Kathleen Eddy1, Adam Beaton2, Richard Eddy3, John Mathieson3. 1. Department of Diagnostic Radiology, Dalhousie University, Halifax, Nova Scotia, Canada. Electronic address: kieddy@dal.ca. 2. Information Management, Island Health, Victoria, British Columbia, Canada. 3. Department of Medical Imaging, Island Health, Victoria, British Columbia, Canada.
Abstract
PURPOSE: A study was performed to evaluate the ACR Select software in determining the level of appropriateness of computed tomography (CT) and magnetic resonance imaging (MRI) in Island Health in British Columbia. METHODS: A total of 1228 consecutive CT and MRI studies performed in a 3-day period were entered into a software program provided by the National Decision Support Company based on the ACR Appropriateness Criteria. The program was able to analyze 93% (1141) of these studies. A subset of these requisitions was manually reviewed. RESULTS: The software program demonstrated a very low 2.5% inappropriate rate and manual review showed an even lower number of 0.6%. In a sample of studies deemed to be appropriate by the software, manual review agreed with this ranking in all cases. In addition, in 20% of cases where CT was done, the software program suggested that MRI would be a more appropriate choice. CONCLUSIONS: First, the ACR Select software is a useful tool to assess appropriateness of CT and MRI, although it may underestimate the level of appropriateness of studies labeled as inappropriate. Second, CT and MRI are being ordered appropriately in Island Health in British Columbia. The software recommendation of MRI as more appropriate in 20% of cases where CT was done suggests a lack of MRI resources in Island Health.
PURPOSE: A study was performed to evaluate the ACR Select software in determining the level of appropriateness of computed tomography (CT) and magnetic resonance imaging (MRI) in Island Health in British Columbia. METHODS: A total of 1228 consecutive CT and MRI studies performed in a 3-day period were entered into a software program provided by the National Decision Support Company based on the ACR Appropriateness Criteria. The program was able to analyze 93% (1141) of these studies. A subset of these requisitions was manually reviewed. RESULTS: The software program demonstrated a very low 2.5% inappropriate rate and manual review showed an even lower number of 0.6%. In a sample of studies deemed to be appropriate by the software, manual review agreed with this ranking in all cases. In addition, in 20% of cases where CT was done, the software program suggested that MRI would be a more appropriate choice. CONCLUSIONS: First, the ACR Select software is a useful tool to assess appropriateness of CT and MRI, although it may underestimate the level of appropriateness of studies labeled as inappropriate. Second, CT and MRI are being ordered appropriately in Island Health in British Columbia. The software recommendation of MRI as more appropriate in 20% of cases where CT was done suggests a lack of MRI resources in Island Health.