Patrick H Do MD, Presenter: Nothing to Disclose

Mingming Ma MD, Abstract Co-Author: Nothing to Disclose

Matthew Alexander MD, Abstract Co-Author: Nothing to Disclose

Benjamin Boun-Ming Chou MD, Abstract Co-Author: Nothing to Disclose

Young S. Kang MD, Abstract Co-Author: Nothing to Disclose

Rajul Parimal Pandit MD, Abstract Co-Author: Nothing to Disclose

Mahesh Ramu Patel MD, Abstract Co-Author: Stockholder, Novartis AG

Automated report generation improves accuracy and conformity with accepted guidelines, as well as improving radiologist efficiency. 

Classification of lumbar disc pathology identified on MRI and the nomenclature for radiology reporting have historically been variable and difficult to interpret. Professional societies, including ASNR, developed a consensus scheme for nomenclature and classification. With contemporary technology, a role for electronically generated standardized reporting has emerged. We developed and evaluated a customized, web-based program for automated generation of reports using consensus nomenclature. 

A web-based structured report generator tailored to lumbar MRI without requiring input of protected health information was created using software implemented on a Linux-based Apache webserver. The involves mySQL databases, AJAX scripting, and additional open source tools. After selection of relevant responses for each field at every spinal level, the program outputs text in a format appropriate for the imaging report that can be simply pasted into the electronic dictation system. A retrospective review was conducted to compare reports generated with and without use of the program. Conformity was evaluated, and resident and attending physician satisfaction with the program was evaluated with a survey. 50 reports each were evaluated with and without program implementation. Conformity with consensus nomenclature was found in 50% of reports without implementation and 96% with implementation. 10 residents and 4 attendings were surveyed. 12 of 14 (86%) users indicated the program increased efficiency and decreased ambiguity of reports compared to the previously utilized macro-based scheme. 2 of 14 (14%) had no preference between the two schemes. 9 of the surveyed residents (90%) indicated improved inter-attending concordance following program implementation.  

Appropriate use of consensus classification and nomenclature aids clinicians in managing disease and assessing response to treatment. The above-described program improves conformity in reporting and increases radiologist efficiency. This process found widespread acceptance by surveyed radiologists. 

Do, P, Ma, M, Alexander, M, Chou, B, Kang, Y, Pandit, R, Patel, M, Web-Based Interactive Generation of Structured Reports for Lumbar Spine MRI Improves Conformity with Consensus Reporting Guidelines.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14009072.html