Global Mycoplasma pneumoniae pneumonia (MPP) outbreaks show increasing refractory cases. Conventional diagnostics lack sensitivity, with limited evidence on factors affecting M. pneumoniae load in mNGS, co-detected fungi, and antibiotic optimization.

To investigate the pivotal role of mNGS in optimizing the diagnosis and treatment of MPP.

We conducted a retrospective analysis of 305 patients with confirmed MPP. Among them, 74 patients (64 serum IgM-negative, 15 PCR-negative, 5 both negative) tested negative by traditional methods but were positive for M. pneumoniae via mNGS. Spearman correlation analysis evaluated associations between M. pneumoniae sequence counts (SC) and inflammation markers. Logistic regression identified risk factors for high SC. The spectrum of fungal co-detections was characterized through frequency statistics and association rule mining (Apriori algorithm). The association between antibiotic adjustment and clinical outcomes was assessed using the chi-square test.

The mean age of the cohort was 27.92 ± 24.06 years. Common symptoms included cough (85.2 %) and fever (69.2 %). The median M. pneumoniae SC was 4.8 × 10⁴ ( 2.3 ×10 ³-2.6 ×10⁵). SC was significantly higher in the adolescent group (≤18 years) compared to the young adult group (18–45 years) (p = 0.007). Independent risk factors for high SC included fever (OR = 1.372 per 0.5°C increase), elevated serum lactate dehydrogenase level (OR = 1.005 per U/L increase), and decreased neutrophil percentage (OR = 0.969 per % decrease) (all p < 0.05). Fungal co-detection was identified in 23.9 % (73/305) of patients; predominant pathogens were Candida albicans (11.80 %), Aspergillus flavus (8.52 %), and Aspergillus fumigatus (8.52 %). The strongest co-occurrence association was observed between A. fumigatus and A. flavus (lift = 4.72, p < 0.001). Antibiotic adjustment was significantly associated with improved treatment effectiveness (OR = 5.33, p < 0.001).

Adolescents exhibit higher M. pneumoniae pathogen loads than young adults. Serum LDH levels and fever demonstrate positive correlations with pathogen load. A synergistic co-detection pattern between A. fumigatus and A. flavus was identified. mNGS, by integrating pathogen load quantification and co-detection profiling, effectively guides MPP diagnosis and antibiotic therapy, delivering pivotal evidence for precision treatment.