RADPEER™ Scoring White Paper

Abdominal and pelvic computed tomography (CT) interpretation: discrepancy rates among experienced radiologists

“…Similar criteria are used by the RADPEER program, which is an online peer-review system developed by the American College of Radiology. The RADPEER task force [3] recently updated the definitions of their scoring language as follows: 1 = concur with interpretation, 2 = discrepancy in interpretation/not ordinarily expected to be made (understandable miss), 3 = discrepancy in interpretation/should be made most of the time, and 4 = discrepancy in interpretation/should be made almost every time-misinterpretation of finding. In each of the scores (except for the score of 1), there are two optional scores available according to whether the reviewer thinks the finding is unlikely or likely to be clinically significant.…”

“…Major disagreement with clinical significance and potential change to patient treatment plan -Typographical or dictation errors that would significantly change the meaning of the reports (e.g., leaving out the word "not" in front of "malignant"), even if the report body contradicted the impression/conclusion -Stable versus more than minimal interval change of tumor -Normal liver versus heterogeneous liver with a recommendation for MRI to evaluate for possible liver disease -Similar description of findings, however one radiologist did not state potential metastasis and another did -Normal colon versus abnormally thickened sigmoid colon with a recommendation for C. difficile assay for possible colitis -No recommendation versus with recommendation for potentially serious condition -Presence or absence of recommendation for a renal mass with calcification on noncontrast CT -Recommendation versus no recommendation for MRI to evaluate small liver lesions in a patient with known malignancy -Absence or presence of a liver lesion in a patient with known malignancy -Stable versus progression of metastasis their definition [3]. Soffa et al [12] reviewed a double reading of 2% of their daily cases (a total of 6,703 examinations read by 26 radiologists) and showed an overall disagreement rate of 3.03-5.79% in the interpretation of general radiology, diagnostic mammography, screening mammography, and ultrasound.…”

“…Two radiologists independently (blindly) interpret the same examination; then they resolve the disagreement by consensus. Another form of peer review can be performed by giving an opinion on the degree of disagreement/discrepancy of the previous interpretation (old examinations) while performing a routine interpretation of the current study [3]. Discrepancy rates of radiological interpretation range widely (2-29%), degree of severity (mild to severe), and potential consequences to the patients [4][5][6].…”

Emergency CT brain: preliminary interpretation with a tablet device: image quality and diagnostic performance of the Apple iPad

“…These two interpretations were then compared, and discrepancies were recorded and were scored under peer review according to the American College of Radiology's RADPEER scoring system (Fig. 3) [6].…”

Survey of Radiologist Attitudes and Perceptions Regarding the Incorporation of a Departmental Peer Review System

“…Numerous audits of radiologist performance have demonstrated error rates in the range of 3% to 5% for all errors in daily practice [6][7][8][9][10]. In response to concerns about quality improvement, the ACR developed the RADPEERÔ system in 2002 [11]. The ACR has made the implementation of formal systems of peer review a key component of hospital accreditation.…”