Predicting refractory mycoplasma pneumoniae pneumonia in children using lung ultrasound and clinical molecu-lar features:A nomogram approach
Objective The early diagnosis of refractory mycoplasma pneumoniae pneumonia(RMPP)is chal-lenging due to the lack of practical diagnostic imaging tools.Pulmonary ultrasound(LUS)is an emerging tool for diagno-sing pneumonia in children.This study aimed to evaluate the role of a nomogram,combining LUS,clinical features,and laboratory markers,in predicting early RMPP in children.Methods A total of 300 children with mycoplasma pneumonia were prospectively enrolled.The cohort was divided into a modeling group(200 children)and a validation group(100 children)with a 2∶1 ratio.The modeling group included 159 cases of typical mycoplasma pneumonia(CMPP)and 41 ca-ses of RMPP,while the validation group comprised 81 cases of CMPP and 19 cases of RMPP.A binary logistic regression analysis was used to assess factors influencing the occurrence of RMPP.Significant variables identified by univariate analy-sis were incorporated into a multivariate logistic regression model to identify independent risk factors for RMPP.The resul-ting predictive model was converted into a visual nomogram.The model's accuracy was evaluated using receiver operating characteristic(ROC)curves,and its calibration was assessed using calibration curves.Results Independent risk factors for RMPP included age(OR=1.286,95％CI:1.031-1.639,P=0.031),duration of fever before admission(OR=1.630,95％CI:1.248-2.219,P＜0.001),high fever(OR=3.650,95％CI:0.889-18.230,P=0.089),erythro-cyte sedimentation rate(ESR)(OR=1.056,95％CI:1.027-1.090,P＜0.001),and bronchial hyperinflation(OR=104.526,95％CI:16.433-1 319.111,P＜0.001).The nomogram for predicting RMPP was visually represented,with an area under the ROC curve(AUC)of 0.943(95％CI:0.908-0.977)in the modeling group and 0.966(95％CI:0.937-0.997)in the validation group.The calibration curves were close to the diagonal line,suggesting good clinical applicability.Conclusion The inclusion of LUS in the nomogram provides a more comprehensive assessment of disease status,enabling more accurate early prediction of RMPP in children.Therefore,this predictive model could be widely ap-plied in the early diagnosis of RMPP in pediatric patients.
国家科技期刊平台
NSTL
万方数据
