Opportunities for a national genomic data governance framework in Australia: a systematic review

Abstract

Background

As with other countries, Australia is seeking to make efficient use of genomic data for use in research, clinical medicine and population health. However, to enable cross jurisdictional consistency in the management of and access to data, it will first need to establish a national framework for governing genomic data. To this end, ethical, legal and social issues are often discussed. However, the literature offers little evidence-based support for such a framework.

Methods

To address this literature gap, we systematically reviewed two databases (Scopus and PubMed) for research articles that discussed issues and opportunities for enacting genomic data governance frameworks in the domains of research, genomic medicine and public (population) health in the Australian context.

Results

Thirty-one relevant articles were included and were analysed using inductive content analysis. Our findings identified that opportunities for implementing a national genomic data governance framework concerned defining roles for patients in data governance, data management processes and increasing the public acceptance of genomic data use in healthcare and research. Additionally, they highlight differences in the opportunities and priorities for clinical and research genomics that hinder further advancement of data governance.

Conclusions

Our synthesis of the current literature on genomic data governance suggests that the current focus on individual consent as the primary mechanism for protecting data subjects and different priorities in clinical and research governance need to be addressed. Given the significance of the role of consent procedures and differences in clinical and research data in generating a data governance framework, our findings hence reveal a critical gap in the research literature. Advancing a national genomic data governance framework will require greater consensus and clarity regarding the application of ethical principles across jurisdictions and institutions.

Introduction

Genomic data governance and regulation are valuable both for producing consistency and predictability in how data are stored, used, and/or accessed, and also in fostering public trust in genomics research and medicine when aligned with community expectations [1]. International organisations, such as the Global Alliance for Genomics in Health (GA4GH), hence encourage the development of standards for genomic data and interoperability between databases across jurisdictions [2]. Yet research and clinical practice internationally are hampered by inconsistencies in regulation that may require negotiation between health consortia to enable genomic data flows across jurisdictional boundaries [3]. In 2017, the Australian National Health Genomics Policy Framework was published and championed a national genomic data governance framework [4]. Movements in many countries towards national genomic data governance frameworks may be seen in the context of broader international regulatory activity towards enabling genomic data sharing between clinical and research contexts, and also across jurisdictional boundaries.

National approaches to genomic data governance have been pursued in some countries but are yet to be realised in the Australian setting. For example, the Pan-Canadian Genomic Strategy identified the need to redress ‘the fragmentation of the data ecosystem’ [5] in its recommendation for the creation of a federated data framework and data governance approach. Additionally, the success of the UK’s 100,000 Genomes Project in coordinating the governance of clinical and research genomics has allowed for cross-disciplinary data sharing between research and a national health system [6]. Such initiatives reiterate the promise of biomedical technology that is advanced by sharing genomic and associated health information between research, clinical, and population and public health settings [2, 7]. However, the realisation of the public benefits of precision medicine and genomics faces ongoing challenges. Such challenges are not only scientific, but also technical, ethical, legal, and sociological.

In Australia, the use of genomics in clinical medicine, research, and population and public health settings is an ongoing focus of government-sponsored initiatives to improve healthcare outcomes and ‘nation-building objectives’ [8]. A 2021 parliamentary report on reforming processes for new medicines and health technologies recommended that the Australian Commonwealth Government should fund a genomics testing program to ‘provide equitable access to genomic testing nationwide’ and ‘ensure the provision of genomics [sic.] counselling for all patients’ [9]. Australia’s healthcare is centred on a primarily public tertiary hospital system that is managed by the seven state and territory governments, with oversight, funding and legislation for healthcare occurring at both state and federal levels. Most Australian jurisdictions’ health departments have publicly available strategies that describe genomic healthcare, highlighting the need for intra-state coordination of genomic data governance [10–17]. Australia’s Commonwealth government has funded the advancement and implementation of genomics into Australia’s healthcare system via research funding programs such as, Genomics Health Futures Mission (GHFM) and Medical Research Futures Fund (MRFF). Australia’s National Health and Medical Research Council (NHMRC) broadly defines and advocates for responsible management of research data by researchers and organisations [18]. It is, however, the responsibility of individual research organisations and their funding recipients to define the framework and practices for genomic data management, which leads to inconsistencies and incompatibility in data management practices. Without resolution, the fragmentation of genomics policy between layers of government and institutions will continue to hamper the delivery of timely and effective genomic healthcare and research.

