Chatbot Docs

Image copyright Getty Images

Could health apps and chatbots eventually replace your traditional doctor?

"Let's talk about how you've been feeling over the past 30 days," says Joy. "This will help me get a sense for your current state."

Joy probes a bit deeper, asking a series of questions: Do I feel hopeless? Do I feel restless? When I respond that I'm a bit stressed, Joy offers me several de-stressing techniques.

Joy might appear to be my counsellor or my life coach, but the conversation I'm having is actually with a chatbot that uses artificial intelligence and machine learning to track emotions and provide mental health support - all through Facebook Messenger.

Welcome to healthcare in the digital age, where smartphone owners have access to their very own doctors and therapists at the touch of a button.

Joy was founded by Danny Freed after one of his best friends committed suicide.

"This planted the seed for me that there must be a way technology can help anyone who is struggling with their mental health," he says.

"I had this realisation that we are tracking all sorts of data and metrics when it comes to our physical health, but next to nothing for our mental health."

He says Joy encourages people to open up about their mental health and feelings, and in return hands out relevant research-backed tips, techniques and exercises.

But how do we know if it works? Mr Freed is certainly careful not to make too many grand claims for the app.

"Joy is not a clinical or diagnosis tool," he says. "You can think of Joy more as a friend or a coach."

Perhaps aware that credibility may be an issue, Mr Freed has recently recruited a PhD student specialising in counselling psychology who is "an expert in mood disorders with extensive clinical training in providing evidence-based therapy with adolescents and young adults".

Since Facebook opened up its Messenger platform to developers in 2016, more than 100,000 bots have been built on the platform, many focused on health and mental wellbeing.

Woebot, for example, helps users track their moods and potentially spot and fend off the early stages of depression.

Created by Alison Darcy, a clinical psychologist at Stanford University, Woebot employs cognitive behavioural techniques.

As it learns more about you, it can see patterns emerging and suggest ways to alleviate your bad moods or negative thinking.

While Joy is free, Woebot costs $39 (£30) a month after 14 free sessions. Given that counselling or psychotherapy sessions with a human can typically cost £30-£200 an hour, it is easy to see the appeal of such digital helpers.

But many potential users may also have concerns about privacy, given Facebook's seemingly insatiable thirst for our data.

Dr Ali Parsa, founder and chief executive of digital healthcare app Babylon, sees the m-health trend as an undoubted force for good.

"It's time to do with healthcare what Google did with information - using the power of technology to democratise access for all, and put a personal doctor in everyone's pocket regardless of geography or income," he says.

Babylon, whose scientists, engineers and clinicians have built a knowledge bank of more than 300 million medical facts, provides users with information and advice on their symptoms based on machine learning and natural language processing.

"It can understand, reason, diagnose, make prognoses, and learn from practice, just as human doctor would," says Dr Parsa.

Babylon also allows users to book an appointment with a GP or therapist, costing from £25.

Medical chatbot Your.MD, which received $10m (£7.6m) in funding in June, works in a similar way to Babylon.

It offers a free pre-primary care information service based on a user's medical history; symptoms; and personal factors, such as age and gender.

Your.MD also recommends relevant local health services and products that can help you with your ailment.

"Your.MD is arguably better than doctors," claims chief executive Matteo Berlucchi. "It has no bias, no preconceptions, and a superior mathematical brain."

Dr Parsa is similarly bullish.

"Babylon scientists predict we will shortly diagnose and prevent personal health issues better than doctors in most cases," he says.

More Technology of Business

• 'Fake news': What's the best way to tame the beast?
• Meet the Guides encouraging girls into science and tech
• Could new data laws end up bankrupting your company?
• The 'Babypod' carrier that comes with an F1 pedigree

But both also acknowledge that human doctors have qualities medical chatbots cannot replicate.

"They have empathy, they can look at you, assess you in the flesh, listen to your breathing, look into your ears, take a blood test, and so on," says Mr Berlucchi.

There are plenty of sceptics out there, particularly given than many of these apps do not have to be regulated at all.

Karen Taylor, director of the UK centre for health solutions at consultancy firm Deloitte, says: "Patients shouldn't solely rely on them as a full replacement for the medical system. There are many areas that apps aren't yet equipped to handle or support."

And Richard Vautrey, acting chair of the British Medical Association's GP [general practitioner] committee, says: "Whilst [AI and machine-learning apps] identify a bit of information and bring it together in a way to provide a diagnosis, I do think we need the skills of a trained doctor or healthcare professional to get beyond present symptoms."

But surely such apps will improve as they absorb and learn from the mountains of data they receive?

Dr Claire Novorol, co-founder and chief medical officer of Ada, a health tech start-up, predicts that as its app integrates more data from sources such as lab test results, genetic tests and wearables, it will be able to identify and track health patterns and flag up possible issues in advance.

"Ada will become much more of an ongoing health companion, supporting long-term monitoring of health data to enable predictive and proactive care," she says.

AI-supported doctor-patient relationships will be "more collaborative", she believes.

• Follow Technology of Business editor Matthew Wall on Twitter and Facebook

• Click here for more Technology of Business features