Probiotics: Help or Harm in Antibiotic-Associated Diarrhea?

David A. Johnson, MD


September 10, 2014

This feature requires the newest version of Flash. You can download it here.

Probiotics: Beneficial or Harmful?

Hello. I'm Dr. David Johnson, Professor of Medicine and Chief of Gastroenterology at Eastern Virginia Medical School in Norfolk, Virginia. Welcome to another GI Common Concerns -- Computer Consult.

Today I want to discuss the issue of probiotics, and whether probiotics are doing an element of benefit or an element of harm.

With access to over-the-counter products, use of probiotics has dramatically increased. Physicians recommend probiotics routinely to patients when they are taking antibiotics to prevent antibiotic-associated diarrhea. I would like to take a time-out and reevaluate what we are doing for these patients.

Antibiotic-associated diarrhea is not uncommon. It occurs in 25%-30% of patients receiving antibiotics, and is more common with amoxicillin/clavulanic acid antibiotics (Augmentin®).

Not infrequently, antibiotics are associated with Clostridium difficile infections, which occur in up to one third of patients with antibiotic-associated diarrhea. C difficile has become rampant, particularly as a nosocomial infection in hospitals. It is now the leading cause of hospital-related infectious mortality from a nosocomial perspective, and is certainly on our radar screen for hospitalized patients. Is there anything we can do to prevent this?

Meta-analyses of Probiotics to Prevent Diarrhea

In 2012, highly publicized meta-analyses were published in JAMA[1] and Annals of Internal Medicine.[2] These studies, and a Cochrane review,[3] suggest that not only can probiotics prevent or diminish antibiotic-associated diarrhea, but probiotics may also be helpful in avoiding C difficile infection.

What are the data on this, and should we all be recommending probiotics? It's difficult to put into perspective, because when you look at the trials that were included in the recent meta-analyses, the individual studies were relatively small.

Furthermore, there was incredible heterogeneity among these trials. The outcomes weren't as specific as they could be, and so combining these in meta-analyses leads at least to a subject bias in terms of the heterogeneity. Are these trials representative of the larger population? With a large prospective study, would we find the same results?


Enter the most recent study, which is called the PLACIDE study, from the United Kingdom.[4] This study was conducted in Wales and northeastern England. It involved 5 hospitals, 68 different medical and surgical units, and more than 17,000 patients aged 65 years or older. All patients were hospitalized and taking an antibiotic.

These patients were randomly assigned, if they met eligibility criteria, to receive either a microbial preparation (which is the term they used for "probiotic") or an identical placebo. These patients were not in intensive care, and none had prosthetic valves or a history of inflammatory bowel disease or C difficile infection.

The microbial preparation had 2 strains of Lactobacillus and 2 strains of bifidobacteria, which patients received for 21 days. The follow-up was 8 weeks in these patients, and loss to follow-up was very small -- only about 10 patients in each group. Data were collected on diarrhea events, and C difficile by standard testing, and these patients were then followed to outcome. Even with evaluation for intention to treat, there was no difference in the outcomes for C difficile infection or antibiotic-associated diarrhea between the microbial preparation (probiotic) and placebo group.

Of interest, there was an increase in flatus in the microbial preparation group, and patients with C difficile diarrhea who received the microbial preparation reported a 3-fold increase in bloating. That is a number needed to harm of 2.5, so now let's back up and think what this means. They didn't get a better reduction in the prevention of C difficile, and now we are talking about the harmful side effects of receiving a medication that was supposed to make them better.

Don't Rush to Give Probiotics

So where are we?

You could argue that this study looked at 2 commonly used bacterial strains (Lactobacillus and bifidobacteria) but not Saccharomyces boulardii, which in one study has shown some benefit. But even when the most recent American College of Gastroenterology guidelines[5] on C difficile included the studies on Saccharomyces, they concluded that there was not enough benefit; only some post hoc analyses suggested some benefit. Saccharomyces is not recommended for the prevention of C difficile.

Now we have an incredibly large, placebo-controlled trial with level 1 evidence to suggest that we are not doing any good for these patients, and we might be causing adverse symptoms with a relatively low number needed to harm. It's not the final answer, but we need to reconsider the use of probiotics.

Although intended to restore good health, we are seeing a dysbiosis. We have disrupted the microflora in the gut, and are trying to jam it back with strains of bacteria that we think are good bacteria, and it may not be the correct answer.

We don't know the right answer. When you alter the microflora, you change some of the metabolism of carbohydrates, bile salts, and complex sugars. We are not clear whether jamming the gut with another strain of bacteria is going to be of benefit.

Probiotics: Do No Harm

I want to posit an element of potential harm, and not rush in to recommend probiotics routinely in patients to whom you prescribe antibiotics.

I would also caution you not to use probiotics in patients in the intensive care unit, or in any patient with an indwelling prosthesis, particularly an intravascular prosthesis. There are reports of fungemia and bacteremia with probiotics, and the first rule is to do no harm.

A report that will come out soon in Clinical Gastroenterology and Hepatology[6] describes a retrospective analysis of more than 12,000 hospitalized patients. Approximately 800 of these patients had received metronidazole for other indications before receiving broad-spectrum antibiotics. The investigators found an 80% reduction in C difficile infection in patients who had received metronidazole. This could be something to look at prospectively.

As we rush to do good, let's pause and ask whether we are doing harm. Probiotics may not be the answer. Although they are certainly appealing, they might not be the right answer for patients. Patients should talk with their healthcare providers, because sometimes these "good bacteria" may not be so good, and they may do an element of harm. We will look for some more prospective studies and update you when we have more answers.

I'm Dr. David Johnson. Thanks for listening.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.