Waking up with a sore throat has never been so scary. Am I right? Before, it was a very normal part of your monthly ‘just feeling a bit run down’ kind of day. Now instead – instant panic floods your body for the fear you’ve contracted COVID-19. Well – that was my experience a couple of weeks back.
Anyways, just like most people (89% actually), before thinking about whether or not I should see a doctor, I Googled my symptoms. And I guess you can’t blame people for that – not knowing what’s wrong with you the moment you feel unwell can be nerve-wracking, especially of late.
As I navigated through several sites – I was shortly confronted by Molly, who popped up seemingly out of nowhere, introduced herself to me and asked me to tell her what my symptoms were.
My first experience of using a chatbot in healthcare.
I was quite surprised – of course I’ve heard all about chatbots and the buzz they’re creating in healthcare – but to be confronted by one so quickly and seamlessly without really knowing it was happening was both startling and impressive in equal measures.
It makes sense though – innovative and nimble tech start-ups identifying a huge problem in human’s behaviour patterns (aka Dr. Google) and using AI to solve this seems like an obvious opportunity. To be more specific, chatbots (for those unaware) are very intelligent algorithm-powered, text or voice-based interfaces that use natural language processing (NLP). In simple terms, a computer that can simulate conversation with humans.
The technology behind it is actually mind-blowing – I won’t go into the ins and outs of that now (there is a very handy article here for those interested), but it’s funny to think that people (aka technophobes like me) don’t realise quite how much chatbots have already infiltrated everyday life.
Virtual assistants like Google Now, Microsoft’s Cortana and of course, Siri, are fine examples of chatbots coming into our lives (and homes), as well as those found in more corporate settings, like in call centers, real estate, banking, and to be honest any business really involving customer-service. Thanks to their intelligence, responsiveness, and usefulness – it seems chatbots are here to stay. And like all fast-moving digital trends – following slightly behind everyone else’s footsteps – is healthcare.
Now the very purpose of a chatbot is to make customer interactions easier – I can’t think of another industry where quick and easy time-saving tricks could improve practices more, than in healthcare. In today’s busy lives, visiting the doctor or hospital may not always be feasible, and even going to the doctor can be a frustrating experience – all the time-consuming miscellaneous admin, forms and paperwork. And the waiting. All of the waiting.
Imagine the autonomy you’d have as a patient by being able to grab your phone, open an app, and quickly interact with an AI-driven health companion to receive a smart and accurate diagnosis. Exciting, right? Everyone wants convenience – including patients – and chatbots seem to have the answer.
And this isn’t just blue-sky thinking. According to a report, over 40% of senior healthcare executives consider AI technology to have the greatest impact on their industry within the next three years. And with venture capitalists having reportedly invested over $800 million into chatbot start-ups with healthcare capabilities – this market is set to skyrocket in the next few years.
As health services move towards being more patient friendly – offering a customized and sufficient experience is of utmost priority. Chatbots are gradually being adopted into healthcare and are generally in the early phases of implementation – the majority appear to focus on checking patient symptoms.
But beyond simple symptom-checking there are countless cases where a chatbot could help. Better organization of patient pathways, medication management, FAQs, help in emergency situations or with first aid, scheduling appointments, health tracking: these are all possible situations for chatbots to step in and ease the burden on medical professionals – which then crucially frees up their precious time to concentrate on providing first-rate care.
This is particularly relevant in today’s situation, where the spread of COVID-19 puts an unprecedented (yes I know I said it) burden on healthcare providers.
And it’s not even just about easing burden – data from Providence Health’s deployment of chatbot Grace suggest 90% accuracy in its recommendations, with patients clicking on the symptom checker’s recommendation 35% of the time. Many award-winning health campaigns have even centred around implementation of chatbots – Burson Cohn & Wolf’s (BCW) Global AI chatbot ‘Tabatha’ in partnership with Boehringer Ingelheim in asthma landed them a handful of trophies – the first ever launched by a pharmaceutical company.
However, naturally, just like any other technology, chatbots come with their shortcomings and disadvantages. One of the key concerns, understandably, is user privacy, and sharing information about your medical condition with a chatbot – which seems less reliable than sharing the same information with a human doctor.
Additionally, and perhaps most importantly, delivering healthcare is different from selling credit cards because of the potential for harm arising from incorrect or confusing guidance. Diagnostic errors in medicine are not only common, they can also have steep cost implications – a study found paid medical malpractice payments over a 10 year period totalled $5.7 billion in payments.
Finally, chatbots could potentially cannibalize revenue for healthcare professionals in countries with fee-for-service (FFS) reimbursement models. And it seems the physician’s jury is still out – at least in the US anyway (perhaps due to my previous point) – a study showed mixed views concerning chatbots. 70-80% agreed they would be useful for administrative benefits, and the same proportion also agreed they could not effectively care for all patient needs or provide detailed diagnosis and treatment.
Let’s be clear: chatbots will never replace doctors – they provide many opportunities to facilitate their jobs or improve performance, but, ultimately, it’s human doctors who are going to deliver the care.
We know that these AI-driven robots will thrive in a world that’s continually short on time. There’s no end to what chatbots may provide us – but more research is clearly needed to prove the value and impact of chatbot use on health outcomes. Whilst we can’t predict where we’ll be in the next 10 years, it’s certainly an exciting time in the healthcare space.