This article is more than 5 years old. Information may no longer be current.
Publication bias found in emerging CVD biomarker research
Click Here to Manage Email Alerts
Reporting of biomarkers for predicting CVD risk may be biased and inflated, researchers reported in a new JAMA Internal Medicine study.
To determine whether there is evidence for biases favoring significant results and inflating associations, researchers conducted a PubMed search for meta-analyses of CV biomarkers that were not part of the Framingham Risk Score.
Ioanna Tzoulaki
Of 56 eligible meta-analyses, researchers found that 49 had statistically significant results. Very large heterogeneity was found in nine analyses and small-study effects were found in 13 analyses. Twenty-nine meta-analyses had an excess of statistically significant results, whereas only 13 of the statistically significant meta-analyses had more than 1,000 cases with no hints of large, heterogeneity, small-study effects or excessive significance. According to the study, these analyses included associations of glomerular filtration rate and albumin-to-creatinine ratio in high-risk and general populations with CVD mortality as well as associations with non-HDL cholesterol, serum abumin, Chlamydia pneumoniae IgG, glycosylated hemoglobin, nonfasting insulin, apolipoprotein B/AI ratio, erythrocyte sedimentation rate and lipoprotein-associated phospholipase mass/activity with CHD.
“We found strong evidence to suggest the effect of biomarkers is exaggerated because the largest studies, which one would expect to produce the most stable estimates, consistently showed smaller effects,” Ioanna Tzoulaki, PhD, of the department of hygiene and epidemiology, University of Ioannina School of Medicine, Greece, wrote. “In most meta-analyses, too many single studies had reported ‘positive’ results compared with what would be expected on the basis of the results of the largest study. This suggests that small studies with ‘negative’ results remain unpublished or that their results are distorted during analysis and reporting to seem more prominent.”
In an accompanying editorial, Steven E. Nissen, MD, Cardiology Today Editorial Board member, said this study “documents another troubling aspect of publication bias — selective reporting of associations between biomarkers and cardiovascular outcomes.
Steven E. Nissen
“Their findings are striking and disturbing, demonstrating that most meta-analyses of biomarkers commonly used in cardiovascular medicine show evidence of publication bias. … In a few cases, the evidence for publication bias was extreme. For example, routine measurement of carotid intima-medial thickness has been advocated as a means to predict CV risk and select patients for treatment, but the current analysis demonstrates a greater than 12-fold excess in the number of favorable studies compared with what would be predicted. … We must emphasize to colleges and trainees that all studies contribute to scientific understanding.”
For more information:
Nissen SE. JAMA Arch Intern Med. 2013;doi:10.1001/jamainternmed.2013.4074.
Tzoulaki I. JAMA Arch Intern Med. 2013;doi:10.1001/jamainternmed.2013.3018.
Disclosure: Tzoulaki and Nissen report no relevant financial disclosures.
Perspective
Back to Top
L. Kristin Newby, MD
My biggest questions are: Is it reporting bias? Are these things consistently associated with the study endpoints? Or, are people jumping on the bandwagon to test a marker in 'their population' once someone else has had success? It is probably a combination of all three questions.
Just because there is a consistent association of a biomarker with outcome, does not mean there is reporting bias. For example, elevated troponin is associated with adverse outcome in almost every disease state and study completed. I would be suspicious if there were negative studies. So, whether this methodology of assessment of heterogeneity suggests bias in reporting of research for a given biomarker requires an understanding of the biomarker-disease state relationship.
Beyond that — and importantly — the study should be a reminder to researchers and a call to action that publishing negative research is important to fully understand and put in perspective the characteristics and performance of a given biomarker, every bit as important and in some cases more important, than reporting positive results.
Going forward, I would encourage a commitment by all investigators to prepare and submit for publication the results of all research carried out on human subjects and a commitment by the editors of the peer-reviewed literature that they will continue to espouse the consideration of all research on human subjects for its merits, not just for the direction of the results.
Click Here to Manage Email Alerts