Reanalysis of randomized clinical trials could change conclusions

Through a reanalysis of previously published randomized clinical trials, researchers from Stanford University School of Medicine found one-third of previously published randomized clinical trials could be reanalyzed in ways that modify the conclusions of how many or what types of patients need to be treated, according to study results.

Researchers completed a Medline search to identify all published studies that completed a reanalysis of individual patient data from previously published randomized clinical trials addressing the same hypothesis as the original randomized clinical trial. Main outcome measures included changes in direction and magnitude of treatment effect, statistical significance and interpretation about the types or numbers of patients who should be treated.

Overall, researchers identified 37 eligible reanalyses, of which five were conducted by researchers who were not associated with the original studies. Study results showed statistical or analytical approaches and definitions or measurements of the outcome of interest most commonly differed in the reanalysis. Researchers found 35% of reanalyses led to interpretations different from the original article with 8% showing different patients should be treated; 3% that fewer patients should be treated and 24% that more patients should be treated.

“There is a real need for researchers to provide access to their raw data for others to analyze,” John Ioannidis, MD, DSc, professor of medicine and director of the Stanford Prevention Research Center, stated in a press release. “Without this access, and possibly incentives to perform this work, there is increasing lack of trust in whether the results of published, randomized trials are credible and can be taken at face value.”

For more information:

Ebrahim S. JAMA. 2014;doi:10.1001/jama.2014.9646.

Disclosure: See the full study for the authors’ relevant financial disclosures.

Leave a Reply

Your email address will not be published.