When eHealth goes viral: The strengths and weaknesses of health tech during COVID-19
Written by Vasileios Nittas, Doctoral Researcher, University of Zurich, Epidemiology, Biostatistics and Prevention Institute & Editor-in-Chief, YRE Series
As the world slows down, the demand for eHealth is going viral; and it is happening fast and globally. Digital solutions, such as telemedicine and online crowdsourced health monitoring, are quickly shifting from a previously slow adoption path to a record pace of uptake. While the temporary new normal settles in, the digitisation of our healthcare is being put on a tough and revealing test; ultimately showcasing the strengths and weaknesses of current eHealth infrastructures.
In response to COVID-19, key health authorities, such as the CDC and the World Health Organization strongly lobby for ways to minimize physical contact between patients and healthcare providers, also known as “medical distancing”. Telehealth services are rapidly becoming a major force in the effort to reduce healthcare-related COVID-19 transmissions, and ultimately protecting our health personnel. The effectiveness of telemedicine has been promising for many health areas, including diabetic care, dermatology and cardiology; allowing for high-quality remote care, while saving time and valuable physical space.
While the evidence for infectious disease telemedicine is not equally strong, its role in the current crisis is expected to be essential. Experts, such as Professor Schwamm from Harvard’s Medical School, believe that telehealth can indeed contribute towards slowing down infections and flattening the current epidemic curve. Symptoms and disease recovery can be monitored through calls or video chats, keeping low-risk and mild-symptom patients, who are also more likely to spread the virus, at home.
Beyond increasing the demand for telehealth, COVID-19 also sheds light on the persisting weaknesses of our telemedical ecosystems. A recent study highlights that costs and reimbursement uncertainty are key barriers to the use of telemedicine, often leaving patients, as well as healthcare providers in the dark regarding payments and insurance coverage.
Other barriers are strongly linked to patient characteristics, such as age and educational background, bearing the risk that the less digitally savvy (e.g. elderly), who are also the most vulnerable, will benefit less by existing telehealth solutions. Uncertainties around legal liability (e.g. lacking rules on who will be held responsible in case of damage or malpractice), as well as loopholes around privacy and confidentiality, pose additional challenges that need to be swiftly addressed as uptake increases. These barriers increase the risk of unmet patient and provider expectations. Failing to meet expectations reduces public trust and hurdles the use of telemedicine as an effective and high-quality substitute for (at least some) physical care.
Crowdsourced disease monitoring
The exponential spread of COVID-19 highlights the need for timely tracking of those who are infected, and ideally, their contacts. Timeliness, as well as flexibility, are common weaknesses of traditional surveillance systems, who often rely on delayed reports by healthcare providers and laboratories; even moreso in the context of a distressed and overburdened system. This is where crowdsourced disease monitoring, also known as participatory surveillance, comes in to fill the gaps.
“Crowdsourced” means that relevant health information is provided by a large number of people, primarily through the internet. The “crowd” can either be actively engaged, such as when answering online surveys or sharing data through smartphone apps and wearable sensors; or be a passive information source, which is the case with publicly available social media data (e.g. tweets).
A successful example of an actively crowdsourced surveillance system is Australia’s and New Zealand’s FluTracking platform. FluTracking is a simple and fast online survey, relying on weekly submissions of quite a large number of volunteers, referred to as “FluTrackers”. The system, recently adjusted to COVID-19, aims to complement traditional flu monitoring while providing early outbreak warnings and contributing to research.
Similar efforts, but targeted to COVID-19 are currently being introduced around the world, aiming to provide a better picture of disease spread, as well as early warnings of future waves. The real power of crowdsourcing lies in its ability to deliver near real-time information, which in turn enables prediction and response in a flexible, usually low-cost, fast and engaging manner.
With real power come real risks. If crowdsourced disease monitoring is to have an impact on current and future epidemics, several of its weaknesses require urgent attention. The first weakness lies in the demographics of the “crowd”. The group of people that is online, digitally engaged and willing to share health information is often quite different from the general population or the population of interest. Users tend to be younger, richer, healthier and more educated, leaving those that are often most vulnerable (e.g. elderly) unaccounted for.
The second weakness lies in the “crowds” data. The internet, and particularly social media platforms (e.g. Twitter), provide a nurturing environment for the rapid spread of false signals and bias. Research shows that certain events or topics that dominate public discussion (e.g. shortages of influenza vaccinations) lead to increased internet (and social media) activity around a topic (e.g. influenza tweets). Analysing these data without appropriate filtering and cross-checks may indeed provide false alarms of disease outbreaks where there are none. Non-credible information can be easily picked up by a few influencing individuals and then quickly amplified by larger crowds, causing enough “noise” to cover up true disease signals.
These risks, as well as issues around privacy and confidentiality, are part of the reason why crowdsourced surveillance faces scepticism by public health officials and policymakers. Despite its potential to complement traditional disease surveillance, lower trust means less use and less use means less impact. Understandably, the last thing a decision-maker would want during a pandemic is misinformed action, the consequences of which (e.g. fear and panic) can spread faster than the virus itself.
The needs of technology in a crisis
The conclusion is clear; COVID-19 has forced eHealth into the lives of many. The influence of health technology in the fight against this pandemic is expected to be significant. However many of these solutions, such as telemedicine and crowdsourced disease monitoring, will likely fulfill their full potential after some of the remaining “needs” are immediately addressed.
We need well defined and easy-to-understand guidelines for the day-to-day use of telehealth technologies in the context of COVID-19. We need these to be adjusted to the most vulnerable (e.g. different age groups) and their needs. Such guidelines need to keep the expectations of users, be it patients or healthcare providers, realistic. They need to convey the message that eHealth solutions are a viable alternative in times of this pandemic and beyond; however, probably not suitable for all problems that arise and certainly not a full replacement of traditional care.
We also need new laws and regulations on reimbursement and liability issues of telemedical care; as well as their clear communication to professionals and the public. We also need guidelines and regulations on how to use online crowdsourced disease monitoring data to complement traditional systems and profit from the internet’s timeliness and flexibility. This requires more research and robust methods on validating online information while ensuring that human rights, privacy, and confidentiality are maintained.
Finally, we, the ones who describe ourselves as “digital natives” need to share our knowledge and skills to help to make eHealth a viable and accessible solution for those around us. We are all in this together, and we can all contribute.
Originally published here.
Originally published at https://digileaders.com on April 7, 2020.