Elizabeth Anne Krupinski PhD, Presenter: Nothing to Disclose

Kevin Staley Berbaum PhD, Abstract Co-Author: Nothing to Disclose

Robert Caldwell, Abstract Co-Author: Nothing to Disclose

Kevin Schartz PhD, Abstract Co-Author: Nothing to Disclose

Mark T. Madsen PhD, Abstract Co-Author: Nothing to Disclose

Our previous studies demonstrated statistically significant decreases in fracture detection accuracy after a long day of clinical reading compared to prior to a day of reading. This study recorded eye-position to determine if efficiency of visual search during interpretation is degraded after a long day of clinical reading, thus contributing to interpretation degradation.

Fifteen plain film cases were used: 5 fracture (fx) free, 5 subtle fxs, 5 moderately subtle fxs. Six radiologists viewed the images twice, once before a day of clinical reading (Early) and once after (Late). Eye position was recorded to characterize: time to first fixate a fracture, total search time, and dwells associated with each decision type (true and false, positive and negative).

Total viewing times were no fx Early = 48. 67 sec; no fx Late = 52.20; subtle fx Early = 45.47 subtle fx Late = 51.13; moderate fx Early = 39.03; moderate fx Late = 44.27. Although times varied significantly by fx type (p = 0.0342), there was no difference for Early vs Late reading (p = 0.0934). Cumulative dwells as well as times to first hit/fixate the fractures for true and false, positive and negative decisions did not differ significantly for Early vs Late reading.

Eye position data revealed that although there were trends towards degraded (i.e., less efficient) visual search after a long day of reading, it does not appear to be the mechanism underlying degraded interpretation accuracy after a long day of clinical reading. Radiologists are reading more studies, containing more images, placing greater demands on them, leading to increased fatigue levels and documented reduction in diagnostic accuracy. Inadequate knowledge about the frequency, cause and impact of medical errors, as well as effective methods for error prevention is a major obstacle for improving the quality of health care.

Degraded visual search does not play a significant role in fatigue-related errors. Future research will explore cognitive fatigue as a mechanism contributing to degraded interpretation accuracy.

Krupinski, E, Berbaum, K, Caldwell, R, Schartz, K, Madsen, M, Impact of Radiologists' Fatigue on Visual Search during the Detection of Bone Fractures.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12024772.html