BACKGROUND: This study assessed the diagnostic value of complete blood count (CBC) and derived immune-inflammatory indicators for Mycoplasma pneumoniae pneumonia (MPP) in children and developed a practical scoring system for early diagnosis. METHODS: A total of 1662 MP-infected children were divided into pneumonia (n = 844) and non-pneumonia (n = 818) groups. CBC parameters, derived immune-inflammatory indicators and C-reactive protein (CRP) were compared. Diagnostic performance was assessed using ROC curve analysis, and diagnostic models were constructed using multivariate logistic regression. Age-stratified analysis addressed confounding. Model stability was evaluated via repeated stratified 10-fold cross-validation, calibration curves with Brier score, and decision curve analysis (DCA). A simplified scoring system was established and validated for clinical application. RESULTS: Children in the pneumonia group were significantly younger (mean age 5.78 ± 3.41 years vs. 8.62 ± 4.92 years, p < 0.001). Lymphocyte count (LYM), lymphocyte percentage (LYM%), platelet count (PLT), plateletcrit (PCT), and the product of absolute lymphocyte count and platelet count (PLT × LYM) in the pneumonia group were significantly higher than in the non-pneumonia group. Age-stratified analysis confirmed these differences across all age subgroups (all p < 0.001). PLT × LYM, LYM%, PLT, and PCT were independent predictors. Internal validation showed a mean AUC of 0.7824 ± 0.0353, good calibration (Brier score = 0.1918 ± 0.0164), and positive net benefit on DCA. The scoring system achieved diagnostic rates of 71.5%, 51.2%, and 29.6% for high-, medium-, and low-risk groups, respectively. CONCLUSION: MPP predominantly affects children under 7 years old. The simple scoring system, supported by robust internal validation, effectively stratifies pneumonia risk and offers a practical early diagnostic tool for primary care.
Liu et al. (Sat,) studied this question.