Abstract
INTRODUCTION: Understanding the safety of pharmacotherapy for irritable bowel syndrome (IBS) enables individuals to make informed treatment decisions. While many studies include the number needed to treat to highlight therapeutic benefits, adding the number needed to harm (NNH), a measure we evaluate herein, could enable more comprehensive risk-benefit assessments. METHODS: PubMed, Web of Science, and Cochrane databases were searched through October 2024. Clinical trials investigating IBS pharmacotherapies including discontinuation rates because of adverse events (AEs) were included. Data were pooled using a random-effects model. The primary outcome was NNH for each pharmacotherapy, defined as the reciprocal of the absolute difference in risk of AEs leading to treatment discontinuation between the experimental and placebo groups. Secondary outcomes included the relative risk of withdrawing because of an AE and the most common AEs for each drug. RESULTS: Fifty-four trials met inclusion criteria. For constipation-predominant IBS pharmacotherapies, the NNH for linaclotide, lubiprostone, plecanatide, tegaserod, and tenapanor was 35 ( P < 0.01), 53 ( P = 0.59), 59 ( P < 0.01), 58 ( P = 0.03), and 16 ( P < 0.01), respectively. For diarrhea-predominant IBS pharmacotherapies, the NNH for alosetron and eluxadoline was 14 ( P < 0.01) and 32 ( P < 0.01) whereas the NNH for both rifaximin and ramosetron was a negative, although statistically insignificant, value. For IBS global symptom pharmacotherapies, the tricyclics, the NNH was 24 ( P < 0.01). Many AEs were transient without long-term sequela. DISCUSSION: Among pharmacotherapies for IBS, tricyclics (especially at elevated doses), tenapanor, and alosetron have the highest absolute risk of discontinuation because of an AE when compared with rifaximin, the safest pharmacotherapy studied.