Tarik K. Alkasab MD, PhD, Presenter: Nothing to Disclose

H. Benjamin Harvey MD, JD, Abstract Co-Author: Nothing to Disclose

David Andrew Rosman MD, Abstract Co-Author: Advisory Board, UnitedHealth Group

Keith J. Dreyer DO, PhD, Abstract Co-Author: Author, Springer Science+Business Media Deutschland GmbH Author, HealthMU Press Editorial Advisory Board, UBM plc Editorial Advisory Board, Imaging Economics Editorial Advisory Board, AuntMinnie.com Medical Advisor, Agfa-Gevaert Group Medical Advisor, Amirsys, Inc Medical Advisor, Carestream Health, Inc Medical Advisor, lifeIMAGE Inc Medical Advisor, Merge Healthcare Incorporated Medical Advisor, Nuance Communications, Inc Medical Advisor, RCG HealthCare Consulting

Daniel Ira Rosenthal MD, Abstract Co-Author: Nothing to Disclose

Giles W. Boland MD, Abstract Co-Author: Nothing to Disclose

Point-of-care CDS software tools can efficiently assist radiologists in the real-time application of standard imaging algorithms to achieve best practices.

Evidenced-based algorithms exist that guide radiologists to make appropriate report recommendations.  Yet, we have separately presented data suggesting poor compliance with these best practice guidelines. To address this problem, we have developed a workstation-integrated, point-of-care Clinical Decision Support (CDS) system that guides radiologists in applying best practices.

The CDS system is activated by radiologists when clinically appropriate and assists the radiologist to use structured descriptions and standardized recommendation language that conform to evidence-based guidelines. The system consists of three layers. First, at the server layer, a database-backed web application applies criteria-driven algorithms to case-specific imaging findings and applicable patient information, such as demographic data and clinical context extracted from the RIS and voice recognition system. Second, a client layer runs within a web browser to mediate the radiologist interaction with the system. Finally, a "glue" layer integrates the system with the voice recognition software and inserts the standardized language into the “Findings”, “Impression”, and a separate “Recommendation” field. For the pilot trial, we deployed this CDS system on a subset of workstations in our abdominal imaging division and limited the application of the tool to consensus-backed criteria for adrenal and pulmonary nodules. Over the 23 week trial, radiologists used the tool to generate recommendations 330 times. Users reported that the CDS tool facilitated appropriate application of imaging criteria without negatively impacting efficiency.  

Point-of-care CDS software tools can efficiently assist radiologists in the real-time application of evidence-based algorithms and can be integrated into the reporting workflow. We expect that use of the tool will increase as more criteria-based algorithms are applied, resulting in improved consistency and accuracy of radiologist recommendations.

Alkasab, T, Harvey, H, Rosman, D, Dreyer, K, Rosenthal, D, Boland, G, Conforming to Best Practice Standards: Development of a Software System to Provide Radiologists with Point-of-Care Decision Support for Recommendations.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13018065.html