Transforming from Radiologist Peer Review Audits to Peer Learning and Improvement Approaches

Abstract

All radiologists actively practicing in the United States are required to undergo some manner of periodic performance evaluation. This should provide an unbiased, fair, and balanced evaluation of radiologist performance to identify opportunities for additional education, error reduction, and self-improvement. By far the most common method of peer review audit system currently used in radiology is RADPEER, developed almost 15 years ago by the American College of Radiology (ACR), in which originally interpreted images are randomly selected and reviewed by a peer radiologist. However, studies have shown that this time-consuming process has inherent sampling bias, has limited value as an educational tool, and is primarily performed to meet accreditation and hospital credentialing requirements. Moreover, it evaluates the performance of a radiologist in terms of a diagnostic discrepancy score, excluding the myriad of other functions and roles that radiologists play, including teaching, consulting, and communicating abnormal results. Consequently, an increasing number of radiology practices are embracing simple scoring systems that either agree with the prior read or score the interpretation as an “apparent learning case.”

Rather than a scoring-based peer review audits of random cases for evaluating radiologist performance, this chapter recommends the adoption of a system based on “peer learning, which consists of peer feedback, learning, and improvement. The goal is not to identify poor-performing physicians, but to improve performance of all members of the group by analyzing the potential contributors to errors through a self-reflection process, as well as peer discussion in a constructive, nonpunitive quality improvement meeting.