Comparing CT interpretations made by emergency physicians with those from radiologists reading the same scans, researchers in Iran found an agreement rate of 68.2 percent, leading them to urge caution among emergency doctors who feel pressured to supply their own reads for critical cases in the absence of an on-call radiologist.
Shahram Bagheri-Hariri, MD, of Tehran University of Medical Sciences and colleagues, including Mona Arbab, MD, of Harvard Medical School, published their findings online Aug. 7 in Emergency Radiology.
The team enrolled all patients who underwent contrast-enhanced abdominopelvic CT in the emergency department over a three-month period.
A total of 170 scans were read first by attending emergency physicians, who then initiated treatment.
Within 12 hours of the first read, radiologists blinded to the first reads interpreted the scans.
For the study, the radiologists’ reports were grouped as either in agreement or disagreement with the ED doctors’ interpretations. A panel of experts evaluated the disagreement groups’ medical charts and placed them in clinically significant or insignificant group based on the follow-up for 28 days.
The researchers found that, in the clinically significant disagreement group, eight patients did not receive the required treatment, and three patients were over-treated.
There were five mortalities, none of which could have been prevented by a prompt radiologist’s report, the authors report.
The team also found the disagreement group had longer hospital stays and more transfers to inpatient units.
“Our findings support the cautious use of ED physicians’ CT scan interpretations for patients’ management,” the authors write. “Ideally, the ED physicians should utilize a real-time radiologist interpretation in critical patients. This collaboration will result in better patient management.”