49th St. Gallen Symposium "Capital for Purpose": 8–10 May 2019


“How are you feeling today, human?"

Data analytics have begun to transform the world of medicine. An algorithm that can scan databases of randomised controlled trials, as well as actual patient outcomes, is something pretty revolutionary, if you think about it. If the whole world used data analytics, then a doctor’s visit could potentially leave you with more answers than questions and concrete treatment plans rather than the usual “we could try you on this.”

Rather than destroying healthcare jobs, in other words, AI has the potential to create the time and tools needed to make doctors more “human” again. “Data analytics and deep learning, transfer learning, machine learning, all these algorithms, are means for me to make best use of the existing data,” says Claudia Süssmuth Dyckerhoff, a member of the board of directors for Roche Pharmaceuticals who holds a PhD in business administration from the University of St. Gallen.

Claudia Süssmuth Dyckerhoff at the 48th St. Gallen Symposium

Receiving a whole stream of data, including information collected by at-home wearable devices that will feed back to the hospital and track a patient’s progress, means that a doctor’s job will include a lot less uncertainty. “It’s not about replacing jobs,” Süssmuth Dyckerhoff explains, but “rather about using analytics to improve support.”

But here’s the fear: if AI can scan databases of every treatment plan, every randomised controlled trial, every case study ever, will certain jobs in healthcare become meaningless? Can a robot do a better job of diagnosis and treatment plans than a human ever could?

The answer is not so clear cut. Radiologists, for example, may be at risk, but according to Süssmuth Dyckerhoff radiologists are in short supply in countries like China to begin with. “What is important is training your own people to be ready,” she says. “Some people need to be retrained. More data could even create jobs.”

Technology is already making significant inroads. A team at Stanford University recently introduced an AI therapist dubbed “Ellie.” The programme uses voice recognition, image recognition, and can scan hundreds of databases to provide appropriate responses to the patient. “Machine learning has been going on for years, but it’s still early in terms of its use in healthcare,” says Harvard University medical student Joshua Onyango. “If this is what it can do so far, in terms of providing very personalised healthcare, what can we imagine it doing in 20 or 25 years?”

Eventually, experts think robots might even allow doctors to be more human, so to speak. When it comes to the innate uncertainty that surrounds disease, treatment, side-effects, and mental health, sometimes a patient really needs a person to talk to. But physicians in today’s overburdened healthcare system struggle with that task. Sometimes doctors simply just do not have the resources or time to be as holistic as we want them to be.

Orest Firsov, an entrepreneur at New Body Technology, a digital healthcare service that tracks posture for users – many who have chronic health conditions – points out the fundamental benefit of AI in medicine. Automation that takes away the burden of “routine checkups and assessments” means doctors have more free time to be both more competent and more caring. Understaffing has left healthcare workers unavoidably robotic in practice, and perhaps we need AI to recieve the burden of routine work as a way to restore humanity to medicine.

Not to be taken lightly, this time to humanise the patient could save healthcare jobs in the age of AI. “One thing that machines are going to have a very difficult time doing,” Onyango says, “is understanding the cultural and emotional nuances that make us human.” Tasks such as navigating social norms, symbolism, body language and empathy – all the little things that make us human and separate us from AI – will be difficult for robots to accomplish.

According to Valeriia Kasatkina, a Leader of Tomorrow researching the connectedness of Australia’s social support systems for domestic abuse victims, AI could also help mobilise policy and care. If we have data analytics to help improve communication between the healthcare and social work sectors, then the role of the social care worker will become more efficient.

“On the organisational level, using data collection and information sharing plat¬forms could provide relevant information about victims,” Kasatkina says. This is vitally important for issues in cases of domestic violence. “Victims could have mental health issues, a child, or problems with drug addiction,” Kasatkina says. “Drawing connections means all of their health and social care services could be coordinated.”

The world of medicine and healthcare is changing, but for now, it seems that the changes are for the better. As long as doctors, healthcare providers and carers can fill the gaps left by machines, providing patients with the human touch, then there really is room for both man and machine in the future of healthcare. “The hospital will always be a space with a human touch,” Süssmuth Dyckerhoff says. “Health is so human.”

“The hospital will always be a space with a human touch. Health is so human.” – Claudia Süssmuth Dyckerhoff
Leaders of Tomorrow at the 48th St. Gallen Symposium: Orest Firsov, Joshua Onyango and Valeriia Kasatkina