A key reason for lack of progress in realising a national genomic data governance framework is ongoing tension and uncertainty regarding the Ethical, Legal and Social Issues (ELSI) across jurisdictions [8]. Existing research into these issues has been piecemeal– tackling small discrete aspects of a large and complex problem– and so has not facilitated effective dataset implementation (for example, see: [19–22]). This body of research has also been disjointed, with numerous different groups (for example, ethicists, lawyers, sociologists, patient and donor advocates, educators, policymakers, and genomic scientists) working on problems in isolation. Similar to other forms of medical ‘big data’ research fields, genomics is characterised by interlocking governance issues, technical issues, and ELSI [23, 24]. There is, therefore, an urgent need for large-scale, transdisciplinary research to develop an ethically defensible, legally robust, socially acceptable, culturally safe and responsive, and implementation-ready governance framework for clinical and genomic datasets in Australia. The development of a fit-for-purpose governance framework, based on an awareness of the ELSI surrounding genomic data, is needed to guide the effective implementation of technical solutions.

To work toward reconciling the breadth of relevant research that may contribute to the development of a national genomic data governance framework, we conducted a systematic literature review of peer-reviewed research of genomic data governance. This systematic literature review is guided by a central research question: What are the opportunities for enacting a national genomic data governance framework in the domains of research, clinical medicine, and public (population) health? What might such a framework aim to achieve? To cast light on these questions, we undertook a review to: (i) generate a current picture of the existing landscape of clinical and genomic datasets in Australia; (ii) identify the factors facilitating the collection, storage and sharing of genomic data; and (iii) categorise the opportunities for improvement of genomic data governance at a national level in Australia.

Methods

Search strategy

A PRISMA compliant search of peer-reviewed literature was undertaken using the checklist and methods described in the PRISMA-P 2015 statement [25]. Two scientific electronic databases, Scopus and PubMed, were selected because of their coverage of journals that address both ELSI and technical issues surrounding genomic and broader medical data. The search was limited by the date of publication (1 January 2014 to 5 May 2023) to align the beginning of the search range with the founding year of the Global Alliance for Genomics and Health. The search used the following keywords and syntax: (genetic OR genomic OR DNA) AND (“data collection” OR “data storage” OR “data sharing”) AND (Australia OR Australian OR Victoria OR “New South Wales” OR Queensland OR “Australian Capital Territory” OR “South Australia” OR Tasmania OR “Western Australia” OR “Northern Territory”) AND (survey OR questionnaire OR interview OR “focus group” OR workshop). The search identified 5,628 unique studies across the two databases (Fig. 1).

Fig. 1.

PRISMA flowchart summarising the identification and analysis of articles

Exclusion criteria

The titles and abstracts of the 5,628 studies identified were manually reviewed by a researcher (FC), to assess their suitability for this review, against six exclusion criteria: (i) not about human genomic data storage, collection or sharing in Australia; (ii) not about genomic data derived from living tissue (i.e., data derived from ancestral DNA was excluded); (iii) not a peer-reviewed source; (iv) not clearly grounded in empirical research (i.e., opinion articles and editorials were excluded, but narrative reviews or ethical articles drawing on empirical findings were included); (v) not about genomic research, genomic medicine, or population and public health domains of genomics (i.e., direct-to-consumer services research was excluded); and (vi) does not describe implications for Australian genomic data governance. Genomic data governance is here defined as the exercise of decision-making and authority over genomic (and related) data, which may hence include organisational rules, roles, responsibilities and the management of technical systems to store, share, and manipulate genomic data and related clinical data and metadata. Two further members of the research team (MV, RM) reviewed 5% (n = 280) of the excluded results (titles and abstracts) to assess the application of the exclusion criteria. The exclusion criteria were refined through this process, resulting in a final 31 studies included for summary and analysis and the exclusion of 5,597 studies.

