Randomized trials are frequently fragmented in multiple secondary publications

Area of Science:

• Clinical Research
• Biostatistics
• Publication Ethics

Background:

• Secondary publications, such as subsequent analyses of randomized controlled trials (RCTs), are common.
• Understanding the frequency and characteristics of these publications is crucial for interpreting research findings.

Purpose of the Study:

• To determine the prevalence and features of secondary publications originating from primary randomized controlled trials (RCTs).
• To identify factors predicting the likelihood of secondary publications.
• To analyze the types of secondary analyses and their reported statistical significance.

Main Methods:

• A cohort of 191 RCTs published in high-impact journals in 2009 was screened for secondary publications authored by at least one author of the primary trial.
• Statistical modeling was used to assess predictors of secondary publications.
• A subset of secondary publications was analyzed for hypothesis testing, outcomes, and eligibility criteria.

Main Results:

• 46% of RCTs (88 out of 191) resulted in 475 secondary publications by 2014.
• Increased sample size (1.32-fold risk per 10-fold increase) and the presence of a design article (1.71-fold risk) predicted more secondary publications.
• Nearly all (91%) examined secondary publications reported statistically significant results, often testing different hypotheses than the primary trial.

Conclusions:

• Secondary publications are frequent, with some RCTs generating a substantial number.
• Factors like sample size and study design influence the generation of secondary publications.
• The high rate of statistically significant findings in secondary publications warrants careful consideration during evidence synthesis.