyte ratio, lactate dehydrogenase, arterial oxygen pressure / fraction inspired oxygen, D-dimer and vasoactive drug use. The area under ROC curve of derivation model was better than sequential organ failure assessment score and APACHE II score (0.777 vs. 0.600 vs. 0.625, p < 0.05). Besides, the proposed model had a significantly higher prediction performance than SOFA and APACHE II scores in the validation cohort (0.803 vs. 0.655 vs. 0.688, p < 0.05). The prediction model was publicly released as an online calculator.ConclusionA prognosis model based on variables of SSP patients at hospital admission was developed.