The pace of innovation and investment in health tech appears to accelerating. In Q3 of 2018 $3.3 billion dollars in deals were made – making it the largest quarter ever for digital health investment. Activity is not limited to OECD countries; we just completed a scan of start-ups innovating in product distribution in four African geographies and surfaced 30 companies seeking to change the way products move using technology. Of course, digital health innovation extends beyond product distribution into other realms of health care delivery, and companies are innovating in a wide number of consumer, payer and provider-facing services across Africa, SE Asia, India and beyond.
In emerging economies, the evolution towards technology-enabled business processes promises unprecedented visibility into utilization, provision, payment, adherence and more. Data from the health sector could be combined with data on consumers’ locations, demographics, purchasing patterns, cell phone usage, internet usage, social networks, etc. to improve the targeting and delivery of information, services and health financing. The possibilities are very exciting, to say the least.
Over the past year we have supported several clients to better understand and optimize health technology platforms in LMICs. It is our experience that in these discussions two important policy challenges are often under-considered:
How do we protect consumers’ data and privacy? National policies need to be crafted and enforced to ensure companies operating in health technology are properly obtaining consent, protecting privacy, storing data securely, sharing data appropriately, maintaining and destroying records properly, and more. These policies should ensure consumers are fully protected, as non-health data (not currently governed by policies like HIPAA) are likely to be increasingly important in the planning and provision of care. As the type and use of data changes, several key challenges with traditional approaches to consent, storage, anonymization and more arise (see below). One can imagine that a lax approach to data protection in contexts where human rights are tenuously enshrined could have quite severe consequences, especially for vulnerable populations.
How do we ensure novel visibility generated by these private companies systematically informs the practice of public health? New data being generated by digital health services is likely to have commercial value. In many OECD countries data aggregators like IQVIA (formerly IMS Health) collect, analyze and sell detailed data on prescribing practices back to manufacturers to inform marketing and sales strategies, generating billions of dollars in annual revenue. However, these data are not systematically available to the public sector to inform planning, purchasing or provision of care. Can governments in rapidly digitizing economies take a different path?
Facebook’s revision of their mantra “move fast and break things” to “be bold and proceed” recognizes the complexity of technology-enabled disruption. As the European Data Protection Supervisor Giovanni Buttarelli stated in 2015, “We live in a world where our digital lives can be acutely analysed. Today, the division between information about our health and information about the rest of our digital lives is disappearing: technology solutions allow devices and apps to connect the dots between different data about us such as location, nutrition and medical. We can put a lot of trust in technology companies to do the right thing with our personal information and to make our lives easier. But we need to have a critical debate about the uses of our personal information that are and are not acceptable to us.” As interest in health tech in LMICs grows, we urge donors and governments to invest in ensuring these innovations protect and serve patients’ interests and strengthen national health systems.
Some of challenges that arise as the use of data in health shifts (1)
- Data are never fully anonymized – re-identification is possible and sometimes quite simple, especially when data sets have been linked or combined.
- Obtaining fully informed consent is very challenging because when data are collected and stored, future uses cannot be fully anticipated.
- A broad range of data can be used to make health inferences about individuals, thus data collected for non-health purposes (which often times is protected differently, especially in the US) can be used to make health-related decisions.
- It is very difficult for individuals to know what data are used to make inferences or predictions about them, and to take action if they feel the data or algorithms have been incorrectly used.
- The concept of “personal data” is changing. Data can have personal implications for a wider range of people than the individual from whom the data are collected, and consent was obtained.
(1) Aicardi, Christine et al. “Emerging ethical issues regarding digital health data. On the World Medical Association Draft Declaration on Ethical Considerations Regarding Health Databases and Biobanks” Croatian medical journal vol. 57,2 (2016): 207-13.