Synthesis of themes for literature appraisal

To develop data themes to categorise and assess reviewed literature, we synthesised categories of governance opportunities from grey literature, such as the health genomics policies noted above, as well as Australia’s National Approach to Genomic Information Management (NAGIM) Blueprint [26]. As shown in Table 1, grey literature from state/territory and Commonwealth government agencies identifies a range of opportunities in their overview of genomic information management. This grey literature consists of a national genomics policy [4], the NAGIM Blueprint [26], as well as most Australian states’ genomics health or clinical genomic strategies [11, 13, 14, 16, 17]. The Tasmanian state health plan [12] and territory health plans [10, 15] are excluded because they only make brief mentions of genomics and/or clinical genetics, rather than having released genomic medicine or clinical genomics strategic plans, reflecting their smaller size relative to the other jurisdictions.

Table 1.

Opportunities for improved governance in genomic data policies

Source*		Description	Opportunities
National
Department of Health, Australia [4]		National genomics policy	Collaborative governance
Queensland Genomics [26]		Blueprint for a National Approach to Genomic Information Management (NAGIM)	National infrastructure can support data sharing; values-based framework to support the interests of a range of communities; richer data sources for researchers; improving quality of care
New South Wales
NSW Ministry of Health [14]		Health genomics strategy	Coordination between states and national government; digitisation of patient medical data in NSW
Victoria
Department of Health and Human Services, Victoria [13]		Health genomics strategy	Coordination between states and national government
Queensland
Queensland Health [16]		Health genomics plan	State-based coordination of genomic services
Western Australia
Amanuel et al. [11]		Genomics strategy	No specific opportunities identified
South Australia
SA Health [17]		Clinical genomics plan	Coordination between states and national government

*Genetics/genomics strategic plans could not be identified for the state of Tasmania, nor for Australia’s two mainland territories, the Northern Territory and the Australian Capital Territory

The opportunities arising from the creation and expansion of a genomic data governance framework referred to in genomic healthcare strategies and plans included three types: (i) opportunities to define the role of patients in data governance, such as improving the quality of care offered through genomic medicine; (ii) opportunities to improve processes of genomic data use, such as through the coordination efforts of national and state-based governments or through utilising widespread systems of digitised patient data; and (iii) opportunities to improve the public acceptance of genomic data use by standardising a values-based framework that serves the interests of communities and marginalised groups. This typology was used to structure the identified themes of the inductive content analysis of peer-reviewed literature.

Inductive content analysis of literature

The 31 studies identified for analysis were read in full by a member of the research team (FC), who coded these documents using inductive content analysis [27], administered through NVivo software. Several rounds of coding, re-coding and sub-coding were used to develop coding units that reflected themes expressed in the reviewed studies, as outlined in Table 2. The content category coded for was responses to the question ‘What are the opportunities for enacting a national genomics governance framework in the domains of research, clinical medicine, and public (population) health?’ Three other members of the research team (AP, MV, RM) each independently reviewed four of the 31 studies to assess the application of coding data categories and the reliability of inductively identified themes. Disagreements and variations in coding were discussed in a team meeting and several refined coding themes were developed and then applied to all reviewed studies.

Table 2.

