The Future of Healthcare | Gareth Thomas, CTO at Healthily
“How well a system is designed must always be distinguished from how well it is funded; the NHS is quite well designed but under-provisioned. By contrast, the US healthcare system is richly funded but designed so that it maximizes waste, inefficiency, and inequity.”
The above quote is from a paper published back in 2003 by the US National Library of Medicine. Even though it’s almost twenty years old, the words are as true now as they were then.
But the purpose of this article is not to get into the rather heated argument around the ideological merits of different healthcare systems, it is to highlight the reality of where we are now with aging populations; unsustainable increases in preventable chronic conditions; and the availability of scalable medical capacity to cope with this growth.
I have had the good fortune to have lived in the UK and US and experienced both healthcare systems, and in both cases I have worked for companies building software products in this space. It is a highly- regulated industry and patient safety is of paramount importance, which means any product or service you build takes years and millions of dollars to get to a point that it is suitable for wide public use.
From a technology perspective, though, a lot has changed in twenty years. We can now build and deploy software in cloud services in a way and at a pace that would’ve been hard to imagine back then. Cloud services have enabled us to abstract away so many elements of complex software systems, which have reached their ultimate manifestation in Serverless, No-Code and Low-code products. We can also now leverage the power of AI and more specifically machine learning to improve the capability of the tools that we produce, enabling us to access enormous amounts of data in the cloud with which to provide feedback to and improve these products.
Which brings me to the focus of this article: telemedicine. While a recent study has shown that growth in this area has slowed in the United States post pandemic, there is no doubt that the future of medicine will be heavily reliant on this technology. It will be needed to provide healthcare in developing countries that are struggling to cope with huge population growth and primitive medical systems. And if you take Elon Musk at his word, we will need it for the first people to leave earth and colonize another planet!
The technology for video calls is now widely used and something we are all familiar with. But although speaking with a physician over the internet provides obvious benefits in terms of convenience and reducing the spread of infection there are still a wide number of conditions that may be better served by seeing a patient in person. What an online consultation can do, however, is enable the use of a very key piece of new technology to aid the physician in making the correct diagnosis: a symptom checker more commonly known as a medical chatbot.
At the moment, there are only a handful of companies in the world whose symptom-checker technology is available and mature enough for wide-spread use. At Healthily, we have spent six years and tens of millions of dollars to build a complex graph of data we call Medibase. At time of writing, it covers almost seven hundred conditions, expressing the relationship between them and their symptoms in order to calculate the probability of a patient having a condition given their own symptoms, age, gender, influencing factors (health background) and season. We then provide a triage recommendation, which incorporates a whole host of other rulesets and “red flags” to ensure the service is as safe as it can be. This triage is the outcome of a conversation with our chatbot (trained on a large dataset of medical questions), based on a Bayesian probability model calculated as the user responds to a series of questions. It’s a completely explainable model and not an AI black box.
The outcome provides what they call in Special Forces, a “force multiplier”, which can preface the appointment with the physician, leveraging the power of algorithms and data in the cloud to augment the physician’s training. This is the future in which we can enable healthcare systems to do more, with less.
Originally published at https://www.siliconroundabout.org.uk.