If you build it, they will come: unintended future uses of organised health data collections

Abstract

BackgroundHealth research increasingly relies on organized collections of health data and biological samples. There are many types of sample and data collections that are used for health research, though these are collected for many purposes, not all of which are health-related. These collections exist under different jurisdictional and regulatory arrangements and include:Population biobanks, cohort studies, and genome databasesClinical and public health dataDirect-to-consumer genetic testingSocial mediaFitness trackers, health apps, and biometric data sensorsEthical, legal, and social challenges of such collections are well recognized, but there has been limited attention to the broader societal implications of the existence of these collections.DiscussionAlthough health research conducted using these collections is broadly recognized as beneficent, secondary uses of these data and samples may be controversial. We examine both documented and hypothetical scenarios of secondary uses of health data and samples. In particular, we focus on the use of health data for purposes of:Forensic investigationsCivil lawsuitsIdentification of victims of mass casualty eventsDenial of entry for border security and immigrationMaking health resource rationing decisionsFacilitating human rights abuses in autocratic regimesConclusionsCurrent safeguards relating to the use of health data and samples include research ethics oversight and privacy laws. These safeguards have a strong focus on informed consent and anonymization, which are aimed at the protection of the individual research subject. They are not intended to address broader societal implications of health data and sample collections. As such, existing arrangements are insufficient to protect against subversion of health databases for non-sanctioned secondary uses, or to provide guidance for reasonable but controversial secondary uses. We are concerned that existing debate in the scholarly literature and beyond has not sufficiently recognized the secondary data uses we outline in this paper. Our main purpose, therefore, is to raise awareness of the potential for unforeseen and unintended consequences, in particular negative consequences, of the increased availability and development of health data collections for research, by providing a comprehensive review of documented and hypothetical non-health research uses of such data.

Choices on selective clinical data sharing by people with Parkinson’s disease

“…6 Although, data collection is regulated, different data collection mechanisms exist under different governance regimes. 1 Depending on the circumstances for which the data is being collected, governance can be a result of law, public health ethics, ethics of clinical care, research ethics, or commercial contracts between individuals and service providers. 1 That said, many medical apps and devices which perform organized collection of personal data are not subject to government regulation and fall short of data protection standards.…”

“…1 Depending on the circumstances for which the data is being collected, governance can be a result of law, public health ethics, ethics of clinical care, research ethics, or commercial contracts between individuals and service providers. 1 That said, many medical apps and devices which perform organized collection of personal data are not subject to government regulation and fall short of data protection standards. 7,8 It is, therefore, immensely important that PwP define how they want their data to be used.…”

“…Digitally generated and stored data has greater longevity than analog data, being potentially infinite in lifespan and uncapped in its capacity for reuse and re-analysis. 1 Collection of this information may be useful in the identification of patterns in diseases, differences in populations, and may ultimately help to identify unmet needs and shape research questions. 2 Unsurprisingly, therefore, this approach has the potential to cause significant impact in numerous disease areas, including Parkinson's disease (PD).…”

“…1 Such data and its sources include, but are not limited to, clinical and public health records, fitness trackers, health apps, biometric sensors, websites, and social media. In the vast majority of cases, the data is in a digital format.…”

Introduction to the article collection ‘Translation in healthcare: ethical, legal, and social implications’

“…The authors chart four waves of biobanking management and policy that have resulted in response to the challenges of ensuring informed consent, standardization, sustainability and public trust [27]. Addressing similar concerns by drawing attention to the lack of regulatory oversight of unauthorized secondary uses of health data and samples, O’Doherty et al argue that without building ethical oversight in tandem with the proliferation of health data collections intended for research, unintended negative consequences are likely to result [28]. …”