Themes and subthemes

Theme	Subtheme	n*	Definition	Example excerpt
Improve Public Acceptance		16	Opportunities for improving the public acceptance of genomic data use.
	Public Acceptance of Data Sharing	12	Opportunities to promote the public acceptance of genomic data sharing.	“Many respondents across the scenarios showed a clear expectation that sharing data would have benefits for knowledge” [28]
	Trust for Data Sharing	8	Opportunities to enhance public, research participant or patient trust for sharing genomic data.	“Those who supported genomic data storage were comfortable with data being reused to benefit the wider community, provided it was collected and stored with proper consent” [22]
	Desire for Common Good	5	Opportunities to promote the use of genomics for the common good.	“Desire to contribute to the global common good and to help those in need” [29]
	Ethical Responsibility	3	Opportunities to promote the use of genomic data through a sense of ethical responsibility.	“Governance and ethical standards for research were also factors that had a significant influence on willingness to donate” [30]
Improve Processes		14	Opportunities for improving genomic data governance processes.
	Standardisation	8	Opportunities to enhance the use of genomic data through standardising formats and data storage.	“Workshop participants suggested that standards need to be developed for curation efforts… to standardize interpretation” [31]
	Infrastructure	5	Opportunities to grow data storage infrastructure for genomic data.	“To promote sharing of knowledge about genes and variants associated with disease, Australian Genomics has deployed two platforms, PanelApp Australia and Shariant” [32]
	Data Access Regulation	3	Opportunities to improve data access regulation for genomic data.	“The access criteria used by Australian biobanks when deciding to share samples or data usually appealed to established or local research governance models.” [33]
	Data Storage and Access	3	Opportunities to improve data storage and accessibility for genomic data.	“We solve the scalability problem by the fragmentation mechanism, providing not only decentralized storage, but also decentralized analysis” [34]
	Data Collection	2	Opportunities to develop data collection procedures for genomic data use.	“A possible solution to this, as one participant suggested, is to “join up with others biobanks in order to work towards epidemiologically valid sample sizes”” [35]
	Align with International Standards	1	Opportunities to alight genomic data governance with international standards.	“GA4GH’s Data Use and Researcher Identity (DURI) Workstream has developed the Data Use Ontology (DUO) technical standard” [36]
	Data Capture Process	1	Opportunities to develop data capture processes for genomic research.	“For those scaling up clinical genomics, the challenge does not necessarily lie with the amount of variation itself but in ensuring that discrepancies are visible” [37]
Define Role of Patients		8	Opportunities for more clearly defining the role of patients (and the public) in data governance.
	Consent for Data Sharing	4	Opportunities to improve consent processes for sharing genomic data.	“The expectation of benefits was the dominant explanation for willingness to share data across all the scenarios” [28]
	Patient Access to Information	4	Opportunities to improve patient information access in genomic research.	“Providing participants with information and support from trusted sources” [36]
	Patient Engagement	3	Opportunities to encourage patient engagement in genomic data use.	“This study demonstrated that participants expect to be asked to share anonymous genomic data at least once. Mechanisms for ‘extended’ or ‘unspecified’ consent… could be included in clinical consent forms to provide an opt-in or opt-out option” [38]
	Patient Autonomy	2	Opportunities to enhance patient autonomy in genomic data use.	“Dynamic Consent platforms are able to facilitate an alternative to broad, ‘all or nothing’ consent, with the opportunity to choose from more granular consent options and revisit consent over time” [36]

*‘n’ denotes the number of studies in which the data category, theme or subtheme appeared. Multiple subthemes may be identified in a single study and hence the total count (‘n’) of subthemes do not add up to the total count of papers in which a theme is identified

Classification and critical appraisal of studies

To assess the opportunities for developing a national genomic data governance framework, we summarised the coding categories related to these themes, presented in the final column of Table 3. This summary of data allows us to (i) identify which themes have been explored and represented in the reviewed studies and (ii) compare these themes to data governance opportunities that have been identified in the grey literature. Consequently, we will use such comparisons to assess any governance opportunities that have been overlooked in peer-reviewed research and identify potential future research areas for the advancement of a national genomic data governance framework in Australia.

Table 3.

