Using an AI-driven dictation platform to free up clinicians’ time
King’s College Hospital serves 700,000 local patients and acts as the end of referral pathways for millions in the south of England and London. This creates a huge amount of paperwork that would usually be typed out by medical secretaries. Due to the pandemic, the speed at which this paperwork is created has slowed dramatically.
Situation
Most users relied on traditional transcription services with digital audio files sent to human typists who would then send text files back to secretaries to enter onto correctly formatted letters with patient details. Dictation files could, albeit infrequently, be lost in upload or transfer. Overall this process could take days.
If used at all, existing dictation technology was substandard, with poor accuracy rates. The reason for this is simple - the platform medical staff were using had not learnt from their individual voice and medical practice. Just like the first time a person hears someone speak, a regular dictation system is in effect hearing you for the first time.
Aspiration
The team wanted to implement an easy to use, full spec transcription system that was accurate enough to be used confidently but also used a range of other well defined features that helps to reduce the administrative burden of writing clinic letters.
Solution and impact
The gastroenterology team at King’s have been using Hepian’s AI-driven platform to decrease the turnaround time for all of their medical letters.
By providing a more accurate transcription solution and a user friendly interface the team has been able to refocus their valuable time on patients.
Functionality
Hepian Write allows doctors to dictate letters quickly and accurately, with the text transcribed in real-time into preset templates, by an algorithm that learns from each user’s voice, speciality and unique speech patterns. The system is accessible with a mobile app and a cloud interface.
Capabilities
- Transcribe dictated text accurately and in real time
- Learn from each individual user’s voice, speciality and speech patterns
- Dictate letters using your phone, and then edit them in an intuitive web interface
- The functionality is quick to set up - users can dictate their first letter within minutes of signing up
- Permits the user to create highly customizable templates to remove duplicate work and completely automate letter formatting
- Send letters by secure email to patients and their GPs or print them for posting
Scope
Hepian can be used in clinics or at home.
Key learning points
Users should speak slowly and clearly for the first few letters as the system gets to know your voice. Users should try to speak in their natural tone - Hepian’s AI listens like a person, so there is no need to talk to it like a computer.
Make any changes you want to your templates so that they fit your requirements.
Key figures/quotes
“Hepian Write learns my voice as I use the tool, so after one clinic it knew my specialty, subspecialty and accent - that's why I think it's more accurate.”
Jonathan Makanjuola, consultant urologist, King's College Hospital NHS Trust
“With Hepian I can quickly turn round medical letters as I am running the clinic - without burdening myself with error correcting.”
Rajesh Nair, consultant urological surgeon, GSTT
Find out more
How Hepian connects AI and physicians to improve healthcare
Key contact
Bu Hayee, consultant gastroenterologist, King's College Hospital NHS Trust
Oliver Bee, founder, Hepian
Disclaimer
These case studies summarise user and patient experiences with digital solutions along the relevant care pathway. Unless expressly stated otherwise, the apps and digital tools referenced are not supplied, distributed or endorsed by NHS England or the Department of Health and Social Care and such parties are not liable for any injury, loss or damage arising from their use.
All playbook case studies have either passed, or are currently undergoing the Digital Technology Assessment Criteria (DTAC) assessment.
Please note the full legal disclaimer: NHS England playbook disclaimer