This site is intended for health professionals only

My vision for algorithm-based GP triage

Dr Ed Turnham argues that technological advances are not a threat to the profession, but allow GPs to provide better access, and even better continuity of care

Last month, the RCGP said it was ‘not convinced the ‘total triage’ model should be the default for all practices once social distancing measures are lifted.

But all GP practices have done triage for years. It is simply the most appropriate person dealing with a problem in the most appropriate way. Dressings are triaged to a nurse, bloods to a healthcare assistant. Receptionists have become ‘care navigators’ and weed out non-urgent issues from the duty doctor’s list. So the question becomes this: could triage be done better by a combination of intelligent software and clinicians, rather than by non-clinicians?

I’ve spent much of the past two years promoting the ‘total online triage’ model and supporting practices to implement it. In my vision, a patient’s encounter with the NHS starts with triage by an algorithm – a bit like 111 but with access to the full medical record. Most will do this online, while others will be led through the questions by a receptionist. Self-limiting problems are managed solely by the algorithm or signposted to pharmacies. Other problems are triaged to face-to-face or remote-first assessment by an appropriate clinician. Patients with, say, dementia will be asked simpler questions.

This may sound soulless and impersonal. But we already encourage self-management of minor problems and long-term conditions. Good algorithms will extend the range of conditions patients can manage without clinical support. Key questions in the history are less likely to be missed. When patients require our input, these thorough histories will feed into clinical decision support tools that complement our human skills and compensate for our cognitive biases, leading to better care.

‘Digital first’ may evoke visions of video-based care with no continuity. But practices like mine have used online triage to increase continuity. Our relationships with patients are now built on a mix of face-to-face, phone and textual contacts – much like those with family and friends. Triage algorithms could further improve continuity, detecting which of the available clinicians knows the patient best. Continuity will be prioritised for those who need it most, while ensuring more ‘popular’ GPs aren’t overburdened.

Data from triage systems such as AskMyGP disprove the idea that all patients want face-to-face appointments. Only about 25% of patients asked for these, even pre-Covid, and satisfaction at online triage practices is high.1

Algorithms could take into account a patient’s preference for face-to-face or remote management. Yes, more work is needed to ensure some groups won’t lose out from new ways of working. But remember that the old system – with its 8am telephone scramble, three-week waits and unnecessary half-days off work – disadvantaged the vulnerable too.

Less face-to-face contact might mean we miss some important ‘and another thing’ door-handle moments. But even in the traditional system many – perhaps most – GP interactions with patients are remote: signing repeat scripts, filing bloods, reauthorising levothyroxine.

Many other interactions aren’t with a GP at all: blood tests, dressings and incoming phone calls. If it’s so important for GPs to see patients in person that the idea of dealing with a UTI or sore throat remotely can be seen as a threat to our profession, why do we already do so much remotely or delegate it to others? By dealing with simple matters quickly,

I provide better access and am often able to give 30 to 45 minutes to complex patients. Could this be a better way to coax out those hidden agendas?

Some GPs feel less face-to-face means less satisfaction – ‘this isn’t what I signed up for’. I never hear such concerns from hospital doctors, who often spend most of their time away from (conscious) patients – in MDT meetings, operating, reviewing scans or helping junior colleagues. Is this perhaps because they aren’t professionally isolated in their clinics, with only patients for company? Triage allows us to work in communal offices, enjoying the company of colleagues and sharing knowledge.

In every other industry and service, it’s uncontroversial to use technology to improve services and optimise the time of highly paid professionals.

It’s reasonable to argue that technology in general practice isn’t ready yet, but can there be any doubt that this way of working should be the future?

Dr Ed Turnham is a GP in Norfolk and chief clinical information officer at Norfolk and Waveney CCG


Hello My name is 31 May, 2021 2:38 pm

I’m not convinced, sorry. We need more staff and demand control.

Patrufini Duffy 1 June, 2021 3:07 pm

One must differentiate between a member of public and a patient. **There is a clear abyss between the two. Treating trivia of individuals (who are categorically not unwell or classified as a patient) with online tools because their boyfriend, mum or best friend couldn’t give them common sense advice may make some doctors feel self-fulfilled, but is a complete waste of time. You are treating nothing – just a mind with a personality disorder. The problem is a global pandemic of health anxiety, funding corporations and individuals with tech degrees and an MBA and lack of any human self-awareness. No computer bot can help that – but can only augment it, for profit purposes. Put these individuals in the dessert, and their issues will disappear when their lust for water is the only thing that matters.

Mark Semmens 5 June, 2021 12:26 pm

Yet to see an on-line triage tool that doesn’t prompt patients to say how urgent they think their issue is or when they would like to be seen. There in lies your problem. Promoting more anxiety of the worried well is one thing. But given patients ammunition for urgent care is another. Let’s educate not facilitate

Not for me !!