Amoxiclav Causes Diarrhea in 1 of 8 Persons Who Take It


Am Fam Physician. 2015 Jun 15;91(12):877.

Clinical Question

How frequently do harms of amoxicillin (with or without clavulanate) occur?

Bottom Line

The risk of diarrhea is significantly higher with amoxicillin/clavulanate (amoxiclav; Augmentin) than placebo, with a number needed to treat to harm (NNTH) of 8, which is not dissimilar from the number needed to treat for conditions such as otitis media. Contrary to popular belief, the risks of nausea, vomiting, and rash are not increased, but the risk of candidiasis is (NNTH = 23). (Level of Evidence = 1a)


Meta-analyses and systematic reviews are increasingly common in the literature, but most focus on benefit and give little attention to harm. This study is a good example of looking carefully at the harms instead. In this case, the authors searched for randomized, placebo-controlled trials of amoxicillin or amoxiclav for any indication. They did a careful search, had two authors review each article and abstract data, and evaluated study quality. Overall, the risk of bias was low, which is good. They found 45 studies of adults or children, but only 25 studies provided data regarding harms that were usable for the quantitative data synthesis.

The most common conditions being treated with amoxicillin were a respiratory infection or an ear, nose, and throat infection. Studies reported between zero and 10 harms, most commonly gastrointestinal and skin adverse effects. There was no increase in the likelihood of diarrhea overall (10 studies with 4,284 patients), but it was increased with amoxiclav (odds ratio [OR] = 3.3; 95% confidence interval [CI], 2.2 to 4.9). The pooled prevalence of diarrhea was 17.5% with amoxiclav and 5.6% with placebo (NNTH = 8). Only three studies (two of amoxicillin, one of amoxiclav) with 456 patients reported rates of candidiasis, and the likelihood was significantly increased (OR = 7.8; 95% CI, 2.2 to 27). The pooled rates were 4.4% in the amoxicillin groups and 0% in the placebo groups (NNTH = 23). Rates of nausea, vomiting, and rash were similar between the amoxicillin/amoxiclav groups and the placebo groups.

Study design: Meta-analysis (randomized controlled trials)

Funding source: Government

Setting: Various (meta-analysis)

Reference: Gillies M, Ranakusuma A, Hoffmann T, et al. Common harms from amoxicillin: a systematic review and meta-analysis of randomized placebo-controlled trials for any indication. CMAJ. 2015; 187( 1): E21– E31.

POEMs (patient-oriented evidence that matters) are provided by EssentialEvidence Plus, a point-of-care clinical decision support system published by Wiley-Blackwell. For more information, please see http://www.essentialevidenceplus.com. Copyright Wiley-Blackwell. Used with permission.

For definitions of levels of evidence used in POEMs, see http://www.essentialevidenceplus.com/product/ebm_loe.cfm?show=oxford.

To subscribe to a free podcast of these and other POEMs that appear in AFP,search in iTunes for “POEM of the Week” or go to http://goo.gl/3niWXb.

This series is coordinated by Sumi Sexton, MD, Associate Medical Editor.

A collection of POEMs published in AFP is available at https://www.aafp.org/afp/poems.



Want to use this article elsewhere? Get Permissions

More in AFP

Editor's Collections

Related Content

More in Pubmed


May 2022

Access the latest issue of American Family Physician

Read the Issue

Email Alerts

Don't miss a single issue. Sign up for the free AFP email table of contents.

Sign Up Now

Navigate this Article