In light of their findings, providers should consider adopting the approach in clinical practice, researchers said during the Transcatheter Cardiovascular Therapeutics conference.
The offending study, which claimed an algorithm could beat out human radiologists, is just one of many examples of this problematic pattern, the group wrote in Nature.
The Information Reporting and Data Systems tool explains if findings are not concerning or if they may require following up with a doctor, experts explained in JACR.