Digital Health at a Tipping Point, at the WEF AMNC 2019

Описание: Digital health has been an industry buzzword for over a decade, but new regulatory approvals for devices such as the Apple watch, major investments and acquisitions such as Roche's $1.9 billion purchase of the 5-year-old health data start-up Flatiron Health, suggest that a new era of real-world evidence is here and is poised to change every sector of the global healthcare industry, doctor-patient relationships and the access to and affordability of treatment.

Panellist
Chen Kuan, Infervision
Pius S. Hornstein, Sanofi China
Liu Jiren, Neusoft Corporation
Jessica Tan, Ping An Insurance

Moderator: Laurent Haug

A large proportion of China’s 1.4 billion people have poor primary care – these include the aged and children left behind as villages “hollow out” when young people migrate to cities. The situation is similar in many developing countries, where the poorest populations are located far from cities and medical hubs.

Telemedicine, especially rapidly improving diagnostics thanks to deep learning and artificial intelligence, is increasingly providing digital solutions in these regions.

Ping An, an insurer, aims to engage more directly and early on with its customers. It also offers a Good Doctor service, which consists of AI-powered kiosks that provide diagnosis and dispense medicines for common conditions. Its Good Doctor app has 25 million users, 550,000 of whom consult its doctors on a given day. Advanced analytics have been used to train its AI assistant in diagnosing 3,000 diseases, which its Village Doctor programme takes to villages. Local volunteers are hired, village doctors are trained, and patients and doctors send their reports on their smartphones and given help to interpret them.

“AI imaging diagnostics is already a fairly mature technology and [is having] a huge impact,” said Jessica Tan, Co-Chief Executive Officer, Ping An Insurance (Group) Company of China Ltd, “The next step is prevention.”

Prevention, along with behaviour change, are the next frontiers for digital healthcare. Prevention (of stroke or cardiac arrest, for instance) and management of chronic diseases (such as diabetes) require attention to patient well-being and care beyond hospital-based treatment, with the key being behaviour change and management.

Sanofi is studying a 50,000-strong patient cohort in China to understand how to improve patient outcomes and treatment adherence, and how to ensure follow-up, which has been shown to significantly improve treatment outcomes – at a fraction of incremental costs – and could also be interesting for insurance companies.

Collaboration is key – Sanofi has partnered with Google, for instance – because the promise of digital health is “mostly about data,” said Pius S. Hornstein, Country Chair, China, and General Manager, Pharmaceuticals, Sanofi China. Chemical, biological and physical science and data engineering are converging today and can detect more precisely how treatment regimens and behaviour influence patients’ health. “Digital healthcare is a tsunami; with 5G it will increase even more,” Hornstein said.

Yet, to enable collaboration, some thorny questions must be settled, such as setting standards for data quality, and rules for data ownership, access and use, for which regulations are still evolving and far from uniform. Often, the biggest hindrance is hospitals’ reluctance to share data, said Liu Jiren, Chairman and Chief Executive Officer of Neusoft Corporation, one of China’s largest healthcare service providers. Another challenge is for digital technology to mesh with the local ecosystem and culture for people to adopt it, he said.

Doctors’ reluctance to share data is a challenge that AI and deep learning start-up Beijing Infervision Technology Co. faced and eventually overcame to launch “a huge wave of innovation,” said Chen Kuan, its Founder and Chief Executive Officer. Currently, the company’s service is available in 300 different hospitals in eight countries including Switzerland. Services include pre-screening (before the doctor performs a diagnosis, with the result available as a reference), and quality check or control (after physician is finished with their day’s work).

Yet, many tricky questions remain:

How to measure the accuracy of digital diagnostics; the consensus is to validate healthcare applications of AI with the same process of academic proof, peer review, clinical review of safety and efficacy as other medical applications
How to ensure follow-up treatment and long-term care
How to get countries to agree upon common standards and enable interoperability
How to ensure high signal quality for accurate remote diagnosis
How to manage real-life experience of people in an app ecosystem focused on a single user
How to digitalize hospital data, make it compatible, comparable, useable at the national and international level