This website is intended for healthcare professionals only.

Newsletter      
Hospital Healthcare Europe
HOPE LOGO
Hospital Healthcare Europe

Speech recognition technology – a sound investment to help provide seamless clinical reports


22 July, 2010  

Great Ormond Street Hospital in London embraces the advantages of new technology, making savings in labour and resources

Dr Melanie Hiorns
Consultant Radiologist
Clinical Unit Chair
Great Ormond Street Hospital
London, UK

Great Ormond Street Hospital is a specialised paediatric hospital, taking care of patients with complex and rare diseases. We treat children from the UK and throughout the world – less than 50% of the patients are from within the boundaries of London. In 95% of patients, the radiology findings are abnormal. This poses a number of challenges to us, particularly in rapid transcription management, with reports having to be highly accurate and yet available to the clinician without delay.
We have found that it may take a transcriptionist a considerable period of time to reach the levels of typing and understanding of terminology required and as the administrative area traditionally has a relatively high turnover of staff, this can be a problem. Speech recognition technology was a clear solution for us.
Great Ormond Street Hospital for Children NHS Trust (GOSH) uses Soliton IT’s Workflow Manager, powered by SpeechMagic from Nuance to bridge the gap between our Radiology Information System (RIS) and our Picture Archiving and Communications System (PACS) and provide seamless front-end speech recognition for clinical reports.
The Soliton product takes into account crucial issues such as work patterns and patient safety and covers every type of reporting case via various modules.
Our system at GOSH is based on unique accession numbers to link RIS data and PACS images, thus avoiding confusion and ensuring accurate reporting. With speech recognition on top, productivity for plain film X-ray reporting increased by more than 40%.
GOSH is an NHS Trust that was early to embrace the advantages of new technology and was among the first in the UK to introduce PACS, with a Siemens system, and then added an iSoft RIS. Interoperability became a particular issue, with no communication between the two systems. There remained a risk that written requests, or images and reports could be mislaid or misdirected and that confusion between different patients could occur as a result of this communication gap between PACS and RIS.
With increasing pressure to meet targets within the NHS for discharge summaries and turnaround times, which included radiology reporting, GOSH had to address the interoperability issues in a way that reduced the risk of error and increased efficiency.
Three measures were identified:

  • Bring in speech recognition to reduce turn around times
  • Create a work list for easier organisation and team communication
  • Establish a RIS/PACS interface to ensure consistency between report patient and image reporting preferences.

Integration of the desktop has brought image, report and scanned paper requests together within RIS and PACS. In terms of productivity, it is the speech recognition element that has proved a significant step forward.
Prior to deployment of Soliton IT’s Workflow Manager 40% of reports involving plain film radiography were available within 24 hours. Using Soliton with Speech recognition pushed that up to 65% very quickly and within two months of implementation 85% of reports were available within that timescale.
We found Soliton Workflow Manager with SpeechMagic simple to operate. The radiologist speaks into the system, which transposes it into text, and then can check instantly whether, or not, the speech recognition equipment has reflected the report accurately. If not, amendments can be made quickly and easily via the microphone. In complex cases, the corrections can be typed in but in more than half of cases the reports on the Soliton/SpeechMagic system are accurate first time. Once the reporter is satisfied, the document is authorised as correct and despatched. From the point at which the radiologist views the image, to the report being authorised and despatched, is now only a matter of minutes.
The system has been highly effective in linking our iSoft RIS into the Siemens PACS.
When an investigation is booked into the RIS, it is assigned a unique accession number which is carried through into the modality. After the examination the images are sent to PACS with the accession number, enabling clinicians across the trust to view the images and also see the corresponding report. This security element with the accession number linking reports and images, eradicates the risk of associating information with the wrong patient.
The clear labour and resource saving attributes have made speech recognition an attractive investment from a cost effectiveness perspective for the Great Ormond Street Hospital for Children NHS Trust. There is the argument that this technology renders secretarial support redundant and leaves highly-trained and well-paid radiologists doing administrative roles.
In effect, Soliton/SpeechMagic is simply cutting out the middle element of the human secretarial support, and offering the radiologist greater control, more instant access and closer responsibility for the final document while speeding up the availability of the report.
The technology transcribes dictations very accurately eliminating the back and forth liaison between radiologists and secretary and leaving less idle time between the dictation, transcription, validation and final authorisation of a report.
In essence, the amount of time the radiologist spends on administration falls and the system becomes less vulnerable to gaps in availability of secretarial support. Transcription work can be mundane, resulting in a high turnover of staff in these roles. Those who had become experienced and got to know the language, terms, vocabulary and nuances of a radiologist were rare and valued. In a health Trust where capable and committed staff are appreciated, there are now opportunities for the more efficient, effective and able transcription secretarial staff members to de deployed in more challenging and rewarding positions, offering additional responsibility and remuneration.
At a time of looming financial constraints within the UK’s NHS, speech recognition has proved a successful investment for us as a Trust enabling us to meet greater demand for reporting and discharge targets.
Speech recognition can pay for itself in just over a year – when compared with the salary outlay for transcriptionist staff the financial benefits are evident.
The GOSH system now also delivers more accurate reports. The operation is streamlined, radiologists do less reporting administration work and the increased accuracy will significantly improve patient safety.

[[HHER22]]