Association Between Caffeine and Invasive Mechanical Ventilation in Bronchiolitis-Related Apnea

OBJECTIVES Caffeine has been shown to reduce apnea of prematurity and prostaglandin-mediated apnea, but prior studies for bronchiolitis-related apnea have been inconclusive. We aimed to describe the epidemiology of bronchiolitis-related apnea and determine the association between caffeine prescription and ventilation across two large databases. METHODS This retrospective cohort study of Epic Cosmos and Pediatric Health Information Systems (PHIS) databases included inpatient admissions for children ≤ 6 months-old with bronchiolitis between 1/1/2015-12/31/2024. Admissions were 1:1 nearest-neighbor propensity matched for caffeine prescription. The marginal risk ratio for subsequent invasive mechanical ventilation was reported. RESULTS Five and a half percent (10,843/195,607) of admissions in Cosmos and 3.0% (5,429/175,945) in PHIS had apnea. Apnea decreased with age; 12.9% (3,469/26,860) and 10.4% (2,732/26,344) admissions for infants aged <1 month had apnea in Cosmos and PHIS, respectively. Of those with apnea 10.2% (1,107/10,843) and 9.2% (498/5,429) received caffeine in Cosmos and PHIS, respectively. Admissions that received caffeine were younger, more premature, had a higher proportion of cardiac conditions, and were more likely to be admitted to an intensive care unit in both databases. After matching, those receiving caffeine were less likely to undergo invasive mechanical ventilation (marginal risk ratio 0.580 [95% CI 0.500, 0.672] in Cosmos and 0.802 [95% CI 0.685, 0.938] in PHIS). The results were robust in sensitivity analyses excluding patients with cardiac disease or prematurity. CONCLUSIONS Caffeine administration was associated with decreased invasive mechanical ventilation use in this propensity-matched database analysis. Multicenter prospective trials are needed to confirm this association.