Richard Earnest Sharpe MD, MBA, Presenter: Nothing to Disclose

Ronald Scott Winokur MD, Abstract Co-Author: Nothing to Disclose

Richard Joseph Thomas Gorniak MD, Abstract Co-Author: Consultant, Bracco Group

Levon N. Nazarian MD, Abstract Co-Author: Nothing to Disclose

Adam E. Flanders MD, Abstract Co-Author: Research Consultant, Geron Corporation

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

Academic radiology attending physicians finalize preliminary reports submitted by radiology residents. Some attending physicians edit reports more than others. The variation of attending physician editing behaviors is not well understood. The purpose of this study is to quantify the variability of attending physician editing behaviors.

Preliminary and finalized radiology reports between from 3/14/2010 to 7/13/2010 and from 12/31/2010 to 2/12/2011 were extracted from the Radiology Information System (RIS) for analysis. Reports were analyzed using open source document comparison software. The Levenshtein distance (LD), the number of character changes required to convert a preliminary report into the finalized version, was calculated. A Levenshtein Percent (LP) was defined as the (LD)/(Final Report Character Length), and normalized for report length variations. Average LP and standard deviation were calculated for each attending radiologist. A line graph was plotted for each attending LP value to facilitate comparison to their peers.

157,745 reports from 48 attendings were included in the analysis. The mean LP for the attending staff is 16.97 with standard deviation of 33.81.The attending LP scores ranged from 2.1 to 85.1. Four attending radiologists had average LPs of 85, 62, 61, and 58. These physicians made many changes to nearly every resident’s reports. Eight attending physicians made very few edits to all resident’s reports, and their LP averaged 2.1, 4.2, 5.2, and 5.8.

There is significant variability in the extent to which attending physicians edit resident preliminary reports. LP analysis may identify attendings with high editing behaviors. Some attendings make enormous numbers of changes to nearly all preliminary reports written by all residents. Such widespread report changes may not all be clinically significant. Since time is required to edit reports, excess editing time represents an increasing relative cost of attending productivity. Such time might be more productively spent on other RVU generating tasks.

Analysis of attending radiologist report editing behaviors can identify outlier attendings that utilize their time editing at the expense of additional productivity.

Sharpe, R, Winokur, R, Gorniak, R, Nazarian, L, Flanders, A, Rao, V, Variations in Attending Physician Reporting Behaviors Suggest Opportunities for Improving Staff Productivity.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11015402.html