Invited Commentary 

Pediatrics

Probiotic-Associated Infection vs Necrotizing Enterocolitis—Weighing Risks and Benefits

Brian Scottoline, Brian K. Jordan

JAMA Netw Open 2026;9;(7):e2626535. doi:10.1001/jamanetworkopen.2026.26535

To our knowledge, the report by Patel et al¹ represents the largest effort to date to estimate the risk of probiotic-associated bloodstream infection in very preterm (VPT) infants (gestational age <32 weeks at birth). They have studied a large cohort of nearly 20 000 VPT infants from the Pediatrix network of neonatal intensive care units (NICUs) who had received enteral probiotics containing Lactobacillus species and/or Bifidobacterium species. Infants in this network of NICUs represent approximately 30% of all NICU patients in the US and, therefore, constitute a substantial representation of VPT infants who received probiotics in the US. The authors report a 0.04% incidence of probiotic-associated bloodstream infection in the VPT population (1 in approximately 2500 infants) and identify only 1 probiotic-associated death in the VPT infants who received probiotics (1 of 19 900). These results are congruent with previously reported incidences of probiotic-associated bloodstream infection of 0.14% in infants born at gestational age less than 34 weeks and 0.4% in infants with extremely low birth weight (LBW) (<1000 g) reported by Alshaikh et al² and less than 0.04% in preterm infants reported by Feldman et al.³ Deaths attributable to probiotic administration were approximately 1 in 6300 in the report by Alshaikh et al² and potentially 1 in 20 323 in the report by Feldman et al.³ Thus, the study by Patel et al¹ is valuable in confirming and refining the estimate of probiotic-associated bloodstream infection previously published in smaller cohorts, and it lends support to the notion that probiotics, or live bacterial products, carry a relatively low potential risk of harm, even among at-risk VPT infants.

Patel et al¹ is published against the backdrop of the use of probiotics in some US NICUs to reduce the incidence of necrotizing enterocolitis (NEC), a disease for which there is currently no US Food and Drug Administration–approved therapy, in VPT or very-LBW (VLBW) (<1500 g) infants (who frequently are VPT). Necrotizing enterocolitis is a devastating inflammatory disease of the infant gut that affects 7% to 12% of VLBW infants, equating to approximately 3500 to 6000 babies per year, and takes the lives of approximately 400 of these infants per year in the US alone, a death rate of approximately 1 in 100 VLBW infants.⁴ If survived, NEC has lifelong consequences, including brain white matter injury and associated neurodevelopmental impairment; gastrointestinal morbidities, including short gut; and increased risk for neonatal chronic lung disease, not to mention the emotional toll on these children’s families.

Necrotizing enterocolitis in VLBW infants is thought to represent one endotype of a spectrum of preterm neonatal intestinal injury,⁵ with a pathophysiology driven by a developing immune response biased toward inflammation and the emergence of a microbiome in a gut that would otherwise be exposed to a sterile or scantly colonized amniotic fluid.⁶ Biochemical and genetic data indicate that Toll-like receptor 4 (which binds lipopolysaccharide) signaling—in the context of a gut microbiome with an overrepresentation of Proteobacteria—may drive overexuberant intestinal epithelial inflammation, subepithelial immune activation, and reactive vasoconstriction compounded by enteric nervous system dysfunction, leading to NEC.⁵ In the most severe form of NEC, intestinal epithelial and subepithelial injury can progress to perforation and peritonitis; bacterial invasion; sepsis; shock; and in some cases, death. Given this model, VLBW NEC has also been referred to as dysbiotic NEC.

The resources dedicated to research on the pathobiology of VLBW NEC have unfortunately been a fraction of those devoted to other, less common and less lethal diseases. Similarly, investment in developing therapeutics to prevent or treat VLBW NEC has been very limited, in part due to a lack of familiarity with the neonatal critical care clinical space compared with, for example, adult oncology and risk aversion associated with infant patients. In light of this research and development gap, and in an effort to reduce the scourge of NEC for patients and families, clinicians and researchers have leveraged NEC pathobiology to develop and implement probiotic-based interventions to shape gut microbial ecology away from a dysbiotic NICU-acquired gut microbiome. Such interventions use ostensibly nonpathogenic, commensal bacterial species, including Lactobacillus and Bifidobacteriaspecies, often in combination and with other strains. While the data are rather heterogeneous, there is a directional signal that probiotics may be efficacious in reducing NEC incidence.⁷

There are risks associated with any therapeutic intervention, and probiotic-associated bloodstream infection is a logical concern, particularly in a preterm infant, although probiotic-associated bacteremia has been reported across all age groups. While there have been more than 3 dozen placebo-controlled randomized clinical trials evaluating the efficacy of probiotics in preventing NEC in VLBW or VPT infants,⁷ and many more observational studies, data collection and reporting on probiotic safety in VLBW or VPT infants have not been systematic. Despite the lack of systematic reporting, there have been scattered reports in the literature of probiotic-associated bloodstream infection in preterm infants.

In 2023, the combination of an increasing use of probiotics to prevent NEC and reports of probiotic-associated bloodstream infections (and potentially death) led the Food and Drug Administration to advise against the use of live bacterial products in preterm infants to prevent NEC unless under Investigational New Drug status.⁸ As a result, the use of probiotics in preterm infants in the US, which were being administered in up to 40% of NICUs, essentially stopped after the advisory. Since the advisory, there have been published reports of increased NEC incidence in VLBW infants in NICUs that stopped using probiotics, but it is unclear whether this finding can be generalized to all NICUs that used probiotics to reduce NEC incidence.

The report by Patel et al,¹ together with those by Alshaikh et al² and Feldman et al,³ makes it possible to compare the risk of bloodstream infection and death due to probiotics in VLBW, VPT, and extremely low-birth-weight infants with the risk of NEC and death due to NEC. Necrotizing enterocolitis may occur on the order of 30 to 300 times more frequently than probiotic-associated bloodstream infection, and death due to NEC may occur on the order of 60 to 200 times more frequently than due to probiotic-associated bloodstream infection.

As highlighted by Patel et al,¹ there remains a need to support large trials to evaluate the safety and efficacy of probiotics or live bacterial products to prevent NEC in VLBW infants. Based on results reported by Patel et al¹ and others, the risk-benefit profile of probiotics in VLBW infants for reducing NEC incidence and death due to NEC appears substantially weighted in favor of live bacterial products. Large trials and, if approved as a drug, patient surveillance could help refine the risk data. Paraphrasing Jennifer Canvasser, parent of an infant who succumbed to NEC and founder of the NEC Society, these results highlight the relatively uncommon risks of probiotic bloodstream infection in preterm neonates compared with the overwhelmingly more common risk of NEC.