Jordan M Brown1, Elliot C Dickerson1, Lee C Rabinowitz2, Richard H Cohan1, James H Ellis1, John M Litell3, Ravi K Kaza1, Alexis N Lopez2, Nikhil R Theyyunni3, Joseph T Weber4, Keith E Kocher5, Matthew S Davenport6. 1. Department of Radiology, University of Michigan Health System, Ann Arbor, Michigan. 2. Department of Internal Medicine, University of Michigan Health System, Ann Arbor, Michigan. 3. Department of Emergency Medicine, University of Michigan Health System, Ann Arbor, Michigan. 4. Pulmonary and Critical Care Associates SC, Cudahy, Wisconsin. 5. Department of Emergency Medicine, University of Michigan Health System, Ann Arbor, Michigan; Institute for Healthcare Policy and Innovation, Ann Arbor, Michigan. 6. Department of Radiology, University of Michigan Health System, Ann Arbor, Michigan; Michigan Radiology Quality Collaborative, Ann Arbor, Michigan. Electronic address: matdaven@med.umich.edu.
Abstract
PURPOSE: To determine whether resident abdominopelvic CT reports considered prospectively concordant with the final interpretation are also considered concordant by other blinded specialists and abdominal radiologists. METHODS: In this institutional review board-approved retrospective cohort study, 119 randomly selected urgent abdominopelvic CT examinations with a resident preliminary report deemed prospectively "concordant" by the signing faculty were identified. Nine blinded specialists from Emergency Medicine, Internal Medicine, and Abdominal Radiology reviewed the preliminary and final reports and scored the preliminary report with respect to urgent findings as follows: 1.) concordant; 2.) discordant with minor differences; 3.) discordant with major differences that do not alter patient management; or 4.) discordant with major differences that do alter patient management. Predicted management resulting from scores of 4 was recorded. Consensus was defined as majority agreement within a specialty. Consensus major discrepancy rates (ie, scores 3 or 4) were compared to the original major discrepancy rate of 0% (0/119) using the McNemar test. RESULTS: Consensus scores of 4 were assigned in 18% (21/119, P < .001, Emergency Medicine), 5% (6/119, P = .03, Internal Medicine), and 13% (16/119, P < .001, Abdominal Radiology) of examinations. Consensus scores of 3 or 4 were assigned in 31% (37/119, P < .001, Emergency Medicine), 14% (17/119, P < .001, Internal Medicine), and 18% (22/119, P < .001, Abdominal Radiology). Predicted management alterations included hospital status (0-4%), medical therapy (1%-4%), imaging (1%-10%), subspecialty consultation (3%-13%), nonsurgical procedure (3%), operation (1%-3%), and other (0-3%). CONCLUSIONS: The historical low major discrepancy rate for urgent findings between resident and faculty radiologists is likely underreported.
PURPOSE: To determine whether resident abdominopelvic CT reports considered prospectively concordant with the final interpretation are also considered concordant by other blinded specialists and abdominal radiologists. METHODS: In this institutional review board-approved retrospective cohort study, 119 randomly selected urgent abdominopelvic CT examinations with a resident preliminary report deemed prospectively "concordant" by the signing faculty were identified. Nine blinded specialists from Emergency Medicine, Internal Medicine, and Abdominal Radiology reviewed the preliminary and final reports and scored the preliminary report with respect to urgent findings as follows: 1.) concordant; 2.) discordant with minor differences; 3.) discordant with major differences that do not alter patient management; or 4.) discordant with major differences that do alter patient management. Predicted management resulting from scores of 4 was recorded. Consensus was defined as majority agreement within a specialty. Consensus major discrepancy rates (ie, scores 3 or 4) were compared to the original major discrepancy rate of 0% (0/119) using the McNemar test. RESULTS: Consensus scores of 4 were assigned in 18% (21/119, P < .001, Emergency Medicine), 5% (6/119, P = .03, Internal Medicine), and 13% (16/119, P < .001, Abdominal Radiology) of examinations. Consensus scores of 3 or 4 were assigned in 31% (37/119, P < .001, Emergency Medicine), 14% (17/119, P < .001, Internal Medicine), and 18% (22/119, P < .001, Abdominal Radiology). Predicted management alterations included hospital status (0-4%), medical therapy (1%-4%), imaging (1%-10%), subspecialty consultation (3%-13%), nonsurgical procedure (3%), operation (1%-3%), and other (0-3%). CONCLUSIONS: The historical low major discrepancy rate for urgent findings between resident and faculty radiologists is likely underreported.