Comparative Effectiveness Analysis of Phase 3 Clinical Trials in Moderately to Severely Active Inflammatory Bowel Disease Reveals Increased Harm From Placebo and Challenges Equipoise
Joëlle St‐Pierre, Zachary D. Fine, Evan N. Fear, Jeremy A. Klein, Natalie K. Choi, David T. RubinABSTRACT
Background
Placebo‐controlled trials are required for regulatory approval of inflammatory bowel disease (IBD) therapies but raise ethical concerns when effective treatments exist. This study quantifies harm associated with placebo in randomized controlled trials (RCTs) for ulcerative colitis (UC) and Crohn's disease (CD).
Methods
A comparative effectiveness analysis of placebo‐controlled Phase 3 RCTs in moderately to severely active UC or CD was conducted. We extracted adverse events (AEs), serious AEs (SAEs), disease‐exacerbation AEs (DEAEs), and serious disease‐exacerbation events (DESAEs), calculating number needed to harm (NNH) and risk ratio (RR) per trial. A positive ΔNNH reflects a higher incidence of harm with placebo.
Results
We analyzed 22 UC and 25 CD trials. AE rates were largely equivalent between placebo and treatment, with median RRs clustered around 1.0. By contrast, SAEs consistently favored active therapy: in UC induction, the median NNH was 31.0 (RR 0.6), in UC maintenance 18.0 (RR 0.8), in CD induction 29.0 (RR 0.8), and in CD maintenance 22.0 (RR 0.9). The most pronounced harms were for disease exacerbations: in UC induction, the median NNH was 12.0 for DEAEs (RR 0.6) and 9.0 for DESAEs (RR 0.5); in UC maintenance, 10.0 (RR 0.5) and 8.0 (RR 0.4); in CD induction, 13.0 (RR 0.6) and 11.0 (RR 0.5); and in CD maintenance, 11.0 (RR 0.5) and 9.0 (RR 0.4).
Conclusion
Placebo exposure in IBD trials leads to clinically significant harm. Clinical trial designs and regulatory guidance should be re‐evaluated to ensure an ethical balance between efficacy and safety.