Details of reviewed studies, including themes and subthemes

First author	Year	Method(s)	Study aim(s)	Opportunities identified
Best [37]	2023	Interviews	To identify the components of the processes of clinical genomics amenable to adaptation when scaling up clinical genomics from the perspective of nongenetic physicians.	Improve processes Data capture process
Bradford [31]	2019	Workshops	To report on the barriers to implementing genomic healthcare faced by clinical demonstration projects in Queensland.	Define role of patients Consent for data sharing Improve processes Standardisation
Chalmers [39]	2016	Review	To identify the issues associated with biobank practices and governance, with a focus on seven countries, including Australia.	None
Critchley [40]	2015	Interviews	To analyse the impact of the commercialisation of biobanks and of access to biobank resources on public trust.	Improve public acceptance Public acceptance of data sharing Desire for common good Trust for data sharing
Critchley [41]	2017	Surveys	To assess the relative importance that the public place on different expectations of biobanks, such as protecting privacy, healthcare benefits, specific consent, benefit sharing, collaborating and sharing data.	Improve processes Data access regulation Improve public acceptance Public acceptance of data sharing Ethical responsibility Trust for data sharing
Critchley [29]	2020	Surveys and Interviews	To identify the impact of biobank location on willingness of the Australian public to donate tissue for research purposes.	Improve processes Standardisation Improve public acceptance Desire for common good Trust for data sharing
Critchley [42]	2021	Surveys and Interviews	To assess the impact of commercialisation of biobanks and biobank resources on public trust.	Improve public acceptance Public acceptance of data sharing Desire for common good
Daniels [35]	2021	Review	To provide an outline of the use of genomic data alongside routinely collected data in health research and identify challenges, avenues for development and data governance models.	Improve processes Data collection Data storage and access Infrastructure
Dive [30]	2020	Surveys and Interviews	To examine public trust in biobanks and attitudes to the networking, globalisation and commercialisation of biobanks.	Improve processes Data collection Standardisation Improve public acceptance Public acceptance of data sharing Ethical responsibility Trust for data sharing
Eckstein [43]	2018	Review	To review the implications of the proposed new requirements within the NHMRC’s National Statement for Ethical Conduct in Human Research regarding the return of genetic research findings and oversight of transfer agreements.	Improve processes Standardisation Improve public acceptance Trust for data sharing
Garrison [44]	2019	Review	To identify areas of research guidelines that pertain to Indigenous peoples from Canada, New Zealand, Australia and the United States of America that need attention, support Indigenous-led governance and support research policy frameworks that are relevant to Indigenous peoples.	None
Haas [36]	2021	Review	To document the design and development of CTRL (a web-based application for dynamic consent), for use in a health services research project building evidence to inform the integration of genomic medicine into mainstream healthcare.	Define role of patients Consent for data sharing Patient autonomy Patient engagement Patient information Improve processes Align with international processes Infrastructure Standardisation Improve public acceptance Ethical responsibility
Levesque [45]	2021	Surveys	To explore the attitudes of healthy volunteers in phase 1 studies to the topics of genetic security, genetic privacy and incidental genetic findings.	Define role of patients Consent for data sharing Patient information Improve public acceptance Public acceptance of data sharing
Light [33]	2021	Surveys and Interviews	To examine the practices and attitudes of Australian biobanks regarding access to samples and data, as well as local and global networking with other biobanks.	Improve processes Data access regulation Data storage and access Standardisation
Lynch [22]	2023	Focus Groups	To describe the Australian public’s views and preferences for storing and sharing genomic data and implications for choices to donate genomic data.	Improve processes Data access regulation Improve public acceptance Public acceptance of data sharing Trust for data sharing
Malakar [46]	2023	Interviews	To investigate four professional groups’ (clinical geneticists, genetic counsellors, laboratory professionals, and researchers) perceptions of the benefits and risks of using genomics in Australian healthcare.	Improve processes Infrastructure Standardisation
Malakar [47]	2023	Surveys	To explore the views of genomic professionals (clinical geneticists, genetic counsellors, bioinformaticians, and researchers) towards patient data ownership in Australia.	Improve processes Standardisation
Middleton [48]	2020	Surveys	To identify factors shaping willingness of potential donors to consent to donate their genomic data.	Define role of patients Patient information Improve public acceptance Public acceptance of data sharing
Milne [49]	2019	Surveys	To examine trust in data sharing among the general public in the United States of America, Canada, the United Kingdom and Australia.	Define role of patients Patient engagement Improve public acceptance Trust for data sharing
Milne [50]	2021	Surveys	To identify which measures contribute to demonstrating trustworthiness to the public when collecting and sharing genomic data.	Improve public acceptance Public acceptance of data sharing Trust for data sharing
Mohammed Yakubu [51]	2020	Review	To discuss the genome privacy problem and review relevant privacy attacks that have breached the privacy of individuals; to classify privacy-preserving solutions that have been proposed to mitigate these attacks; and to identify the ongoing issues in the field of genomic privacy.	None
Nicol [52]	2016	Multiple Empirical Methods*	To analyse public reactions to the commercialisation of biobanks and their outputs.	Improve public acceptance Public acceptance of data sharing
Prictor [53]	2020	Review	To examine questions of First Peoples’ ownership, ongoing custodianship, potential repatriation, and consent for research use, of heritage materials (such as bones, hair and blood), from legal and normative perspectives.	None
Schlosberg [54]	2016	Review	To describe the relation of privacy principles to data protection and cryptographic techniques with regards to the archival and backup storage of health data in Australia, and to propose a means of implementing the secure management of genomic archives.	None
Stark [32]	2023	Multiple Empirical Methods**	To report on the outcomes of a 5-year national program to accelerate the integration of genomic testing into healthcare in Australia’s decentralised healthcare system.	Improve processes Infrastructure
Trudgett [55]	2022	Review	To identify characteristics of Indigenous Data Sovereignty (IDS) principles and consider a framework for operationalisation of those principles in Australia.	None
Tudini [56]	2022	Multiple Empirical Methods***	To describe the evolution and implementation of Shariant (a controlled-access platform designed to simplify sharing of variant interpretations and associated evidence); and to present initial data demonstrating its benefit for diagnostic accuracy.	None
Vidgen [38]	2020	Surveys	To inform public policy and discussions around genomic data sharing through describing public opinions on using genomic data contained in medical records for research purposes in the Australian state of Queensland.	Define role of patients Patient autonomy Patient engagement Improve public acceptance Public acceptance of data sharing
Warren [28]	2022	Surveys	To investigate factors affecting public attitudes to data sharing through responses to diverse genomic data sharing scenarios.	Define role of patients Consent for data sharing Improve public acceptance Public acceptance of data sharing Desire for common good
Wilson [57]	2022	Surveys	To understand individuals with cerebral palsy and their family’s attitudes and preferences to genomics research, data sharing and biobanking.	Define role of patients Patient information Improve public acceptance Public acceptance of data sharing Desire for common good
Zhang [34]	2019	Experiment	To propose a preventive approach for privacy-preserving sharing of genomic data in decentralised networks for Genome-Wide Association Studies.	Improve processes Data storage and access Infrastructure

