The Evolution of Digital Health and the Era of Cognitive Computing
The Evolution of Digital Health and the Era of Cognitive Computing By Tim Davis, CEO and Founder, Exco InTouch I recently discovered the concept of telehealth was first put forward in the 1920s by Hugo Gernsback. Gernsback was a pioneer in both radio and publishing and in 1925 wrote an article in one of his magazines proposing a device called the “teledactyl” which would allow doctors to not only see their patients through a view screen, but also touch them from miles away with spindly robot arms1. The premise behind the concept was not too dis-similar to the issues we face today predicting in 1925 that “the busy doctor, fifty years hence, will not be able to visit his patients as he does now” so technology could provide the conduit to allow patients to come to him. This visionary concept was clearly ahead of its time as the technology required to achieve such a service was not sufficiently advanced, nor was it adequately available to provide practical solutions for delivering healthcare in remote settings. Although basic computing was being developed, it was not until the 1950s that programmable systems began to appear, allowing simple responses to yes/no styled questions, and even this was not sufficiently developed to provide the depth of information required to make telehealth systems a reality in delivering healthcare. Since then we have, of course, come a long way. Over the last 20 years or so telehealth, and more recently digital health solutions, have become more valuable, and therefore more commonplace in healthcare delivery. The real advance in digital health however has come in the twenty-tens, enabled by the significant advances in technology capabilities combined with its availability across the globe. This has allowed us to develop more advanced digital health solutions that utilise cognitive algorithms to assess a wide variety of data, enabling systems to learn at scale, reason with a clear purpose and interact with users naturally, hence it has now been termed the ‘Cognitive Computing Era’2. Through this era of cognitive computing the age of digital health is set to spiral. It is predicted that by 2020 there will be over 44 zettabytes (1021) of data collected from 20bn connected sensors via the Internet of Things each year3. With this quantity of data collection options, the possibilities for health data are almost endless, giving rise to the concept of the ‘Internet of Medical Things’, the ability for each of us to be connected via a raft of sensors, from wearable technology through to biomedical sensors that report back on physiological data such as medication release, body temperature and heart rate. This will enable people to monitor their health as they walk around their home, with sensors continually reporting key measures to track progress. For digital health providers, the challenge will be to develop programs that are able to take this level of data and use
32
December 2015
it to provide meaningful feedback to patients, their caregivers and healthcare professionals. To achieve this there needs to be a paradigm shift away from the relatively basic, consumer driven health apps that we see today, towards interactive programs that use a variety of source data (for example medication tracking, medical devices or biometric sensors) alongside patient assessments – diaries and questionnaires – to create a complete picture of patient progress. It is essential to tailor personalised responses, for example, progress feedback, automated alerts to care teams, relevant content and encouragement, in order to make the service interactive and personalised to each individual. Through this approach digital health programs can be designed to engage users over extended periods of time, which is key to long term health improvement. However, we should also be aware that people are fallible, and no matter how interactive and engaging a program is, life will get in the way so adherence to the program may fall from time to time. In that sense we should accept that patients are allowed to fail, in fact we should plan for them to fail and build in mechanisms to help them come back on board and re-engage even after significant periods of inactivity. Beyond this, we understand the psychology of user behaviour. For example trust in the service provider is key; in a talk at the recent mHealth Summit in Washington DC Pfizer reported that 90% of users said they would use a health app if their doctor had prescribed it, but were less trusting of health programs provided by pharmaceutical companies, with 76% of patients stating they would expect to see evidence that Pharma had evaluated the value of a program before they would commit to using it. We have come a long way from the original telehealth concept proposed by Gernback in the 1920s, technology is now widely accessible across the globe and sufficiently advanced to make a valuable contribution to modern healthcare. The challenge, and enormous opportunity for healthcare providers, is how we harness this level of health data and yet still develop mHealth programs for individual users. How can we truly personalise mHealth programs to each individual patient and how can we minimize the barriers to long-term engagement? To me, the answer is to start from the insights into users’ lives, then develop health programs that incorporate features which address their needs. Through this approach we can make it simple, make it relevant, but most of all, make it enjoyable to use. References
1. Telemedicine Predicted in 1925: With video screens and remote control arms, any doctor could make a virtual house call http://www.smithsonianmag.com/history/telemedicine-predicted-in-1925-124140942/#BCx5S06Awzkoxirp.99 2. Smart Machines: IBM’s Watson and the Era of Cognitive Computing https:// books.google.co.uk/books?hl=en&lr=&id=U9arAgAAQBAJ&oi=fnd &pg=PA301&dq=cognitive+computing+era&ots=LGvK5V6pla&sig= oAHgjTQguwcYo3_0Jm_3hvC3Lc#v=onepage&q=cognitive%20computing%20era&f=false 3. The Digital Universe of Opportunities: Rich Data and the Increasing Value of the Internet of Things http://idcdocserv.com/1678 n