Eduard Michel MD, Presenter: Officer, Virtual Radiologic Corporation Shareholder, Virtual Radiologic Corporation

Julie Shields, Abstract Co-Author: Employee, Virtual Radiologic Corporation

Fatigue has been implicated as a factor in diagnostic radiology interpretation accuracy. We investigated the incidence of discrepancies at the beginning and at the end of our radiologists’ work shifts and found that the timing of the reads during their shifts did not negatively impact the accuracy of interpretations, despite a demanding schedule, which consists of overnight shifts. Additional factors which may positively impact our results include lack of interruptions and dedicated support staff which allow for a smooth workflow for our radiologists.

Fatigue has been implicated as a factor affecting the accuracy of interpretations by radiologists (Krupinski et al, 2009). This study measured the discrepancy rate for diagnostic interpretations in a teleradiology practice at two different time points during the radiologists’ shifts.

A random sample of 19 radiologists with identical work schedules during 2009 was included. All radiologists worked a fixed schedule of 7 nights on (9pm-7am) followed by 7 days off duty. The overall distribution of study assignments remained constant during the study period. All cases were preliminary emergent interpretations. Each case was overread by a local radiologist who provided a final report. Conflicting reports were reviewed by the QA committee and categorized as: no discrepancy, minor discrepancy with no impact on patient care, minor discrepancy with impact on patient care and significant discrepancy. The discrepancy results were correlated with the time of interpretation based on the electronic signature time in our RIS

During 2009 the 19 radiologists included in the study read 343,587 studies. Of these, 60,816 cases (17.7%) were read during the 9-11pm period and 61,595 cases (17.9%) during the 5-7am period. A total of 1,804 cases were submitted as discrepancies. Of these 673 (37.3%) were judged to be without merit. Of the remaining discrepancies 167 (14.7%) were for cases read during the 9-11pm period and 208 (18.4%) for cases read during the 5-7am period

Michel, E, Shields, J, Assessment of Interpretation Accuracy as a Function of Timing during Radiologists' Shift.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9015641.html