From online grocery shopping to smart watches, most of us generate some form of health data outside of the healthcare system. Even during the depths of COVID-19, anxious internet searches for symptoms and GPS-tracked walks within 5km limits generated huge amounts of data and took on new significance as digital tools for infection control. So what if we could harness this data to improve healthcare outcomes?
Historically, the generation and collection of health data has been confined to hospitals, clinics and the like. However, in today’s digitised world, individuals are increasingly generating personalised health-related data outside of formal healthcare settings, either intentionally via health tools (fitness trackers, home monitoring devices) or passively through environmental sensors, phone apps and even online search activity (See our article 'Should Googling your illness be a medical device?'.
According to a recent report by Cambridge think-tank the PHG Foundation, this citizen-generated data — or CGD — has enormous potential to inform preventative healthcare. However, its use raises uncomfortable questions about health inequality, privacy, consent and regulation.
Given prevention is a key pillar of our Long Term National Health Plan, Australia must now decide if harnessing CGD is the right move to advance our digital health landscape, or if its problems in application outweigh its potential.
Making the case for Citizen-Generated Data (CGD)…
Virtually all of our current health data comes from patients presenting to the healthcare system with symptoms. Therefore, a challenge for predictive prevention — that is, the use of personal data to predict future illness — is how to gain information about people who are not yet engaging with the healthcare system to prevent or delay ill-health. CGD fills this gap by collecting health data produced by individuals outside of health settings and using this to identify and inform those at increased risk of ill-health before it worsens.
While having a personalised prevention tool at the touch of your fingers seems futuristic, CGD is in fact already part of some preventative health landscapes. Abroad, CGD has had success in providing data-driven marketing to prevent mental health decline on social media, predicting schizophrenia relapse through phone monitoring, and — confirming our worst fears about googling our symptoms — using online search history to predict cancer diagnoses.
Even in Australia, we have some familiarity with CGD from the COVIDSafe app, which combines location data from COVID-positive individuals with information about other proximate app users to determine at-risk contacts. Questionable efficacy aside, it shows the innovation potential of aggregated CGD to address population health concerns.
Presently, most uses of CGD are ad hoc and locally driven. If Australia wants to get serious about its National Digital Health Strategy — which lags behind other jurisdictions on digital innovation — then harnessing CGD for predictive prevention is one avenue to regain our footing as a global health leader.
Citizen-Generated Data…at what cost?
Like all medical tech, CGD has its issues.
Digital health initiatives are unlikely to benefit all people equally. While using CGD could reduce health inequalities by increasing access to health information, it can also further disadvantage groups who may be unable to use digital devices, or for whom financial, physiological or psychological constraints prevent engagement. If we only collect data from people with the resources, skills and literacy to access CGD tools, then others may miss out on clinically relevant insights.
Privacy and consent are also big-ticket complaints. In particular, the unauthorised sharing of CGD and the pervasive nature of passive data capture stoke typical Orwellian fears. Although people are usually prepared to accept some privacy loss for a clear short-term benefit, this enthusiasm wanes in the face of opaque, long-term data capture policies, particularly where data is not de-identified.
As such, transparency and public trust are critical to the successful use of CGD. Ensuring clarity around who is using CGD, for what purposes, and which parties will benefit will help the community develop realistic expectations of the short- and long-term benefits and potential risks arising from using CGD. Only then can the public make informed decisions about contributing to CGD.
Private data, public ends
Most CGD is generated by devices and services provided by commercial entities, so gaining access to personal data from the private sector is a challenge for the healthcare system. Who should control the dissemination of step-count data on your phone — you or the software creator? One option to address this issue is to increase the level of open-source CGD (for example, public social media feeds), but this doesn’t work for all data types. Data sharing arrangements between healthcare providers and private companies are another option, but these must be highly transparent to overcome public scepticism.
As CGD use is still in its infancy, our healthcare system has little control over the tools developed by industry, meaning they often do not align with its needs. Of course, private companies currently design most CGD devices with technical and commercial viability in mind rather than public health interests. If our healthcare system wants to prioritise using CGD, then it needs to engage with industry partners to ensure that mutually beneficial tools are created — tools that address the areas of greatest clinical need while also meeting consumers desires.
And then we turn to regulation. The challenges that exist with in-system health data — such as quality, safety and security standards — will also apply to the data collected by commercial providers. Regulation will be key to the success of CGD in Australia, however the Therapeutic Goods Association’s lengthy approval process and antiquated approach towards software-based medical devices is likely to disincentivise public-private collaboration in the CGD space.
An opportunity for leadership
Using CGD in preventative health asks us to accept blurring the boundary between ‘patient’ and ‘citizen’, revamping discussions around the commodification and commercialisation of health. In this domain, Australia is at a crossroads.
We can take a cautious approach and wait for leadership abroad, or we can begin to address structural and cultural barriers which prevent Australia from leading the world in digital health innovation.
Either way, the use of CGD to inform our healthcare practices — and the ever-growing pool of technology involved — isn’t going away. It is up to us to determine if we want to harness it.
Authors: Tamara Ernest, Madelaine De-Iudicibus and Peter Waters
Read More: Citizen generated data and health