This article was published in Medical Observer.
The article is also available on the Medical Observer website (may need registration).
Jack, a health conscious 48-year old man, came in a while ago with cellulitis following a small injury. As I printed the prescription, he mentioned that he had previously suffered from diarrhoea from antibiotics and asked whether eating yoghurt would help. I recently had a personal stake in this question as I was on antibiotics myself! I realised on reflection that I had hitherto given vague answers to my patients.
Does eating yoghurt reduce the risk of antibiotic-associated diarrhoea?
What does the research evidence say?
Step 1: The Cochrane Library
A quick search of the Cochrane Library revealed a number of reviews for probiotics – results of which are not entirely translatable to store bought yoghurt. The most relevant review was in children  rather than adults – probiotics (particularly high doses) might protect against antibiotic-associated diarrhoea, but the quality of the evidence was low in part due to a high number of dropouts.
Step 2: The Trip Database
I next went to the Trip Database and used their PICO search tool using the terms: P (antibiotics), I (yoghurt), C (placebo), O (diarrhoea).
Ah ha! The first result seemed promising – a trial specifically of yoghurt conducted in the general practice (GP) setting. Let’s have a closer look at this study by Conway et al. (2007) published in the British Journal of General Practice: 
We should focus on methods before results. Using the critical appraisal sheet from the Centre for Evidence Based Medicine at the University of Oxford: 
Participants: who was studied?
407 patients aged over 1-year, presenting to a single rural GP clinic in Norfolk, UK, who required a 1-week course of antibiotics. The average age was 38 years. A quarter of the participants were children, and 40% of the participants were male.
Intervention: what was the exposure?
“Bio yoghurt” produced by Yeo Valley Organics – 150 mL per day for 12 days. The yoghurt contained S. thermophillus, L. delbrueckii bulgaris, L. acidopohilus, and B. anamalis lactus. Participants also received and completed a symptom diary and received a phone call from the researchers.
Comparator: what was the control/alternative?
“Commercial yoghurt” produced by the same company in identical pots, consumed in the same manner. This yoghurt was not pasteurised and contained S. thermophillus, and L. delbrueckii bulgaris, but not the other “probiotic” organisms.
There was a third “no yoghurt” group who agreed to abstain from yoghurt during the trial.
Outcomes: what was measured?
Primary outcome: incidence of diarrhoea (defined as ≥ 3 loose stools per day, over at least 2 consecutive days during the trial).
Internal validity: are the trial results valid?
Randomised patient assignment?
Groups similar at the start?
Yes. There were no major differences between the two groups (Table 2, pg 955). 
Groups treated equally apart from assigned treatment?
Yes and no. The two yoghurt groups were treated the same, and the no-yoghurt group was treated differently.
All patients accounted for?
Yes. Dropout rates were low (9%); analyses were performed on intention-to-treat basis.
Measures objective? Or patients and clinicians kept blinded?
Yes and no. Patients who received yoghurt were blinded to the type they were allocated. It was not possible to be blinded to “no yoghurt”. “Diarrhoea” was self-reported.
What were the results?
Primary outcome: the differences in rates of diarrhoea between the three groups were not statistically significant – 7% (95% confidence interval: 3.7 to 12.5) in the “bio yoghurt” group, 11% (95% CI: 6.6 to 17.9) in the “commercial yoghurt” group, and 14% (95% CI: 9.0 to 21.5) in the “no yoghurt” group.
Questions to consider when deciding if these results help you care for your patients: 
- Is my patient so different to those in the study that the results cannot apply?
- Is the treatment feasible in my setting?
- Will the potential benefits of treatment outweigh the potential harms of treatment for my patient?
Unfortunately, this interesting study which is externally valid for the GP setting (adult patients seen in a GP clinic, using realistic interventions) suffers from a critical flaw. The “commercial yoghurt” was not pasteurised and is arguably an invalid placebo comparator to the “bio yoghurt”. Effectively, the comparison may have been between two very similar substances. Furthermore, the study was substantially underpowered – the investigators were not able to recruit their target number of participants. This was reflected in the imprecision of the results (the wide confidence intervals).
As such, we cannot and should not make any definitive conclusions from this study. Nevertheless, the results do seem to suggest that yoghurt eaters in this study suffered from less diarrhoea as compared to no yoghurt. This result is consistent with other research from the paediatric  and hospital inpatient  populations – settings which may not be externally valid to adult patients in primary care.
In my opinion, the best way forward with this lack of evidentiary clarity is with a cautious but pragmatic approach. There is no good evidence that commonly available yoghurt protects against antibiotic-diarrhoea in patients like Jack in the GP setting, so we should refrain from routinely recommending it. However, individuals who enjoy yoghurt, or who want to try, could definitely eat it more regularly while on antibiotics.
The attrition of participants from a study is problematic as it may bias the results. For instance, if those who drop out of a study do so because they found the treatment ineffective, then the study results will overestimate the intervention’s efficacy. As a rule of thumb, a study with a dropout rate over 20% may have a serious flaw.
- Johnston BC, Goldenberg JZ, Vandvik PO, Sun X, Guyatt GH. Probiotics for the prevention of pediatric antibiotic-associated diarrhea. Cochrane Database of Systematic Reviews 2011, Issue 11. Art. No.: CD004827. DOI: 10.1002/14651858.CD004827.pub3.
- Conway S, Hart A, Clark A, Harvey I. Does eating yogurt prevent antibiotic-associated diarrhoea? A placebo-controlled randomised controlled trial in general practice. British Journal of General Practice 2007; 57(545): 953-959. DOI: 10.3399/096016407782604811
- RCT Appraisal Sheets. Centre For Evidence Based Medicine [website]. Retrieved on 17 April 2013. http://www.cebm.net/index.aspx?o=1097
- Hickson M, D’Souza AL, Muthu N, Rogers TR, Want S, Rajkumar C, Bulpitt CJ. Use of probiotic Lactobacillus preparation to prevent diarrhoea associated with antibiotics: randomised double blind placebo controlled trial. BMJ 2007; 335. DOI: 10.1136/bmj.39231.599815.55