Accuracy of the neutrophil-lymphocyte ratio in predicting thirty-day all-cause mortality among patients hospitalized with community-acquired pneumonia in Uganda a nested cross-sectional study

Abstract

Background: Community acquired pneumonia (CAP) contributes to approximately 3 million deaths globally. The assessment of severity plays a crucial role in effectively managing CAP and accurately predicting mortality outcomes. The neutrophil-lymphocyte ratio (NLR) has shown promise in assessing the risk of mortality for CAP patients, however, its role in predicting CAP mortality has not been extensively evaluated in Sub-Saharan Africa. This study sought to determine the accuracy of the NLR in predicting 30-day all-cause mortality among adults admitted with CAP to Mulago, Kiruddu, and Naguru Referral hospitals in Uganda. Methods: A nested cross-sectional study was conducted among consecutively selected adults with CAP admitted at Mulago, Kiruddu, and Naguru Referral Hospitals in Uganda, between April 2022 and January 2024. This study was nested in the LoRTISA study that seeks to determine the etiology and resistance patterns of clinically diagnosed CAP among hospitalized individuals, using a comprehensive microbiological approach. Data concerning the clinical history, physical examination, and laboratory tests were abstracted from the primary study Research Electronic Data Capture (REDCap) dataset, in February and March 2024. Data on the vital status of the participants at day 30 was also abstracted. Receiver operating characteristic (ROC) curve analysis was employed to determine the optimum cut-off point of the NLR for predicting mortality. Sensitivity, Specificity, Positive and Negative predictive values (PPV &NPV) of the optimum cut-off were then determined. Results: A total of 200 participants were enrolled. The median age and NLR were 41 years (IQR: 32-60years) and 4.0 (IQR:2-8) respectively. Majority of the participants were male (54%, 108/200). Mortality at day 30 was 17.5%. The optimum cut-off point of the NLR for predicting 30-day mortality was 4.23, with a corresponding sensitivity and specificity of 63% and 58%, respectively. The PPV was 24.2% and NPV was 88.1%. The area under the ROC curve (AUC) was 0.63. Conclusion: The NLR cut-off of 4.23 has a modest accuracy in predicting 30-day mortality and identifying high-risk CAP patients needing closer monitoring or treatment. Additional studies are needed to validate and refine NLR cut-off points through prospective studies, which will enhance its clinical utility in CAP management and improve patient outcomes.