Importance: For neonates with very low birth weight (VLBW), randomized clinical trials (RCTs) indicate that probiotic treatment decreases the risk of necrotizing enterocolitis (NEC), with smaller decreases in the risk of sepsis and death. There is little evidence on the rate of probiotic adoption in US neonatal intensive care units (NICUs) and whether the benefits seen in trials have materialized in practice.
Objective: To estimate changes in probiotic use among neonates with VLBW and to test whether neonates with VLBW treated at NICUs adopting routine probiotic use experience better outcomes compared with neonates treated at nonadopting NICUs.
Design, setting, and participants: This cohort study used Vermont Oxford Network data on neonates with VLBW in US NICUs from January 1, 2012, to December 31, 2019. Data were analyzed from January 2022 through February 2023.
Exposure: Probiotics adoption vs nonadoption. Adopting NICUs were defined as those that currently or previously treated at least 20% of neonates with VLBW with probiotics.
Main outcomes: The primary outcomes were rates of NEC, in-hospital mortality, and sepsis, defined as bacterial or fungal infection occurring after day 3 from birth. A difference-in-differences analysis compared changes in VLBW infant outcomes between adopting and nonadopting NICUs before and after hospital-level adoption of probiotics. Additional analyses used the proportion of neonates treated with probiotics in each neonate's birth NICU and year.
Results: The analysis included 307 905 neonates with VLBW (mean [SD] gestational age, 28.4 [2.9] weeks; 50.0% male) at 807 US hospitals. The rate of probiotic treatment of neonates with VLBW rose from 1572 of 38 296 neonates (4.1%) in 2012 to 4788 of 37 910 (12.6%) in 2019. Only 123 of 745 NICUs (16.5%) adopted probiotics by 2019, with 4591 of 6017 neonates with VLBW (76.3%) receiving probiotics in 2019 at adopting NICUs. Incidence of NEC declined by 18% at adopting NICUs (odds ratio [OR], 0.82; 95% CI, 0.70-0.95; P = .10) compared with nonadopting NICUs. Probiotic adoption was not associated with a significant reduction in sepsis (OR, 1.11; 95% CI, 0.98-1.25; P = .09) or mortality (OR, 0.93; 95% CI, 0.80-1.08; P = .33).
Conclusion and relevance: In this cohort study, adoption of routine use of probiotics increased slowly in US NICUs and was associated with lower NEC risk but not with sepsis or mortality among neonates with VLBW. The findings for probiotic adoption and NEC, sepsis, and mortality were smaller than would have been predicted by the totality of RCT evidence but are consistent with a meta-analysis restricted to studies at low risk of bias.