Robert Schwartz, MD, cardiologist at the Minneapolis Heart Institute and Jeffrey Hellinger, MD, director of cardiovascular imaging and the director of the 3D laboratory at the Children’s Hospital of Philadelphia, disagreed about the value of overreading cardiac CT angiography (CTA) images. The debate was presented in a series “Controversies in Cardiac Noninvasive Imaging” this week at the Transcatheter Cardiovascular Therapeutics conference in Washington, D.C.
Schwartz said that the rationale for overreading is a decrease in liabilities. It also could detect lung cancer earlier and allows the presentation of a complete report. However, Schwartz does not believe that those reasons necessitate the need to overread.
Schwartz examined a study from the New England Journal of Medicine, which consisted of 31,567 patients. Out of that number, 320 patients (1 percent) had stage 1 cancer; 22 had lymph node cancer. According to Schwartz, “we can use CT scans to help detect lung cancer at their earliest, most curable stage.” He said only 20 percent of patients who have a nodule detected on CT scans can be expected to have a problem.
Schwartz said that “overreading does not pass the evidence-based standards” because there is low prevalence of significant disease (1 percent, with a 99 percent false-positive rate). Schwartz does not believe that overreading is cost-effective and said it could even worsen the “quality of life of its patients because of over-testing.”
Schwartz highlighted computer-aided detection (CAD) software in which the application overreads images. CAD brings attention to false-positives and with the software, there is a 26 percent reduction in missed actionable nodules.
In conclusion, Schwartz stressed that a “cardiologist is best to read coronary and vascular structures” and not venture into fields that are not his/her specialty.
In full-support of overreading, Hellinger said that cardiologists “must overread.” He said it is the responsibility of the cardiologist to overread because “it’s on the film.”
Hellinger said that lung cancer prevalence in CT screenings is between 0.3 percent and 1 percent. He reported that out of 258 patients, 145 will produce non-cardiac findings, which necessitates overreading. He said that cardiologists need recommendations for non-cardiac findings, which should be on a “case-by-case basis.” Hellinger also suggested that the cardiologists need to protect themselves from being liable for overreading.