In a study published in JAMA Internal Medicine, a team of researchers at Florida International University Robert Stempel College of Public Health and Social Work, led by Dr. Craig A. Garmendia, found that almost half of all meta-analyses had conclusions altered by falsified data publications. This resulted in nearly one third of all analyses having considerable changes in outcomes.
“The desire to include all available data in a meta-analysis to obtain the ‘best estimate’ of effect size may result in the inclusion of falsified data,” said Dr. Garmendia. “This in turn may provide biased results that compromise future research, policy decisions, and even patient care.”
The study analyzed the effect of publications with falsified data on the results of meta-analyses, which can be affected by the inclusion of publications with falsified data. Robust sensitivity analyses and prevention of publication of falsified data, affecting not only the original publication but also any subsequent meta-analyses, is needed to avoid compromising future research, policy decisions, and patient care.
This study found that 45.5 percent of all meta-analysis publications had conclusions altered by publications with falsified data and 32.3 percent of all the analyses having a considerable change in the outcome. Full meta-analyses were more robust against the effects of publications with falsified data. For odds ratios not statistically affected, the estimates generally moved towards the null when more than one publication remained.
Falsified data can affect not only the original publication, but also any subsequent meta-analyses and any resulting clinical or policy changes resulting from the findings of these studies.
Read more about the study.Tags: Friday Letter Submission