RSNA 2005 

Abstract Archives of the RSNA, 2005


SSG19-04

Long-term Impact of PACS/RIS Integrated Speech Recognition on Radiology Report Availability: 3-year Follow-up Study

Scientific Papers

Presented on November 29, 2005
Presented as part of SSG19: Health Services, Policy, and Research (Picture Archiving and Communication Systems, Teleradiology)

Participants

Christoph Gregor Trumm MD, Presenter: Nothing to Disclose
Christian Glaser MD, Abstract Co-Author: Nothing to Disclose
Petra Popp, Abstract Co-Author: Nothing to Disclose
Benjamin Küttner, Abstract Co-Author: Nothing to Disclose
Sven Nissen-Meyer, Abstract Co-Author: Nothing to Disclose
Maximilian Ferdinand Reiser MD, Abstract Co-Author: Nothing to Disclose

PURPOSE

To quantify the long-term effect of a PACS/RIS integrated speech recognition system on radiology report availability (RA) in a university hospital over a 3-year period.

METHOD AND MATERIALS

In a retrospective PACS/RIS database analysis, report turnaround times (RTT) were calculated for 3 comparable time periods: 1. 11/02-02/03 (only tape-based dictation with transcription), 2. 11/03-02/04 (0.5 year experience with speech recognition system[SRS]), 3. 11/04-02/05 (1.5 year experience with SRS). Only remote (RSR) and online (OSR) speech recognition with transcriptionist and self-correction respectively were used in period 2 and 3. Cumulative RA (ratio of available RIS/HIS reports vs. total number of examinations) within 2/6/12h after examination completion was calculated for period 1/2/3 respectively including all modalities. For period 3, mean RTT was computed and cumulative short-term RA differentiated into reports available 10/30/60 minutes after completion of conventional radiography (CR), computed tomography (CT) and ultrasound (US) examinations.

RESULTS

37898/39680/42565 datasets were analyzed for time period 1/2/3, the number of reports available in less than 24h was 3485/11453/24753 respectively. Including all modalities, cumulative RA within 2/6/12h after examination completion was 2.4/6.4/9.0% for period 1, 13.6/24.5/28.6% for period 2 and 33.5/53.9/58.0% for period 3. Mean RTT in period 3 was 15.8/21.4/26.3h for 24285 CR/8782 CT/3050 US examinations. Cumulative short-term RA after 10/30/60 minutes was 3.3/21.1/34.9% for CR, 1.4/7.5/15.6% for CT and 44.6/62.1/66.4% for US examinations.

CONCLUSION

Institution-wide use of a PACS/RIS integrated SRS over a 1.5-year period facilitates a continuous improvement of report availability. Reasons are an increasing acceptance of OSR by radiologists and shift from RSR to OSR for creation of routine reports. While mean RTT is still comparably high for high turnover modalities like CR, CT and US, ratio of instantly available reports is growing.

DISCLOSURE

P.P.: Siemens AG Medical Solutions, Director of Product Management & Innovations, Erlangen / Germany

Cite This Abstract

Trumm, C, Glaser, C, Popp, P, Küttner, B, Nissen-Meyer, S, Reiser, M, Long-term Impact of PACS/RIS Integrated Speech Recognition on Radiology Report Availability: 3-year Follow-up Study.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4409992.html