* Methods included surveys and a community consultation (deliberative democracy event)

** Methods included surveys, interviews, observational notes, genomic analysis, and program evaluation methods

*** Methods included workshops and program evaluation methods

Results

Key characteristics of studies

The key characteristics of the 31 studies identified for analysis, including year of publication, first author listed, study methods, and a description of each study’s aims, are summarised in Table 3.

There has been an increase in studies related to genomic data governance in Australia over the past decade, with only six studies found from the years 2014 to 2018 and then 25 studies identified in the years 2019 to May 5, 2023.

Opportunities for a national genomic data governance framework

Three categories of opportunities for genomic data governance in Australia were coded within the reviewed literature: opportunities to improve the public acceptance of genomic data use (n = 16), opportunities to improve genomic data processes (n = 14), and opportunities to define the role of patients in data governance (n = 8).

Improving the public acceptance of genomic data use

Four subcategories of opportunities related to improving the public acceptance of genomic data use were: descriptions of the public acceptance of data sharing (n = 12), where trust has been established for data sharing (n = 8), where a desire to use genomic data for the common good are expressed (n = 5), and genomic data sharing being driven by an ethical responsibility (n = 3). Much of the research surrounding the public acceptance of data sharing focuses on comparisons of context, with publicly-funded and not-for-profit uses of data broadly correlating with greater public acceptance of data sharing [38, 40, 42] and broad-based forms of consent (as opposed to narrow, project-specific consent) noted as ‘being the most preferred form of consent’ for such data sharing [30]. Notably, when given the choice between commercial funding of biobank research and no funding, most participants in Nicol et al.’s [52] deliberative democracy event agreed to a resolution that, ‘commercial funding was better than no funding at all’. Collaboration between biobanks was also suggested to be an expectation among the general public in one study [41], while Levesque et al. [45] note that re-identifiability was an important component of the future use of genomic information in the view of clinical trial participants (so much so that they were willing to trade-off privacy in order to possibly benefit from future uses of their information). The expected uses of genomic information play a role in shaping the acceptability of genomic data sharing, with participation in future clinical trials [45], rare disease research [22], participants feeling ‘familiar with genetics’ [48], the transparency of research using genomics [50, 57], and providing health benefits [28] all being associated with the acceptability of sharing genomic data.

The associated data category of trust for data sharing includes descriptions of participants trusting the use of their data by custodians and data governance organisations such as biobanks [40, 41], medical professionals [29] and Human Research Ethics Committees [43]. Milne et al. [49] have noted that, in studies of the Australian public’s views, ‘trust is strongest in individuals’ own doctors and lowest for other actors, particularly for companies and governments’.

Genomic data were also identified across a number of studies as integral to achieving common goods. Studies focusing on public views of data sharing, for example, often identify desires among both (biobank and research) data donors and clinical patients for their genomic data to be used for the common good of others [29, 41, 42] and scientific advancement [28, 57]. A related opportunity for improving the public acceptance of genomic data use lies in leveraging a common public perception that progressing genomic research and medicine are parts of an ethical responsibility towards the general public. For example, Critchley et al. [41] note that, when discussing who should benefit from genetic biobanking, participants in their study identified the ‘general public’, ‘those in need’, and those living in low-income countries as appropriate beneficiaries.

