The Nuffield Trust has published a new report called Delivering the benefits of digital health care. It is a well-written, timely piece of research which sets out how we can transform healthcare systems using digital technology. It succinctly identifies the conundrum we currently face: that although healthcare delivery is being transformed by new technology, strategic decisions about clinical transformation and the investment needed in information and digital technology are often relegated to the end of NHS board discussions.
Culture change is critical for deploying digital technology in healthcare.
The report includes seven lessons for success. The one that jumped out at me is that culture change is critical. This is something we are very familiar with at Nuance because we work hard with clinical staff at the frontline to ensure speech recognition software becomes embedded in their everyday tasks.
The role of speech recognition in healthcare
Speech recognition is mentioned briefly in the report in the description of Nuffield’s own vision of how healthcare is likely to change in the next 10 years. It suggests that ‘…though there is interest in new models of care, the most significant improvements in productivity over the next few years are likely to come from the combined impact of large numbers of small changes and extracting the full benefit from the technologies currently available…A lot less time will be spent by staff on administrative tasks and routine communication, as automation, voice recognition and natural language processing become more commonplace.’
Speech recognition is a mature technology that has now taken the consumer market by storm. If you don’t use speech recognition software to command your phone, your car, even your TV, you probably know someone who does. Within healthcare, speech recognition will help drive significant improvements in productivity from the huge investments already made, but not yet fully realised, in the EPRs as well as improving the day to day lives of the staff by freeing them from the keyboard.
Speech recognition changes how healthcare professionals work.
In healthcare speech recognition software understands clinical language. This is quite different from the software you might get on an iPad, or iPhone. Radiologists have been our Trojan Horse with hospitals throughout the land using speech recognition as their standard method of reporting and managing patient and medical records. Take it away and they would more than likely bite your arm off.
Over the years, healthcare speech recognition software has improved in leaps and bounds, whether it be around the medical vocabulary, an individual’s phonetics or the mathematical probabilities of the language model. So too has the hardware upon which it runs.
Sadly there has been a lack of published research in this area. However, that does seem to be changing with recent studies providing good evidence. A German university hospital undertook a study that showed how a web-based medical speech recognition system for clinical documentation increased documentation speed by 26 per cent, increased the amount of content by 82 per cent, and also resulted in enhancing user satisfaction. In the UK, mental health has also taken to the stage:
South West London and St. George’s Mental Health National Health Service (NHS) Trust showed that time spent entering patient notes was reduced by almost 50 per cent when using speech recognition.
Surrey Borders Partnership NHS Foundation Trust observed the average turnaround for letters or reports dropped from 6-7 days to just 1-2 days.
The EPR (Electronic Patient Records) also adds a new dimension to the challenge for healthcare professionals. The Nuffield Trust report highlights that the EPR ‘straddles the system as a whole, reflecting the pivotal role it plays in any digital strategy’. It acts as a ‘foundation on which many of the other tools are built’.
However, clinical studies have shown that EPRs can add as much as 90 minutes to a medical professional’s day.
Clinical speech recognition – here to stay
Without good quality information in an EPR, the seven opportunities to drive improvements outlined in the Nuffield Trust’s report will not be realised. So this conundrum must be solved with smart technology.
Huge strides have been made in the quality, accuracy, performance, affordability and time-to-value of speech recognition solutions and their use in healthcare is on the rise. The benefits of improved clinical documentation using speech recognition are broad and varied – in essence better quality of care, improved patient safety and health professional satisfaction. The result is the ultimate goal – health professionals have more time to spend with their patients like we see it for example with the nursing team at at Alder Hey Children’s Hospital PICU. They are saving 40 minutes per patient per day using speech recognition versus typing, with an improved hand over and faster referral and discharge process to other units, and most importantly freeing up more time to care – as also published in the Nursing Times last year.
We are keen to demonstrate how speech recognition can help clinicians save time and record a greater amount of detail more accurately, so if you have any personal experiences you would like to share, please let me know. I look forward to getting your feedback.
Dr Simon Wallace, Medical Consultant, Nuance Communications UK