D105-19 Co-infection Burden and Elevated CRP Independently Predict Pneumonia in Rhinovirus/enterovirus-positive Children

Importance Co-infections are common in pediatric acute respiratory illness, yet the clinical implications of specific pathogen combinations superimposed on rhinovirus/enterovirus (RV/EV) infection in children remain unclear. Objective To quantify the association of co-infection burden and specific co-pathogens with inflammatory C-reactive protein (CRP) levels and pneumonia among RV/EV-positive in children. Design, Setting, and Participants Retrospective cohort (January 2019-December 2024) of 11,926 emergency department encounters at a regional pediatric center using ICD-10. Viral detections were recorded in binary columns (1 = positive, 0 = negative). By design, analyses here focus on RV/EV-positive encounters (n = 4,732); co-infection was defined using additional pathogen flags (adenovirus, RSV, parainfluenza, influenza, human metapneumovirus [hMPV], common human coronaviruses, SARS-CoV-2, and Mycoplasma pneumoniae). Exposures (1) Co-infection burden among RV/EV positives: RV/EV alone, RV/EV + 1 other pathogen, RV/EV + 2+ pathogens; (2) Specific RV/EV pairs and triplets. Main Outcomes and Measures (a) First CRP (mg/L), (b) pneumonia (binary; 1 = present). Secondary covariates: age (months) and severity (ordinal: 1 mild/2 moderate/3 severe). Results 8,802/11,926 (73.8%) had ≥1 pathogen; 4,732 (39.7%) were RV/EV-positive; 1,418/11,926 (11.9%; 29.9% of RV/EV-positives) had RV/EV with ≥1 additional pathogen. Median CRP was higher with pneumonia across RV/EV alone, pairs, and triplets (all Mann-Whitney p < 3.0 × 10-5). In adjusted logistic models among RV/EV-positives (n = 944), compared with RV/EV alone, RV/EV + 1 other showed higher odds of pneumonia (OR 1.48, 95% CI 1.04-2.12; p = 0.030) and RV/EV + 2+ others was still higher (OR 3.38, 1.77-6.45; p = 2.25 × 10-4), independent of log(CRP), age, and severity. Log (CRP) was strongly associated with pneumonia (OR 1.60, 1.41-1.82; p < 10-12). In pair-only models (n = 231; ref = RV/EV+ adenovirus), RV/EV+ Mycoplasma pneumoniae (OR 44.51, 8.35-237.24; p ≈ 8.8 × 10-6) and RV/EV+RSV (OR 4.72, 1.79-12.43; p = 0.0017) were associated with higher pneumonia odds after adjustment. Triplet-specific effects were unstable due to sparse cells; ridge-penalized models shrank triplet pathogen coefficients toward null while retaining positive effects of log(CRP), age, and severity. Conclusions Among RV/EV-positive pediatric encounters, greater co-infection burden and certain co-pathogens—especially Mycoplasma pneumoniae and RSV—are associated with higher pneumonia risk, independent of CRP, age, and severity. CRP remains a robust severity biomarker across co-infection strata. These findings refine risk stratification for RV/EV-positive children when atypical bacteria or RSV co-detections are present. This abstract is funded by: None