Association between trial registration and treatment effect estimates: a meta-epidemiological study

Area of Science:

• Clinical Trials
• Meta-epidemiology
• Research Integrity

Background:

• International Committee of Medical Journal Editors mandated prospective clinical trial registration in 2005 to enhance research transparency.
• A significant number of clinical trials remain unregistered or are registered retrospectively, potentially compromising data integrity.

Purpose of the Study:

• To investigate the association between prospective registration of clinical trials and the magnitude of estimated treatment effects.
• To determine if unregistered or retrospectively registered trials yield different effect sizes compared to prospectively registered trials.

Main Methods:

• A meta-epidemiological study analyzed Cochrane reviews published between March 2011 and September 2014.
• Included meta-analyses of binary outcomes with at least three randomized controlled trials published after 2006.
• Trials were classified as prospectively registered, retrospectively registered (before or after primary completion date), or unregistered, with data extracted from reviews and registries.

Main Results:

• Out of 322 trials across 67 meta-analyses, 70% were registered (33% prospectively, 63% retrospectively).
• Unregistered or retrospectively registered trials showed a tendency towards larger treatment effect estimates (ROR=0.81).
• Trials registered after the primary completion date also exhibited larger treatment effects compared to those registered before (ROR=0.84).

Conclusions:

• The lack of prospective clinical trial registration is potentially linked to inflated treatment effect estimates.
• Ensuring prospective registration is vital for maintaining unbiased research findings and reliable evidence synthesis.