683 - Risk of Probiotic-Associated Invasive Infection in a Large Cohort of Very Preterm Infants in the United States

Publication Number: 683.6363

Ravi M. Patel, Emory University & Children's Healthcare of Atlanta, Atlanta, GA, United States; Monica M. Bennett, Baylor Scott & White Research Institute, Dallas, TX 75204, TX, United States; Jennifer M. Canvasser, Necrotizing Enterocolitis (NEC) Society, Davis, CA, United States; Rachel G. Greenberg, Duke Clinical Research Institute, Durham, NC, United States; Kaashif A. Ahmad, Pediatrix Medical Group, Houston, TX, United States; Veeral Tolia, Pediatrix, Dallas, TX, United States

Background: In 2023, a warning was issued from the U.S. Food and Drug Administration regarding the risk of invasive disease in preterm infants given probiotics. Within the context of potential benefits of probiotics supplementation, the absolute risk of probiotic-associated invasive infections is uncertain. Large cohorts are needed to quantify uncommon risks.

Objective: We characterized invasive infection with Bifidobacterium or Lactobacillus, two types of bacteria commonly contained in probiotics used in neonatal intensive care units, in a large cohort of very preterm infants.

Design/Methods: We included infants 22-32 weeks gestation who were discharged from 1/1/2016-12/31/2023 in the Pediatrix Clinical Data Warehouse, which comprises U.S. neonatal units. Infants were followed until discharge, death, or transfer to another hospital. We defined invasive infection as identification of either Lactobacillus or Bifidobacterium in a blood culture >72 hours of age. Infections and mortality were compared among probiotic supplemented and unsupplemented infants, given these bacteria may normally colonize unsupplemented infants, using risk differences with 95% confidence intervals. We conducted sensitivity analyses limited to infants with a birth weight < 1500g.

Results: We evaluated 107,505 very preterm infants (27% < 1000g, 35% 1000-1500g, 38% >1500g), of whom 19,900 (19%) received probiotic supplementation (Table 1). There was a total of 11 infants with Bifidobacterium or Lactobacillus invasive infection, with 8 (0.04% or 0.4 per 1000 supplemented infants) occurring in probiotic-supplemented infants and 3 occurring in non-supplemented infants (0.003%, Table 2). This corresponded to an absolute risk increase in probiotic supplemented infants of 0.04% (95% CI 0.01%, 0.07%), equivalent to 1 invasive infection per ~2,500 supplemented infants (95% CI 1428 - 10,000). Among the 613 deaths in probiotic-supplemented infants, there was 1 death in an infant with invasive infection (>30 days after the positive blood culture). Overall mortality was lower among probiotic supplemented infants (risk difference -3.3%, 95% CI -3.6%, -3.0%). The risk of any bloodstream infection when considering incident infections among probiotic supplemented infants was not different (risk difference -0.01%; 95% CI -0.4%, 0.4%). Results were similar in a sensitivity analysis limited to infants with a birth weight < 1500g.

Conclusion(s): In this large cohort of very preterm infants cared for in U.S. neonatal intensive care units, the risks of invasive infection with Bifidobacterium or Lactobacillus were extremely uncommon among probiotic supplemented infants.

Table 1. Baseline Characteristics


Table 2. Clinical Outcomes