Out of hours multidetector computed tomography pulmonary angiography: are specialist resident reports reliable?

The purposes of this study were to assess the accuracy of trainee radiologists' reports for computed tomographic pulmonary angiographic (CTPA) imaging and to determine agreement or discrepancy with final verified consultant reports.

A total of 100 consecutive out-of-hours CTPA examinations were prospectively analyzed. Fifty-one male and 49 female subjects were included in the study. The mean age of patients scanned was 63.7 years (range, 17-98 years).

Eighteen of the 100 subjects (18%) had findings positive for pulmonary embolism. The interobserver agreement for pulmonary embolism between on-call radiology residents and consultant radiologists was almost perfect (κ = 0.932; 95% confidence interval, 0.84-1.0; P < .0001). There was one false-negative CTPA report. Eighty-two CTPA scans (82%) were reported as negative for pulmonary embolism by consultant radiologists. In this group, there was a single false-positive interpretation by the on-call specialist resident. The interobserver agreement for all findings between resident and consultant reports was almost perfect (weighted κ = 0.87; 95% confidence interval, 0.79-0.96; P < .0001). The overall discrepancy rate, including both false-positive and false-negative findings, between the on-call radiology resident and consultant radiologist was 8% (eight of 100).

CTPA reports by radiology residents can be relied and acted upon without any major discrepancies. There is a relatively much higher proportion of patients with alternative diagnoses, mainly infective consolidation and heart failure presenting with similar symptoms and signs as pulmonary emboli. It is imperative for trainees to be systematic and review all images if observational omissions are to be reduced.