Online out of hours triage or 20 questions

Talk Code: 
Adrian Brown
Author institutions: 
St George's University of London


There is an increased demand for out of hours services and pressure on primary care and A+E. Triaging out of hours contacts is a significant challenge for the health service. The nurse led NHS Direct was seen as too expensive and there are concerns about the effectiveness of NHS 111. The NHS is considering innovative solutions to manage cost and demand and is piloting an app in areas of North London starting in Feb 2017. This uses software from Babylon which is also a free online app. The free app offers a text based triage with options for self care, book an telephone or GP video consultation. This study considered the effectiveness of triaging presentations, that could be managed by self care, in the online Babylon app.


10 scenarios of minor illness, presenting acutely, considered not in need of an out of hours consultation were inputted into the app. For example “I have a cough” was the start of a consultation for 25yr old man with a productive cough for 24hrs with no temperature no sob and no past medical history.


5 of the 10 presentations, diarrhoea, headache , knee pain, sore throat and back pain resulted in self care advice, or advice to see a pharmacist, The other scenarios cough, rash, bilateral conjunctivitis, earache, and dyspepsia ended with advice either to message or consult with a GP. For one presentation, “I have a rash” no questions were asked before the recommendation to seek advice, although there was the option to send a photo of the rash. For the others the range of number of questions was 12-15.


Any presentation dealt with entirely automatically is a cost saving over the current 111 system, and future healthcare interactions are likely to become more automated. However, there remains a risk of creating demand on services if the advice is too cautious. The online version tested offers a GP video consultation for a fee, but there may be cost implications for the NHS if common self-limiting illnesses are not able to be screened out effectively without losing the sensitivity to exclude serious illness. It would be interesting to see the algorithms that resulted in the advice given but they are part of a proprietary system. Evidence based medicine with links to evidence and guidelines supports routine clinical practice Is there a role for developing open source diagnostic algorithms in the future of the NHS.

Submitted by: 
Adrian Brown
Funding acknowledgement: