Baseline Neutrophil-To-Lymphocyte Ratio Versus NT-ProBNP for Predicting Mortality in Severe Pneumonia
Main Article Content
Abstract
Background: Severe pneumonia is associated with substantial short-term mortality, and early identification of high-risk patients remains essential, particularly in settings where rapid access to complex prognostic tools is limited. Objective: To compare the performance of baseline neutrophil-to-lymphocyte ratio and N-terminal pro-B-type natriuretic peptide for predicting all-cause 30-day mortality in adults hospitalized with severe pneumonia, and to evaluate whether their combined use improves prognostic accuracy. Methods: This prospective observational study was conducted at the Department of Medicine, Pak Emirates Military Hospital, Rawalpindi, Pakistan, from July 2023 to June 2024. A total of 130 adults meeting IDSA/ATS 2007 criteria for severe pneumonia were enrolled. Blood samples were obtained within six hours of admission for calculation of neutrophil-to-lymphocyte ratio and measurement of NT-proBNP. Receiver operating characteristic curve analysis, DeLong comparison, and binary logistic regression were performed. Results: Thirty-eight of 130 patients (29.2%) died within 30 days. Non-survivors had higher median neutrophil-to-lymphocyte ratio (19.4 vs 9.1; p<0.001) and NT-proBNP (4,820 vs 1,240 pg/mL; p<0.001). Neutrophil-to-lymphocyte ratio showed an AUC of 0.812 (95% CI 0.736-0.888), while NT-proBNP showed an AUC of 0.779 (95% CI 0.697-0.861), with no significant difference between them (p=0.38). The combined model improved AUC to 0.856 (95% CI 0.789-0.923). On multivariable analysis, both biomarkers independently predicted mortality. Conclusion: Both neutrophil-to-lymphocyte ratio and NT-proBNP are effective admission biomarkers for mortality risk stratification in severe pneumonia, with neutrophil-to-lymphocyte ratio offering particular practical value in resource-limited settings; combined assessment may further improve prognostic performance.
Article Details
Issue
Section

This work is licensed under a Creative Commons Attribution 4.0 International License.