Table 3 Probiotic trials in mechanically ventilated patients for prevention of ventilator-associated pneumonia/respiratory tract infection

VAP/respiratory tract infection as primary outcome

Forestier and colleagues [49]

Mixed ICU, >18 years old, requiring MV >48 hours

Lactobacillus casei rhamnosus, 10⁹ CFU vs. placebo ng/og twice daily until ICU discharge or death

Time to first colonisation/infection of the gastric and respiratory tracts with Pseudomonas aeruginosa strains. Median delay 50 days probiotic vs. 11 days placebo (P = 0.01)

P. aeruginosa VAP reduced in probiotic group: 2.9% vs. 7.5% (P = NS). No adverse effects

Single centre. More patients in probiotic group than placebo received antipseudomonal antibiotics during their admission (55% vs. 43%)

Knight and colleagues [51]

Mixed ICU, >16 years old, requiring MV >2 days

Synbiotic 2000 Forte: 10¹⁰ CFU vs. placebo ng/og twice daily until day 28, ICU discharge or death

Incidence of VAP: 9% probiotic vs. 13% placebo (P = 0.42)

Hospital mortality: 27% vs. 33% (P = 0.39). No adverse effects

Single centre. Overall VAP rate lower than anticipated

Morrow and colleagues [38]

Mixed ICU, >18 years old, expected to require MV >72 hours

Lactobacillus rhamnosus GG 10⁹ CFU vs. placebo per orally and ng twice daily, started within 24 hours until death, extubation or tracheostomy

Incidence of VAP: 19.1% probiotic vs. 40% placebo (P = 0.007)

Significant reduction in Clostridium difficile-associated diarrhoea and ICU-associated diarrhoea, fewer antibiotic days, delay in onset of VAP, reduction in gastric and oral colonisation with pathogenic species, preferential reduction in VAP caused by Gram-negative pathogens. No adverse effects

Single centre. Small sample size. High-risk population, selected: mean APACHE II score 23, mean days ventilated 10 days. Extensive exclusions: pregnancy; immunosuppression; prosthetic heart valve or vascular graft; cardiac trauma; history of rheumatic fever, endocarditis or congenital heart defect; gastro-oesophageal or intestinal injury or foregut surgery; oropharyngeal injury; tracheostomy

VAP/respiratory tract infection as secondary outcome

Kotzampassi and colleagues [47]

Severe multiple trauma patients, >18 years old, requiring MV

Synbiotic 2000 Forte: 10¹¹ CFU vs. placebo once daily via gastrostomy or ng for 15 days

Systemic infection rate during ICU stay or the development of SIRS and MODS. Overall infection rate: 63% probiotic vs. 90% placebo (P = 0.01)

VAP rate reduced in probiotic group: 54% vs. 80% in placebo group (P = 0.03). Central line and urinary tract infections also significantly reduced. Severe sepsis: 17% vs. 40% (P = 0.04). Ventilation days (P = 0.001) and ICU length of stay (P = 0.01) significantly reduced with probiotics. Reduction in mortality (14.3% vs. 30%) nonsignificant (P = 0.12). No adverse effects

Small sample size, severe trauma patients only

Spindler-Vessel and colleagues [48]

Severe multipletrauma patients requiring MV, at least 4-day ICU stay

Synbiotic 2000 Forte, 10¹⁰ CFU vs. glutamine or fermentable fibre or peptide diet once daily ng/og for 15 days or until ICU discharge or death

Effect on intestinal permeability reduced on day 7 in probiotic group only (P <0.05)

Probiotic group also had fewer pneumonias (P = 0.03) and total infections (P = 0.003). No adverse effects

Single centre. Small sample size, comparing multiple interventions, no placebo group

Klarin and colleagues [50]

Mixed ICU, 18 years old, requiring MV >24 hours

Lactobacillus plantarum 299, 10¹⁰ CFU vs. 0.1% chlorhexidine per orally twice daily until ICU discharge or death

Pathogenic bacterial load in oropharynx. New colonisation rate: 34.8% probiotic vs. 61.9% chlorhexidine (P = 0.13)

Emerging bacteria largely Gram-negative species. Noncolonised patients had lower ventilator days (P <0.001). Incidence of VAP: 4% probiotic group vs. 14% chlorhexidine group (P = NS). No adverse effects

Single centre. Not powered for incidence of VAP as primary outcome. Small sample size

Barraud and colleagues [53]

Medical ICU, >18 years old, MV > 48 hours

Ergyphilus 2×10¹⁰ lactic acid bacteria, mostly L. rhamnosus GG, once a day vs. placebo via enteral feeding tube.

28-day mortality. No difference: 25.3% probiotic vs. 23.7% placebo (P = 0.8)

Mortality rates in ICU and at 90 days were also unaffected by the treatment. Incidence of ICU-acquired infections including VAP not significantly different except for catheter-related bloodstream infections that were lowered by probiotics. Reduced 28-day mortality in severe sepsis patients given probiotics (P = 0.035) but higher mortality rate in nonsevere sepsis patients (P = 0.08)

Single centre. Small sample. Stopped early

Oudhuis and colleagues [52]

Mixed ICUs, consecutive ICU patients with expected MV ≥48 hours and/or expected ICU stay ≥72 hours

L. plantarum 299/299v in a dose of 5×10⁹ CFU together with 6 g of rose-hip, twice daily via ng vs. SDD

ICU-acquired infection rate: 31% probiotic vs. 24% SDD (P = 0.10)

No significant difference in VAP rate (7.7% vs. 7.2%.), 28-day or ICU mortality between groups

Small sample size. Stopped early. Crossover of units was not completed, resulting in unequal mix of patients and disease burden. Not DB. Infections defined retrospectively