Richard Joseph Thomas Gorniak MD, Abstract Co-Author: Speaker, Koninklijke Philips NV

Adam Eugene Flanders MD, Abstract Co-Author: Nothing to Disclose

Christopher Geordie Roth MD, Presenter: Author, Reed Elsevier

Paras Lakhani MD, Abstract Co-Author: Nothing to Disclose

Vijay Madan Rao MD, Abstract Co-Author: Nothing to Disclose

This system increases the educational value of peer review data captured during the normal course of clinical operations, potentially reducing future errors.

Peer review is commonplace in radiology practices and is necessary to meet regulatory requirements. While there are differing methodologies for collecting and presenting peer review, typical systems include data gathering, validation, benchmarking and individual feedback. Beyond individual feedback, peer review data is sometimes used more broadly for education such as in “missed case” conferences or teaching files. While these educational methods have utility, they do not emulate the actual image interpretation process.

Using a PACS-integrated tool, our institution’s peer review database was queried for discrepancies occurring in the last 2 years and the discrepant cases were reviewed for educational value. The entire DICOM file set of 45 selected cases and reports were extracted from PACS and anonymized. The anonymized DICOM files were then stored on a nonclinical PACS server. Using the PACS javascript API, a user interface was made to list cases, display images and history, show the report and recorded discrepancy, and record user self-assessment.

Peer review data offers a rich source of educational material highlighting real world mistakes of attending radiologists. While feedback to the individual who committed the error is valuable, this project increases the educational value of peer review by allowing multiple trainees and attendings to learn from an error. By reviewing the mistakes of others, radiologists may be able to avoid making similar future mistakes. Unlike a typical teaching file or conference, this system more closely simulates normal practice where the radiologist has the entire image set to evaluate instead of a few key images, mimicking the condition of the original reader. There is likely a greater potential for retention by actively reviewing the error as an unknown in a familiar reading-room workflow, rather than viewing cases passively in a conference. This system accomplishes this, while keeping both the radiologist who erred and the radiologist who flagged the error anonymous.

Gorniak, R, Flanders, A, Roth, C, Lakhani, P, Rao, V, Learning from the Mistakes of Others: Utilizing Peer Review Data for Education via a PACS-based Teaching System.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14002349.html