These tools will no longer be maintained as of December 31, 2024. Archived website can be found here. PubMed4Hh GitHub repository can be found here. Contact NLM Customer Service if you have questions.


Pubmed for Handhelds

PUBMED FOR HANDHELDS

Search MEDLINE/PubMed

  • Title: Success with voice recognition.
    Author: Sferrella SM.
    Journal: Radiol Manage; 2003; 25(3):42-9. PubMed ID: 12817421.
    Abstract:
    You need a compelling reason to implement voice recognition technology. At my institution, the compelling reason was a turnaround time for Radiology results of more than two days. Only 41 percent of our reports were transcribed and signed within 24 hours. In November 1998, a team from Lehigh Valley Hospital went to RSNA and reviewed every voice system on the market. The evaluation was done with the radiologist workflow in mind, and we came back from the meeting with the vendor selection completed. The next steps included developing a business plan, approval of funds, reference calls to more than 15 sites and contract negotiation, all of which took about six months. The department of Radiology at Lehigh Valley Hospital and Health Network (LVHHN) is a multi-site center that performs over 360,000 procedures annually. The department handles all modalities of radiology: general diagnosis, neuroradiology, ultrasound, CT Scan, MRI, interventional radiology, arthography, myelography, bone densitometry, nuclear medicine, PET imaging, vascular lab and other advanced procedures. The department consists of 200 FTEs and a medical staff of more than 40 radiologists. The budget is in the $10.3 million range. There are three hospital sites and four outpatient imaging center sites where services are provided. At Lehigh Valley Hospital, radiologists are not dedicated to one subspecialty, so implementing a voice system by modality was not an option. Because transcription was so far behind, we needed to eliminate that part of the process. As a result, we decided to deploy the system all at once and with the radiologists as editors. The planning and testing phase took about four months, and the implementation took two weeks. We deployed over 40 workstations and trained close to 50 physicians. The radiologists brought in an extra radiologist from our group for the two weeks of training. That allowed us to train without taking a radiologist out of the department. We trained three to six radiologists a day. I projected a savings of 5.0 FTEs over two years. The actual savings were 8.0 FTEs within three weeks for the first phase and an additional 4.3 FTEs within two weeks of the second phase. The transcription staff was retained to perform other types of transcription and not displaced. The goal was to reduce Medical Records' outsourcing expenses by $670,000 over three years. The actual savings are in excess of $900,000. The proposed payback period was 17 months, and the actual was less than 12 months. For two years prior to implementing the voice system, the turnaround time at Lehigh Valley was 41 percent within 24 hours. One week after implementation, the turnaround time was 78 percent within 24 hours. Today it ranges between 85 percent and 92 percent. Overall, the radiologists at Lehigh Valley Hospital did an excellent job with the cultural change to voice recognition. It has made a major impact on our ability to get reports to physicians in a timely manner so they can make treatment decisions.
    [Abstract] [Full Text] [Related] [New Search]