Abstract / Summary
Abstract Background Epidemiological patterns of pediatric Mycoplasma pneumoniae pneumonia changed during the implementation and adjustment of COVID-19-related non-pharmaceutical interventions. Multicenter evidence on pathogen co-detection and clinical outcomes remains limited. We aimed to characterize its epidemiology and assess these associations. Methods This two-center retrospective study included hospitalized children aged 0–16 years with community-acquired pneumonia (CAP) from January 1, 2022 to December 31, 2025. M. pneumoniae positivity among CAP admissions was compared by year, sex, age, season, and intervention stage. Children with M. pneumoniae pneumonia (MPP) were classified as M. pneumoniae alone, with bacteria, with viruses, or with both bacteria and viruses. Length of stay was analyzed using a Gamma generalized linear model; pulmonary consolidation and pleural effusion were analyzed using multivariable logistic regression. Models adjusted for center, sex, age, season, and intervention stage. Results Among 28,067 children with CAP, 8,328 had MPP (29.7%). Annual positivity was 19.7%, 38.3%, 38.2%, and 12.0% from 2022 through 2025 ( P < 0.001). Compared with Mycoplasma pneumoniae alone, bacterial co-detection and bacterial-viral co-detection were associated with longer hospital stay, with adjusted mean ratios of 1.071 (95% confidence interval 1.047–1.094) and 1.064 (1.037–1.092), respectively (both P < 0.001); viral co-detection was not ( P = 0.173). Center-by-pattern interactions were significant, and associations with imaging outcomes were less consistent. Conclusions Pediatric Mycoplasma pneumoniae infection showed marked temporal, seasonal, and demographic variation. Bacterial-containing co-detection patterns were associated with modestly longer hospitalization in overall models, but effect estimates varied between centers, supporting cautious interpretation of co-detection as a clinical marker.