Improving processes for governing genomic data

The most common opportunities identified for improving the processes for governing genomic data identified in reviewed studies concerned standardising data (n = 8), improving genomic (and related) data infrastructure (n = 5), improving data access regulations (n = 3) and improving data storage and access processes (n = 3). Opportunities for the standardisation of genomic data practices differed across research and clinical domains. They included: (i) calls from the clinical setting to standardise genomic data and analysis pipelines [33], including the incorporation of genomic information within ‘Australia’s national digital health’ infrastructure [31, 36]; (ii) training for clinicians and genetic counsellors to standardise interpretation of genomic analyses [31, 46]; and (iii) the research-driven call for reconciliation of ethical standards across national and international settings [29, 33, 43, 47], including more standardised transfer agreements [30]. The related subcategory of improving genomic data infrastructure identifies the associated issues of scalability [36], security [34], and the need for interoperable genomic data interfaces [32], which chiefly related to research genomics. Daniels et al. [35] note that the need for secure storage space for ‘raw genomic data and for specialist platforms required to conduct analysis’ presents an ongoing challenge for genomics broadly. Malakar et al. [46] claim that not having a reference database that is large and representative enough will lead to the generation of avoidable inconclusive results.

Studies of public perspectives of genomic data sharing and biobanks’ uses of data identified opportunities to improve data access regulation. Biobank participants indicated that data access regulations should be enforced by external agencies, such as governments [41] and that the reasons for data being accessed was an important factor in determining the need for regulation [22]. Studies describing opportunities to improve data storage and access processes identified interoperability and the need for researchers to pool rare sample types as essential for achieving research objectives [33, 34]. The less often described opportunity of improving data collection (n = 2) similarly focuses on the potential that sharing interoperable data and its analyses will advance research [30, 35]. Other less frequently mentioned opportunities for improving processes for governing genomic data include one study that identified the GA4GH’s Data Use Ontology technical standard as an opportunity to align data management processes with international standards (n = 1) [36], and another that cautioned that improving data capture processes (n = 1) would need to ‘identify non-negotiable features’ of genomic data– that is, those qualities that should be recorded across genomic tests and data uses– ‘to ensure quality of practice is maintained’ [37].

Defining the role of patients in genomic data governance

Opportunities for clarifying the role of patients in genomic data governance identified in the reviewed literature included improving consent processes for data sharing (n = 4), enhancing patient access to information (n = 4), increasing patient engagement in their genomic data governance (n = 3) and enabling patient autonomy over uses of their genomic data (n = 2). While these four categories are conceptually related, they entail distinct conceptions of the role of the data subject (i.e. patient, participant or donor) in genomic data governance: as subject of pre-defined consent processes, as seekers of information, as participants in data governance, and as having control over uses of their data. The focus of studies describing opportunities for improving data subject consent are directed at enabling choice in the consent process. Australian Genomics has developed a national clinical consent form with the aim of producing consistent data management and sharing across jurisdictions [32], while a dynamic consent platform, CTRL, has also been trialled in research, with a key aim of reducing ambiguity about the meaning of consent clauses, and thereby facilitating data sharing more readily [36]. Other studies have suggested that genomic research consent processes need to be improved to include options for the return of incidental/secondary findings and actionable findings affecting healthcare decisions [28, 45]. The closely related finding of enhancing patient information about genomics is similarly related to both the return of relevant findings [57], and the need to broadly educate the public about genomics and its role in contemporary medicine and public health [48].

Studies identifying opportunities to increase data subjects’ engagement with genomic data governance (namely, how data are used and shared) have explored how ‘choice’ could be modified in research consent processes [36, 38] and have also pointed to the importance of primary healthcare professionals, such as general practitioners, in encouraging patient engagement with secondary data uses (such as research studies) in clinical contexts [49]. The related theme of enabling data subject autonomy is addressed through the same set of studies, which characterise autonomy as enabling choice and access to information during research consent processes [36, 38].

