DIGITAL DICTATION, DOCUMENTS & DATA
Rolling out speech recognition at Poole
John Stubbs, IT project manager at Poole Hospital NHS Foundation Trust and The Royal Bournemouth & Christchurch Hospitals NHS Foundation Trust, discusses the implementation of new speech recognition software.
Foundation Trust just over two years ago, it was decided that rather than cutting jobs, the trust would look at embracing modern technologies to save time, money and resources.
F
In particular, the trust recognised three areas for development: electronic document management, which is going live at the moment; the adoption of a patient pathway coordinator model, to realign the secretarial areas; and the introduction of digital dictation/ speech recognition.
Prior to going down the digital dictation/ speech recognition route, tapes were mostly used by the trust’s consultants and secretaries and there was some outsourcing in a couple of departments.
John Stubbs, IT project manager at Poole Hospital NHS Foundation Trust, told NHE that digital dictation was piloted last September in the department of medicine for the elderly (DME) and orthopaedics, where it was well received.
Then, at the end of 2013, the Child Development Centre (CDC) was brought on board too, but using speech recognition.
Stubbs said: “At that point we were trying to ascertain the best way to implement services. For instance, should we go straight to speech recognition? Or do a two-step with digital dictation and then speech recognition?”
Going straight to speech recognition was seen as the best way forward. In fact, the paediatrics, DME and orthopaedic departments have recently started using speech recognition as well.
First 10 hours ‘critical’
Poole decided to use G2 Speech UK Ltd for its speech recognition software, because of the company’s ability to implement its services trust-wide. But the company is clear that to achieve its stated 98% accuracy, there needs to be about 10 hours of dictation recorded per author.
Stubbs noted that CDC was achieving 98% accuracy by Christmas 2013, and now when his team check the statistics for the secretaries
ollowing a series of admin and clerical staff workshops at Poole Hospital NHS
in this department, “for every minute of dictation it takes them about one- and-a-half minutes to convert that into a letter,” he said.
The departments that went from digital dictation to speech recognition have struggled to achieve this level of accuracy – although week by week, improvements are being made.
Increasing accuracy
James Stapleton, head of operations at G2 Speech UK Ltd, told NHE that the company’s technology is now being used in at least 35 trusts in at least one department, and trust-wide in 14 trusts, with more currently rolling it out.
With regards to the first 10 hours of dictation, he said: “People start off at around 70-80% accuracy. After two to four hours it can go up to 90%, and then after eight to 10 hours they can get up to 98% accuracy – but people vary.”
G2 is moving to a new, upgraded recognition engine. “We’re finding that the results are much quicker – but our guidelines are still eight to 10 hours for 98% accuracy, as we don’t want to mislead people,” said Stapleton.
“That first 10 hours of dictation is critical, and should not be underestimated,” said Stubbs, noting that at times the system can be running perfectly from the off, but in other instances it can take longer to perfect. And, unfortunately, there can be backlogs, which have been seen in DME and orthopaedics.
“But we produce some charts each week, and we can see the volume of letters going up and up, and the time taking to convert that one minute of dictation into a letter is going down,” he said.
Currently the trust has 120 authors and 37 transcriptionists using the G2 technology, with the emergency department, rheumatology, palliative care and dermatology recently coming on board. However, by mid-2015, when the system goes trust-wide, Poole aims to have 400 authors and 125 transcriptionists trained up using speech recognition.
How it works
NHE was told that the trust’s secretaries, even in DME and orthopaedics, have started seeing the positive benefits the system can bring in time and cost savings. In fact, just by introducing the technology there has been a step-change in how secretaries produce the letters.
“The author does their dictation, then it goes through to the secretary and they do the corrections and fire it back to the consultants, who approve that it is an accurate representation of what they have said,” stated
John Stubbs James Stapleton
TELL US WHAT YOU THINK
opinion@nationalhealthexecutive.com
national health executive Sep/Oct 14 | 77
Stubbs. “But what is a big change is that consultants aren’t seeing the final letter – they are signing off that it is what they said. It then goes back to the secretary, gets put in Word and is then posted into our electronic patient record (EPR) and copies are printed.”
This has fitted in extremely well with Poole’s decision to implement admin and clerical secretarial review (along the lines of the patient pathway coordinator model), which has restructured the secretarial PAs. The main idea has been to free up the Band 4 and 3 secretaries, helping them look after the consultants’ caseloads more effectively and not worrying so much about the typing, with Band 2s doing all the topping and tailing of the letters.
Stubbs welcomed the support from G2 during the implementation, and said he expects to see speech recognition used much more widely across the NHS.
“By freeing up time and resources from the back office,” he said, “it means there is more money to invest in frontline staff.”
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89 |
Page 90 |
Page 91 |
Page 92 |
Page 93 |
Page 94 |
Page 95 |
Page 96 |
Page 97 |
Page 98 |
Page 99 |
Page 100 |
Page 101 |
Page 102 |
Page 103 |
Page 104