Harms Reporting in Randomized Controlled Trials of Interventions Aimed at Modifying Microbiota: A Systematic Review

Ann Intern Med. 2018 Aug 21;169(4):240-247. doi: 10.7326/M18-0343. Epub 2018 Jul 17.


Background: Probiotics, prebiotics, and synbiotics are used increasingly, although the safety and potential harms of these interventions are poorly understood.

Purpose: To examine how harms-related information is reported in publications of randomized controlled trials (RCTs) of probiotics, prebiotics, and synbiotics.

Data sources: Cochrane Central Register of Controlled Trials, PubMed, EMBASE, and Web of Science (without language restrictions) from 1 January 2015 to 20 March 2018.

Study selection: RCTs assessing the safety or efficacy of at least 1 intervention involving probiotics, prebiotics, or synbiotics alone or in combination with another intervention compared with any control (such as a placebo or an antibiotic) for any clinical condition.

Data extraction: 4 reviewers independently assessed study characteristics, the reporting of harms, and the presentation of safety results.

Data synthesis: Of 384 trials conducted in healthy volunteers (n = 136) or patients with any of several medical conditions (n = 248), 339 (88%) were published in specialty journals. Trials most often evaluated probiotics (n = 265 [69%]). Studies in persons with medical conditions enrolled outpatients (n = 195) and high-risk patients (n = 53). No harms-related data were reported for 106 trials (28%), safety results were not reported for 142 (37%), and the number of serious adverse events (SAEs) per study group was not given for 309 (80%). Of 242 studies mentioning harms-related results, 37% (n = 89) used only generic statements to describe AEs and 16% (n = 38) used inadequate metrics. Overall, 375 trials (98%) did not give a definition for AEs or SAEs, the number of participant withdrawals due to harms, or the number of AEs and SAEs per study group with denominators.

Limitation: Journal publication processes may have affected the completeness of reporting; only English-language publications were examined.

Conclusion: Harms reporting in published reports of RCTs assessing probiotics, prebiotics, and synbiotics often is lacking or inadequate. We cannot broadly conclude that these interventions are safe without reporting safety data.

Primary funding source: No specific funding.

Publication types

  • Systematic Review

MeSH terms

  • Humans
  • Microbiota / drug effects*
  • Prebiotics / adverse effects*
  • Probiotics / adverse effects*
  • Publishing / standards*
  • Randomized Controlled Trials as Topic / standards*
  • Research Design
  • Synbiotics / adverse effects*


  • Prebiotics