Nomogram Prediction Model of the Severity of CAP Patients Based on Blood Indicators.

Abstract

Community-acquired pneumonia (CAP) is a common cause of hospitalization, characterized by high mortality, morbidity, and cost and has serious human health implications. Various scoring criteria have been used to predict the severity of pneumonia, including the CURB-65 score, Pneumonia Severity Index (PSI) score, and Severe Community-Acquired Pneumonia (SCAP) score. However, these scoring criteria have shortcomings such as low sensitivity, cumbersome clinical application, and limited application. This study aimed to construct a nomogram model to predict the severity of CAP patients by blood indicators. This is a retrospective study. Patients with CAP were enrolled and then tested by routine blood, coagulation series, biochemical and inflammatory indicators, and general information such as imaging findings and disease history of the patients were recorded. The main observation was the progression of the patients' disease. Univariate analysis and binary logistic regression analysis were used to explore the independent risk factors for the severity of CAP patients, followed by plotting a nomogram to obtain the prediction model and constructing calibration curves to assess the authenticity and accuracy of the prediction model. Finally, the sensitivity, specificity, and other evaluation indexes were calculated by the receiver operating characteristic curve (ROC) to evaluate the clinical application value of the nomogram prediction model. Univariate analysis and binary logistic regression analysis of 277 hospitalized patients yielded platelet to lymphocyte ratio (PLR) and serum amyloid A (SAA) as independent risk factors for the severity of CAP patients. The AUC of the nomogram model for PLR and SAA was 0.889 (95% CI 0.845 - 0.932). The sensitivity was 77.3%, and the specificity was 85.3%, which had an excellent predictive value. The nomogram model based on PLR and SAA to predict the severity of CAP patients has a high specificity and sensitivity.