From the Executive Director: An Explanation of ABR Oral Exam Scoring
By ABR Executive Director Brent Wagner, MD, MBA, and Associate Executive Directors Mary S. Newell, MD (Diagnostic Radiology), James B. Spies, MD, MPH (Interventional Radiology), Geoffrey S. Ibbott, PhD (Medical Physics), and Michael Yunes, MD (Radiation Oncology)
June 2026;19(3):3
The ABR has used a standard oral exam scoring process for decades. The model is a psychometrically reliable and reproducible method that aggregates performance across multiple cases and a range of categories for each examinee.
ABR oral exams are one-on-one videoconference sessions between a subject matter expert (examiner) and a candidate for certification (examinee). A composite score is generated by an individual examiner for a specific examinee for each of these exam periods, during which the examinee is tested on three to 10 cases, depending on the overall complexity of the content. For interventional radiology (IR), radiation oncology (RO), and diagnostic radiology (DR), the exam period corresponds to a content category (e.g., arterial diagnosis and intervention in IR, gynecological in RO, neuroradiology in DR). For medical physics (MP), a period corresponds to interactions with a single examiner, but the cases represent five distinct (and separately scored) categories within the discipline. There are four categories in IR, eight in RO, seven in DR, and five in MP.
Performance on an individual case is scored as 68, 69, 70, 71, or 72. Although a candidate does not need to pass every case to pass the category, the average of the case scores must equal or exceed 70 to pass the category. Hence, a candidate who is shown five cases and receives scores of 69, 70, 70, 70, and 71 would pass because the average of those scores is 70.
Examiners are trained to assess performance relative to a standard of competence, and standardization (by minimizing interobserver variability) is a specific goal of the examiner orientation process.
Although specific performance elements will vary among cases and the disciplines, the general definitions and numerical correlates of performance on individual cases or scenarios are presented here. More discipline-specific information and an opportunity to ask questions are provided for candidates at an orientation webinar prior to each exam.
A 72 represents exceptionally strong performance (“outstanding”). This would include a confident and concise discussion that reflects a superlative understanding of the elements of the case, which might include technical features, understanding of anatomy and disease, safe practice, and a clearly stated rationale addressing the important elements of the presented scenario.
A 71 represents a strong performance (“good”). This would characterize a pattern of responses to a case that is better than acceptable and represents a reasonable and well-presented analysis with few knowledge gaps.
A 70 represents acceptable performance (“satisfactory”). This is the passing standard and is generally understood to correspond to the examiner’s confidence that the candidate has the knowledge, skill, and understanding for independent practice that is safe and effective. The satisfactory candidate demonstrates and applies substantial understanding of fundamental principles but may lack highly detailed depth or breadth of knowledge.
A 69 represents weak performance (“marginally unsatisfactory”). This score would be assigned to a candidate’s case discussion that fails to convey an understanding of important and fundamental concepts that are needed for safe and independent practice.
A 68 represents poor performance that is potentially unsafe (“absolutely unsatisfactory”). This score is uncommon and is typically a manifestation of significant knowledge gaps or poor judgment that may also be apparent in other cases.
The exam software averages the scores for a particular category; if the average is 70 or above, the category is considered to have been passed. Average scores below 70 indicate a fail in that category. To mitigate potential bias, examiners document their scoring determination without knowledge of the scores assigned by other examiners who met with the same candidate. Very rarely, markedly poor performance in a particular case (for example, insisting on a diagnosis or course of action that is potentially dangerous) is of sufficient concern that the category score is a fail regardless of performance on the other cases.
Candidates who fail one or two categories (depending on their specialty) will be discussed in a panel meeting among the group of examiners who met with that candidate. Based on this holistic view of the candidate’s performance, failing scores may be revised to a pass (but passing scores are never revised to fail).
A pass is awarded to candidates who pass all categories. Depending on the discipline, failing one or two categories may result in a conditional pass. If the candidate has a conditional pass, they will have to retake those portions of the exam they failed — not the entire exam. A conditional pass is not possible in IR because the exam consists of only four sections and performance below the standard in a single section means that a substantial portion of the exam material was not mastered.
The ABR seeks to support the integrity and validity of the oral exam process through a robust and fair process that reinforces the value of the certificate as an indicator of the attainment of knowledge and skill that serves the public and the profession.
