Second reading of coronary angiograms by radiologists

In many hospitals in the provinces of Quebec and Nova Scotia, as well as in some hospitals in the rest of Canada, coronary angiograms are performed and interpreted by invasive cardiologists, and are later reinterpreted and reported by radiologists.

To evaluate the value of second readings of coronary angiograms by radiologists.

Cardiology and radiology reports of a total of 160 consecutive coronary angiograms were compared from patients at three hospitals. Ten segments of the coronary tree were considered and 1582 segments were included. Agreement between cardiology and radiology interpretations was evaluated using per cent agreement, Pearson correlation and Bland-Altman limits of agreement. Agreement was calculated for each arterial segment and for each hospital.

Excellent agreement was found between cardiology and radiology interpretations of coronary angiograms. Per cent agreement ranged from 94.9% to 100%, Pearson correlation ranged from 0.83 to 0.97 and Bland-Altman limits of agreement ranged from -18.1 to 19.4. Agreement was similar for each segment and for each hospital. Agreement remained excellent after exclusion of normal angiograms (n=348 segments), with a per cent agreement of 96.3%. Secondary analyses demonstrated a mean time delay of 13 days between angiograms and the subsequent radiology reports.

There are minimal differences between the cardiology and radiology interpretations of coronary angiograms. Routine second reading by a radiologist may be redundant.