Development and validation of a novel nomogram to predict severe adenovirus pneumonia in children with adenovirus pneumonia

Abstract

Background Severe adenovirus pneumonia (SAP) in children is characterized by multiple complications in the respiratory system, a high mortality rate, and a long recovery period. The study aimed to develop and validate a nomogram for predicting SAP in patients with adenovirus pneumonia (ADVP).Methods In this study, a total of 202 children with ADVP were collected between January 2019 and December 2020. Demographic and clinical characteristics of patients who participated in this study were utilized to develop a nomogram for predicting SAP. The data were categorized as training and validation datasets using random split sampling (split ratio = 7:3). Univariate logistic regression was used to select predictors. Multivariate logistic regression analysis was applied to construct a predictive model by introducing the predictors. The nomogram was visually developed on the basis of the selected predictors. The discriminatory ability of the model was determined using the receiver operating characteristic curve. Moreover, the prediction accuracy was evaluated using a calibration curve, and clinical effectiveness was evaluated by decision curve analysis (DCA).Results Univariate and multivariate logistic regression demonstrated that the duration of fever (OR: 1.500, 95% Cl: 1.261–1.783), atelectasis (OR: 12.581, 95% Cl: 1.323–119.615), L% (OR: 0.938, 95% Cl: 0.905–0.972), and FER (OR: 1.006, 95% Cl: 1.002–1.010) were independent predictors of SAP in patients with ADVP. The nomogram exhibited good discrimination with area under the curve (AUC) in the training dataset (0.860, 95%Cl: 0.800–0.920) and validation dataset (0.818, 95% Cl: 0.690–0.947). Through the calibration plot and Hosmer–Lemeshow test, the predicted probability was consistent with the actual probability in the training dataset (P = 0.545) and validation dataset (P = 0.545), and DCA showed good clinical utility.Conclusions In this study, a nomogram for predicting SAP among ADVP was developed and validated. It also showed good performance, indicating its discrimination ability, calibration ability, and clinical value. Thus, it may be used for the early identification of SAP, which will help physicians take timely intervention and appropriate management.