Objective To analyze the clinical characteristics of children with severe Mycoplasma pneumoniae pneumonia (SMPP) treated with different doses of glucocorticoids (GC) and to investigate factors associated with GC treatment efficacy on the basis of propensity score matching (PSM).

Methods Clinical data from 271 children who were hospitalized at Tianjin Children's Hospital between January 1, 2021 and December 31, 2022, and who received GC treatment for SMPP were retrospectively collected. The patients were divided into low-dose (methylprednisolone ≤ 2 mg/kg·d) and high-dose (methylprednisolone > 2 mg/kg·d) groups. A 1∶1 PSM based on the matching variables, including time from onset to admission, time from fever onset to GC administration, and time from GC administration to follow-up chest X-ray, was performed to reduce confounding factors, resulting in 90 matched pairs in total in the low-dose and high-dose groups. Clinical characteristics were compared between the two groups after PSM. Multivariate logistic regression was performed to identify risk factors for high-dose GC use. The 90 pairs of children were further divided into an ineffective group (n = 38) and an effective group (n = 142) on the basis of clinical treatment outcomes, and multivariate logistic regression was conducted to determine risk factors for clinical GC treatment failure.

Results Children in the high-dose group were younger and had higher platelet counts and a higher incidence of atelectasis (all P < 0.05). The high-dose group showed a shorter time to fever resolution and overall duration of fever, but higher rates of adverse drug reactions (nausea, vomiting, abdominal pain, diarrhea, rash, etc.) compared with the low-dose group (all P < 0.05). No severe adverse events, such as hyperglycemia, hyperlipidemia, hypertension, gastrointestinal bleeding, or perforation, were observed in either group. Radiographic assessment showed a significantly higher rate of marked absorption in the high-dose group (P = 0.009). There were no significant differences between the groups in the length-of-stay, time to cough relief, or time to disappearance of pulmonary rales (all P > 0.05). Younger age, atelectasis, and elevated platelet count were identified as risk factors for selecting high-dose GC therapy using multivariate logistic regression analysis. Compared with the GC-effective group, the GC-ineffective group had a higher proportion of refractory Mycoplasma pneumoniae pneumonia (RMPP), a higher incidence of chest pain, and significantly increased levels of neutrophil-to-lymphocyte ratio, peak fever, procalcitonin, ferritin, lactate dehydrogenase, alanine aminotransferase, aspartate aminotransferase, and D-dimer (all P < 0.05), along with decreased lymphocyte counts (P < 0.05). There were no significant differences between the 2 groups in age or the proportion of children receiving high-dose GC (all P > 0.05). Multivariate analysis showed that RMPP and high peak fever were independent risk factors for GC treatment failure.

Conclusion High-dose GC therapy may facilitate fever resolution, shorten the duration of fever, and enhance radiological improvement. However, its overall efficacy is not significantly superior to low-dose therapy and is associated with a higher incidence of adverse reactions. Younger age, atelectasis, and elevated platelet count are key factors influencing the use of high-dose GC in children. Notably, RMPP and high peak fever are independent risk factors for GC treatment failure, irrespective of the GC dosage. Clinicians should carefully evaluate the individualized risk-benefit profile when considering high-dose GC therapy for pediatric SMPP.