— Conclusions still stand, group says after fixing a raft of data errors
by Crystal Phend, Senior Editor, MedPage Today August 25, 2020
Corrections to a study of lingering cardiac abnormalities after COVID-19 recovery, criticized for data analysis problems, didn’t alter the overall conclusions.
The study from Germany had been flagged on Twitter for using inaccurate metrics for the data analysis, as well as inconsistencies internally in the paper.
For example, “the fact that it’s not just one but literally dozens of similarly implausible data points suggests that the most generous interpretation of their work is ‘extremely careless,’” tweeted David Cohen, MD, of Kansas City, Missouri.
After reviewing the data and repeating analysis, Eike Nagel, MD, and Valentina O. Puntmann, MD, PhD, both of University Hospital Frankfurt, published a correction in JAMA Cardiology.
While various points changed, like correctly reporting means (SDs) or medians (interquartile ranges) and recalculation of all percentages for the risk-matched controls, the statistical effect was relatively minor, they noted.
“We are pleased to confirm that reanalysis of the data has not led to a change in the main conclusions of the study,” they wrote. “As we originally reported, compared with healthy controls and risk factor-matched controls, patients recently recovered from COVID-19 had lower left ventricular ejection fraction, higher left ventricle volume, and elevated values of T1 and T2.”
Only the comparison for left ventricular mass index between COVID-19 patients and healthy controls changed from a significant to a nonsignificant association.
Darrel Francis, MD, of the National Heart and Lung Institute of Imperial College London, had criticized the journal for dragging its feet since original publication on July 27.
“I am delighted that this, the most important cardiology paper of the decade, according to the 9000+ Altmetric score, is going to be corrected,” he told MedPage Today.
The study “has been referenced thousands of times to shut down the #Big10 and #PAC12 seasons by media including @espn,” noted one critic on Twitter.
How COVID-19 affects the heart long-term is a “very relevant” clinical question, too, noted Behnood Bikdeli, MD, of Brigham and Women’s Hospital and Harvard in Boston.
However, the small differences affirmed after the corrections probably aren’t a big problem, Francis argued on Twitter.
“The revised Jama Cardio paper is brilliant. It’s an all-clear! The interesting features on the scans in Covid survivors are real, but they are caused by the risk factors, not by the covid,” he tweeted.
High-sensitivity troponin T and high-sensitivity C-reactive protein as markers of cardiac injury and inflammation, respectively, were both significantly higher in the COVID-recovered patients than controls but still within the normal range.
