Guest Post: Angela Ballantyne and G. Owen Schaefer
Health systems are producing exponentially more data about patients and there is increasing demand to use that data – for predictive modelling, precision medicine, funding decisions and health system design. One of the features that makes data exciting is it flexibility – it can be repurposed, combined and recombined, shared across sectors and jurisdictions. Health data can be combined with a wide variety of other data, from both the public and private systems. A traditional health research model conceives of research data as something that sits within a research study framework – to be collected, analysed, stored and then destroyed. But in modern research, the data sits outside the confines of a specific project and has a life-cycle of its own. Data may be initially collected for clinical purposes but can and is easily adapted for administrative or research purposes. Examples of systematic gathering of clinical data for administrative and research purposes include the Scottish Primary Care Information Resource (SPIRE) and, more controversially, NHS England’s Care.data system that was shuttered in 2016 after public outcry.
All of this innovation poses challenges for consent models concerning the use of patients’ health information (de-identified or otherwise). How to make consent work in the age of data, and when consent requirements may be waived, are vexing ethical and policy questions.
A traditional research ethics paradigm defines the use of potentially identifiable patient data as observational research that prima facie requires individual consent. Many jurisdictions adopt this view, but allow Institutional Review Boards (IRBs) or Research Ethics Committees (RECs) to waive the consent requirement, typically on the grounds that consent would be impossible or impractical (in terms of numbers, time or cost) and/or would impede the scientific validity of the study (in research where the entire dataset is required). These waivers suggest that patient autonomy and consent are not absolute, and that the interests of patients can be overridden when risks are low and there is a compelling public interest in the research.
In this paper we consider recent literature that proposes that citizens have an ethical obligation to participate in research, specifically that citizens have an ethical duty to share their health information for research purposes. We consider three of the most powerful objections put forward in the literature to the proposed obligation to contribute to research, or to the enforceability of such an obligation. Our aim is to show that these objections are not compelling in the case of data research. Therefore some who believe that the obligation to participate in research is not enforceable in general, could support enforcement in the specific case of data research.
The upshot of this proposal would be to re-orientate IRB/REC review processes away from requiring consent for data use, towards a more robust review of the proposed public good of the research, opportunities for public engagement and research transparency. Some jurisdictions already require that a waiver of consent meet a public good criterion. But we contend that the concept remains under-defined in regulation and under-theorised in the bioethics literature.
The later part of the paper fleshes out what we mean by public good. We present minimal requirements and factors that should be considered. The minimum requirements are that there be no restrictions on publications; in addition to a lack of patents or other use-restricting copyrights on results. Health data is a strategic asset of huge commercial value, and regulatory models need to ensure that public system datasets are carefully governed with a view to promoting public good, not simply private intellectual property. In addition there are four factors an IRB/REC should weigh in their assessment of the potential public good of the research:
- Whether the research addresses health inequities or the needs of disenfranchised or vulnerable patient groups
- Whether the research addresses the needs of groups traditionally excluded from clinical research and therefore lacking in robust evidence to guide clinical care
- Whether the research will be made publicly available and open-access
- Whether the research has engaged with patient groups in terms of pre-research consultation, ongoing partnership and/or patient involvement with the research governance.
In many ways our revised regulatory model is actually more demanding than existing criteria for waivers of consent for the use of patient health data. The key difference is that instead of focusing on the practicality or otherwise of gaining consent, the review process would focus on whether the research will serve the public good. Our approach is grounded in a sound ethical duty to share data, and is a legitimate policy insofar as it enforces that duty without unduly burdening potential subjects of the data research.