Risk of Mycoplasma pneumoniae-related hepatitis in MP pneumonia pediatric patients: a predictive model construction and assessment

Abstract Background A predictive model for risk of Mycoplasma pneumoniae (MP)-related hepatitis in MP pneumonia pediatric patients can improve treatment selection and therapeutic effect. However, currently, no is available. Methods Three hundred seventy-four with/without serologically-confirmed infection ninety-three health controls were enrolled. Logistic regressions performed to identify the determinant variables develop model. Predictive performance optimal diagnostic threshold evaluated using area under receiver operating characteristic curve (AUROC). Stratification analysis by age MP-IgM titer was used optimize model’s clinical utility. An external validation set, including 84 patients, verify efficiency. After univariate screen significant variables, monocyte count (MO), erythrocyte distribution width (RDW) platelet (PLT) identified as independent predictors multivariate analysis. Results We constructed MRP model: MO [^10 9 /L] × 4 + RDW [%] – PLT 0.01. achieved an AUROC 0.754 sensitivity specificity at cut-off value 10.44 71.72 61.00 %, respectively predicting MP-related from pneumonia. These results verified whereas it merely 0.540 without infection. The 0.812 0.787 infants toddlers (0–36 months) low subgroup (1:160–1:320), respectively. It achieve 0.804 toddler with subgroup. Conclusions effective especially titer.