Transformative technologies, medicines and insurance that could vastly improve the reach and quality of healthcare are on the horizon — but they also bring new risks and trade-offs for health systems and societies. The 2020 Global Risks Report, produced by the World Economic Forum in partnership with Marsh & McLennan and Zurich Insurance Group, discusses the benefits of these new breakthroughs, as well as the challenges they present.
Over the centuries health systems have embraced many innovations, sometimes without waiting for them to be proven safe and effective. Healthcare providers and payers are already using today’s emerging technologies — machine learning and artificial intelligence (AI), sensors, digital therapies, telemedicine and so on — to support both clinical and operational decisions: to triage symptoms (1), interpret diagnostic tests (2), create personalized treatment plans (3) and predict re-admissions at a hospital or epidemics in a population. Combined with human capacity, these technologies could ultimately make it possible for everyone — even in currently fragile, over-burdened health systems — to access high-quality, consistent, affordable, timely and convenient care.
But new technologies also raise risks, including risks of compromising patient safety and privacy, as well as introducing bias. Errors by individual health workers affect only their patients, whereas the consequences of AI errors could unfold at a whole new scale. Since training data sets in health often skew white and male (4), AI could fail to spot symptoms or devise effective treatment plans for everyone else. These outcomes will be tough to predict or avoid because AI’s black-box nature makes it difficult to understand how it reaches conclusions — making it hard to spot bias. Health data are especially vulnerable to cyberattacks (5), with risks of individuals being identified even from anonymized data (see Chapter 5, Wild Wide Web).
Highly complex, specialized new drugs promise radically better treatment for devastating diseases — but they come at exorbitant prices. For example, three recently launched cell and gene therapies cost up to USD 2 million per patient. Over the next few years, between 15 and 30 new million-dollar drugs are expected to enter the market, mostly for cancer (6). New pricing models — such as multi-year payments contingent on patient outcomes — are starting to emerge to address the high costs and risks of these treatments.
But health systems are finding it difficult to adapt amid questions over who should pay, how high a price can be justified, and what can be given up to afford new therapies. As people’s expectations rise, unequal access to better therapies could deepen health inequalities within and across countries, eroding trust in health systems and societal cohesion. In the longer run, if (or when) gene-editing technologies become available to enhance physical, cognitive or behavioral capabilities, these could result in a society of genetically enhanced haves and the merely natural have-nots.
Risk pools of one.
Health insurance looks set to be transformed by big data and analytics. As with in-car devices used by car insurers to reward responsible drivers with lower premiums, health insurers can (with the consent of customers and the appropriate levels of data security), capture, store and analyze personal health and behavioral data from wearable — and eventually implantable — devices. Personalized risk assessment could lead to rewards and incentives for people to live healthier lifestyles, but if unchecked by regulation, it could also potentially put insurance beyond the reach of people judged to be higher risk for genetic, environmental or behavioral reasons.
In some jurisdictions, steps have already been taken to mitigate this risk in response to the concerns of people who have taken predictive genetic tests for certain diseases. In 2018, the UK Government, together with the Association of British Insurers (ABI), consolidated existing agreements on the use of genetic information and created the Code on Genetic Testing and Insurance. Given the rapid advances taking place in genetic research, this Code will be reviewed every three years to consider the technical, ethical and societal implications of insurability. Among other principles, the Code commits insurance companies to treat applicants for insurance fairly and not require or pressure any applicant to undertake a predictive or diagnostic genetic test (7).
1 Babylon GP at Hand application. https:// www.gpathand.nhs.uk/
2 Davis, N. 2019. “AI Equal with Human Experts in Medical Diagnosis, Study Finds”. The Guardian. 24 September 2019. https://www. theguardian.com/technology/2019/sep/24/ ai-equal-with-human-experts-in-medical-diagnosis-study-finds
3 Wels-Maug, C. 2019. “HUG to Become First University Hospital Using IBM’s AI Tool to Personalise Cancer Treatment”. Healthcare IT News. 12 September 2019. https:// www.healthcareitnews.com/news/europe/ hug-become-first-university-hospital-using-ibm-s-ai-tool-personalise-cancer-treatment
4 Gershgorn, D. 2018. “If AI Is Going to Be the World’s Doctor, It Needs Better Textbooks”. Quartz. 06 September 2018. https:// qz.com/1367177/if-ai-is-going-to-be-theworlds-doctor-it-needs-better-textbooks/
5 Lee, K., J. P. Raman, W. Hedrich, R. Lam and P. Daga. 2018. Holding Healthcare to Ransom: Industry Perspectives on Cyber Risks. Marsh & McLennan Companies. http://www. mmc.com/insights/publications/2018/jul/ holding-healthcare-to-ransom-industry-perspectives-on-cyber-risks.html
6 MIT NEWDIGS. 2018. “Projections from the Existing Pipeline of Cell and Gene Therapies: Launches and Patent Numbers”. Research Brief 2018F210-v027-Launches. https://newdigs.mit.edu/sites/default/ files/FoCUS%20Research%20Brief%20 2018F210v027.pdf
7 ABI (Association of British Insurers). 2018. Code on Genetic Testing and Insurance. https://www.abi.org.uk/data-and-resources/ tools-and-resources/genetics/code-on-genetic-testing-and-insurance/