Hospital administrators aiming to go filmless may have a new arrow in their quiver, thanks to a recent study in the American Journal of Roentgenology. Researchers from Philadelphia found soft-copy interpretation as reliable as hard-copy in the emergency department.
A group from the department of radiology at the University of Pennsylvania Medical Center gathered radiographs from a sampling of 100 patients seen in the ER. Only images obtained during the day shift were included in the study. Images were categorized as chest, abdomen, spine and pelvis, or extremities, and all of the standard views were included with each case.
The images were obtained using CR (STIII, Fuji Medical Systems, Stamford, CT) plates and were printed to film (DryView 8700, Eastman Kodak Health Imaging, Rochester, NY) for interpretation. Seven staff radiologists, who either interpreted emergency department cases or rotated through the department’s reading room, read the cases. None had given the primary interpretation or had reviewed any of the cases (AJR September 2001, Vol. 177:3, pp. 525-528).
The cases were divided into 10 groups of 10 cases, and the radiologists were asked to decide if the findings were positive, equivocal, or negative. If the results were deemed equivocal, the radiologist was asked to rule if the decision was due to technical or diagnostic reasons.
The same group of radiologists reviewed the cases six months later on two 2K x 2.5K display workstations (Advantage, GE Medical Systems, Waukesha, WI), and were free to use any of the image-processing functions available on the system.
The research team then used mixture distribution analysis to compute the relative percentage of agreement between the hard-copy and soft-copy interpretations, in order to calculate a measure of reliability on the basis of observer performance. Reliability was defined as the likelihood of agreement of one typical observer, relative to the interpretation of the majority by the study authors.
The results of the mixture distribution analysis showed a relative percentage of agreement by the radiologists of 90.1% for hard copy interpretations and 89.8% for soft-copy interpretations. The team did note that when cases were interpreted using hard copy, the proportion of negative cases was slightly lower and the proportion of positive cases slightly higher.
By Jonathan S. Batchelor
AuntMinnie.com staff writer
October 1, 2001
Related Reading
BJR article recommends international viewing quality standards, September 16, 2001
Image quality is still job one for PACS, August 31, 2001
Strategic plan allows smooth transition from film library to image service center, August 29, 2001
Soft-copy review of CT studies bolsters productivity, April 26, 2001
Baltimore VA survey says human factors count in soft-copy reading, August 29, 2000
Copyright © 2001 AuntMinnie.com
