We know that our gut microbiome is altered by critical illness, and that gut translocation is a key component of the pathophysiology of multi-organ failure. Giving a probiotic aims to ‘restore the deranged microbiome to health’. Is that plausible? And if so do we really understand the implications of our meddling?
The most recent meta-analysis is this one published (open access) in Critical Care. They used a random effects model (as described in this post).
- 30 RCTs were included. Five were considered to be high quality (by the authors own criteria).
- For the 14 trials with infection as an outcome measure, there was a significant reduction in (all) infections (RR 0.80, 95% CI 0.68,0.95 P=0.009 I2=36%).
- For the 9 trials considering ventilator aquired pneumonia, probiotic use showed a rate reduction (RR 0.74, 95% CI 0.61,0.90 P-0.002, I2=19%).
- Perhaps unsurprisingly there was also a reduction of antibiotic duration (of a day).
- There was no difference in mortality (the authors had to publish an erratum, but whilst the number changed the meaning did not. The error was pointed out in this letter, to which the response is here).
- There was no difference in ICU or hospital length of stay (it always makes me suspicious when a reduction is shown in something that affects ICU length of stay but the ICU LOS remains unchanged – all other things being equal that shouldn’t be so).
- There was no effect on diarrhoea.
Subgroup analysis was also undertaken. Interestingly the dose made no difference, which suggests if beneficial that only a low dose is required, but it might also suggest that probiotics are ineffective!
In some cases there was a significant benefit in one group but not the other, yet when they were compared there was no significant difference between the groups. This was the case for the presence of L. planetarium, and for probiotics not containing L. rhamnosus or symbiotics, as well as for the use of probiotics in groups with a higher mortality.
The only positive result for the subgroup analysis was that lower quality studies were more likely to give a positive result (P=0.03). Publication bias was also identified for the overall finding of reduced infection with probiotics which is perhaps unsurprising but worth noting.
Whilst this meta-analysis only considered potential benefit, it’s clearly important to look at potential harm. The ‘bad press’ if there is any comes from the PROPRIATA trial which looked at probiotics in pancreatitis; the abstract of which is certainly enough to put you off. A subsequent meta-analysis has calmed the waters a bit as far as use in pancreatitis is concerned. The British Society of Gastroenterology guidelines make no reference to probiotics at all (they are however 12 years old).
We only have two interventions to influence the gut flora (3 if you include faecal transplant) – probiotics and selective gut decontamination. Both are blunt instruments for a complex problem and both seem resigned to a series of small studies, multiple meta-analysis and patchy implementation. This is an area with huge potential, but maybe with our current understanding it’s just too difficult. To future might be exciting though!
What you you think – probiotics for all??
*By way of definitions:
- A probiotic is a living microorganism which is non-pathological and administered to prevent disease.
- A prebiotic is a substance given to aid the growth or activity of microorganisms that are beneficial to the host. In the gut this is commonly fibre to act as a nutrient.
- A symbiotic is a combination of the two.