Probiotics in the ICU: a scoping review of evidence for infection prevention

Intensive care unit-acquired infections (ICU-AIs) are associated with substantial morbidity and mortality in critically ill patients. Although probiotics have been investigated as a preventive strategy, clinical trials have reported conflicting results, largely attributable to heterogeneity in strain composition, treatment duration, and dosage. This scoping review aimed to examine the influence of strain specificity, intervention duration, and daily dosage on probiotic efficacy in preventing ICU-AIs in adults. We conducted this scoping review in accordance with PRISMA-ScR guidelines. Five electronic databases (PubMed, Scopus, Web of Science, MEDLINE, and Embase) were systematically searched from January 2014 to September 2024 for English-language interventional and observational studies. Data on probiotic strain(s), duration, dosage (CFU/day), and infection outcomes were extracted and synthesised narratively. Ten studies (seven randomised controlled trials and three observational studies; total > 6,500 patients) were included, encompassing 18 distinct probiotic strains. Multi-strain preparations containing Lactobacillus acidophilus LA-5 (evaluated in four publications from two RCTs) were most consistently associated with benefit. All five studies using both intervention duration ≥ 14 days and dosage ≥ 5 × 10⁹ CFU/day reported significant reductions in at least one ICU-AI (100%; Fisher exact p = 0.007), whereas none of the five studies with shorter duration or lower dosage achieved significance (0%). Single-strain Lacticaseibacillus rhamnosus GG and low-dose regimens showed no effect. Current evidence indicates that probiotic efficacy against ICU-AIs is highly strain-, duration-, and dose-dependent, with multi-strain regimens containing Lactobacillus acidophilus LA-5 administered for ≥ 14 days at ≥ 5 × 10⁹ CFU/day demonstrating the most promising results. However, the limited number of studies and persistent heterogeneity preclude firm recommendations. Large, multicentre, strain-specific randomised trials are needed to confirm these observations and to evaluate safer alternatives such as postbiotics.