Timing of corticosteroid initiation and severe in-hospital outcomes in pediatric Mycoplasma pneumoniae pneumonia: a propensity score-matched cohort study.

Abstract

Background: Among hospitalized children with Mycoplasma pneumoniae pneumonia (MPP) who received systemic corticosteroids, the association between treatment timing and subsequent in-hospital outcomes remains uncertain. We evaluated whether earlier initiation was associated with a more favorable hospital course.

Methods: We conducted a retrospective propensity score-matched cohort study at a tertiary pediatric center in China between January 2023 and August 2025. Children aged 1-18 years with confirmed MPP who received systemic corticosteroids 3-14 days after symptom onset were included. Symptom onset was adjudicated by chart abstraction from caregiver-reported histories documented in the electronic medical record. All included patients received macrolide-based first-line antibiotics. Patients were grouped by corticosteroid initiation timing as early (3-7 days) or late (> 7 days). The propensity model used prespecified pre-treatment demographic, clinical-course, admission respiratory, inflammatory, radiographic, comorbidity, and resistance variables. The primary outcome was a severe composite endpoint; secondary outcomes included fever duration after corticosteroid initiation, hospital stay, oxygen therapy, ICU admission, pleural effusion, and antibiotic escalation. Exploratory analyses compared 3-5, 6-7, and > 7 days, with initiation day additionally modeled continuously for trend.

Results: Among 902 eligible patients, 482 received early corticosteroids and 420 received late corticosteroids. After 1:1 matching, 299 pairs were analyzed; 183 early initiators and 121 late initiators were unmatched and excluded from the matched cohort. Early initiation was associated with lower odds of the severe composite outcome than late initiation (12.0% vs. 19.4%; OR 0.57, 95% CI 0.36-0.89; P = 0.018), less oxygen therapy (8.7% vs. 13.7%; OR 0.59, 95% CI 0.35-0.99), less antibiotic escalation (9.4% vs. 15.7%; OR 0.56, 95% CI 0.34-0.91; P = 0.019), shorter fever after corticosteroid initiation (3 [IQR 2-4] vs. 4 [3-5] days), and shorter hospitalization (7 [6-9] vs. 9 [7-11] days; both P < 0.001). In exploratory timing analyses, only initiation on days 3-5 was associated with lower odds of severe outcomes (OR 0.35, 95% CI 0.18-0.68), whereas initiation on days 6-7 was not (OR 0.83, 95% CI 0.48-1.43). Continuous modeling suggested a graded association between later initiation day and worse outcomes (P for trend = 0.021).

Conclusion: Among corticosteroid-treated hospitalized children with MPP, earlier corticosteroid initiation-particularly on days 3-5 after symptom onset-was associated with a more favorable in-hospital course. Because residual confounding, potential collider bias, and time-dependent bias cannot be excluded, these findings should be interpreted as associative and require prospective confirmation.