The E-Health Insider website Two years ago Northumbria Healthcare NHS Foundation Trust set itself the goal of ensuring that every GP and patient receives correspondence relating to a clinical episode within 48 hours. With correspondence backlogs of up to 13 weeks, the trust faced a stiff challenge. Digital dictation and speech recognition were seen as the foundation for improving document management – but, as far as anyone knew, no trust in the UK had tried to use the technology on the scale the trust was proposing. It was a bold move. A contract was put in place to implement G2 Speech’s MediSpeech with Philips SpeechMagic and work started in April 2006. Project manager overseeing the introduction of the new system, Jen Henderson, says: “The trust went for a speech recognition system for one reason and that was to make the turnaround time for all documentation less than 48 hours, which was a large undertaking when we had clinics where the turnaround could be 12-13 weeks.” She cites the example of a pain management clinic with four consultants, one specialist nurse, a psychologist, four secretaries – and a 53 clinic backlog. “We put an audiotypist in for two weeks while the full-time secretaries worked on G2 Speech. They cleared the backlog and it has not returned.” Focus on batch speech recognition Efforts have been focused on introducing batch speech recognition in which the dictated sound file is transcribed digitally, but a secretary corrects the document produced by speech recognition before producing the completed document. It is a radical change of role for the secretary turning her from a typist to a proof reader and editor of letters, clinic notes and other documents. Denise Patterson, (pictured) a senior secretary at Wansbeck Hospital, Ashington, was the first adopter. “It thought it was going to be a better way of managing the workload,” she explains. She feels that one of the best aspects of the changeover have been the move from the old system in which tapes containing unknown quantities of work piled up and were shared between secretaries for typing when one person’s backlog became unmanageable. “You can actually see what’s in the inbox,” she explains, showing Monday’s list of pending files on her screen following a busy weekend. “You can see who’s got what and where the problems are.” Making the workload more visible This transparency brings benefits by making the workload much more visible and highlighting urgent work clearly. Work can be shared quickly and colleagues are happier to take on a known number of files displayed on the computer screen to process, rather than a set of tapes that may contain dozens of notes. “You could spend half the day organising the work. Now it’s just there,” she says pointing to her screen. Such clear visibility was uncomfortable initially for some who found the system a little Big Brother-ish. A recent straw poll among secretaries, however, revealed that only two out of 13 preferred the old system and their preference was due to a fondness for typing rather than any active dislike of G2 Speech’s system. To their surprise, the secretaries found that the system transcribed the medical terminology well, but needed time to ‘learn’ the voice of the user and pick up the everyday language used. Correcting the text and feeding the corrections back into the system is vital in the process of fine tuning the software to recognise the characteristics of a user’s voice. ‘Little and often’ rule for switching Henderson advises new users to use the new system ‘little and often’ at first so that the secretary can take time over the files created through speech recognition and process others quickly under the routine to which she is accustomed. Gradually, as the speech recognition files are processed faster, a full switch can be made. Teamwork and partnership between the consultants and secretaries is essential, she says. “A lot of it was down to Denise. She was a real driving force,” says Henderson. Clinical staff have found benefits once they become used to the system. Consultant orthopaedic surgeon, Simon Jones, says: “You can get urgent files typed up and brought back quickly. Tapes don’t go missing, unravel or get stuck.” Just over a year on, how is the trust doing with the system? Out of a potential 325 ‘dictators’, 87 have converted - including all the orthopaedic surgeons at Wansbeck - and 90 transcriptionists have been trained to use the system. Another achievement is a web-based training facility on the trust’s intranet developed to overcome difficulties in getting busy staff together for training sessions. Jen Henderson praises G2 Speech’s contribution saying: “Everything we’ve asked of them, they’ve done.” This included a lot of support for early users and adaptations such as a template to record documents from users working away from their base. This enables a consultant based at Wansbeck Hospital to take a clinic at Alnwick Hospital, for example, dictate the notes at Alnwick, dock the dictation machine there and upload the files to a server in North Tyneside ready for transcription back at base in Wansbeck. A totally different way to produce text The trust’s executive director for elective care and surgery, Ann Wright, who started life in the health service as a medical secretary, saw the possibilities opened up by speech recognition in an area that had changed little apart from shifting typing from typewriters to PCs. “I saw it as a totally different way to produce correspondence and text,” she says. Several Northumbria staff, including Wright, spoke of the new system as the trust’s foundation for electronic records. Certainly the trust's work is attracting a steady stream of visitors from around the country to see what has been achieved. Asked when the task might be finished Wright says this will happen when the system is “in everywhere, into every specialty, into every consultant’s office and when it’s in and it’s working well and we’ve re-engineered the whole secretarial and administrative service.” Links Northumbria Healthcare NHS Foundation Trust
http://www.northumbria.nhs.uk/ G2 Speech
http://www.g2speech.com/en/ Philips SpeechMagic
http://www.philips.com/speechrecognition
|