Discussion

This review revealed two key findings that help better understand the opportunities for a national genomic data governance framework:

1. informed consent is consistently conceptualised as the primary mechanism for protecting data subjects and hence legitimising current genomic data governance arrangements; and.

2. opportunities relevant to clinical care and research genomics seem to stem from different priorities, so both will need to be considered alongside one another in any national data governance framework.

Beyond informed consent as the marker of legitimate governance

Across the reviewed studies, individual consent is conceptualised as the primary mechanism for protecting the rights of data subjects. While most Australian states and territories have proposed some combination of coordinated (inter-state) action and digitisation of health records to integrate genomics into health care, the reliance on informed consent as the gold standard of ethical data sharing risks oversimplifying data governance, treating consent processes as proxies for assessing the quality of data governance. In both clinical governance and research ethics, individual consent is the most common standard for guiding custodians and data access assessors as to how, to whom, and when genomic information should be disclosed or reported. While the development of consent processes is informed by discussions of data ethics, the application of consent processes is procedural rather than ethical, and hence such processes should not be used as a proxy for assessing the quality of data governance regimes themselves.

Moving beyond the equating of informed consent with legitimate governance requires a reconceptualisation of the role of ‘trust’ in good governance. Rather than consent to use genomic material or data being ‘informed’, genomic data governance should be ‘trusted’ by data subjects, and worthy of that trust. As Boniolo et al. [58] argue, in relation to biobanks, ‘“trust” concerns a positive attitude of the donor towards the competence, expertise, and moral integrity of both researchers and [bio-]bank governance’. This change in qualifier entails more than a public relations exercise on the part of data custodians, as ‘a trust-based relationship must be monitored to protect patients and prevent abuse’ [58]. A national approach to genomic data governance may therefore need to include a mechanism for fostering the trustworthiness of data management, beyond the well-established norm of informed consent. Barbara Koenig’s [59] model of giving data subjects the option of consenting ‘to be governed’ by a citizen-led Community Advisory Board, rather than being expected to achieve informed consent on every use of their genomic data, is one potential means to preserve the sovereignty of data subjects’ consent while also fostering trust in a governance process that deliberates and acts on their behalf. The challenge here would be establishing the governance of such a Community Advisory Board in a manner which is both independent of genomic data users and representative of the interests of potential data subjects.

The ethical risks of assuming that informed consent is a marker of good governance also rests with the cultural assumptions that shape common informed consent processes. For example, despite genomic data’s inherently shared properties, a methodological individualism characterises discussions of genomic data use and protection; namely, the data subject (or their legal guardian or representative) is often treated as an isolated unit whose rights are protected through the consent process. The potential for benefits or harms to flow on to communities or relatives, which are exacerbated for Australian Indigenous peoples and minority ethnic communities, are secondary considerations in comparison to the individual data subject from whom informed consent is sought. As genomic data are both inherently personal and related to both family and community, more group-oriented forms of governance should be explored as pathways to protecting the rights of data subjects and those with whom they share sensitive genetic information or genome-related social relationships (for examples from the context of Indigenous genomics, see [60, 61]). Such ethical considerations are precluded by commonly used models of consent that are based on the rights of individuals. Accounting for the interests of groups is aligned with the need to explore methods for protecting data subjects and communities beyond the individualised model of consent. For example, Taylor and Whitton [62] have explored the public interest test in the United Kingdom’s Data Protection Act 2018 as a means of advancing public trust in genomic research governance and promoting the social legitimacy of data protection legislation in that context.

Addressing clinical care and research governance

The opportunities for improving processes for governing genomic data identified in this review highlight different priorities described in relation to clinical care and research governance settings. While the identified opportunities for clinical governance largely focused on improving clinical infrastructure and training for practitioners such as counsellors, research opportunities focused on the standardisation and integration of data management with international standards that would allow for more effective data sharing and use. These differing priorities likely reflect the difference in the primary purpose of research data and the use of genomics in clinical care. The national research partnership that commenced in 2016, Australian Genomics, piloted a whole-of-system approach, aiming to deliver a collaborative, federated model for integrating genomics into the Australian healthcare system [63]. Its model addresses both the clinical and research opportunities identified in this study in developing a data governance framework that enabled the creation of ‘integrated national data resources’ such as a genomic data repository, an open source dynamic consent platform, and platforms for disseminating knowledge about genes and variants [32]. This initiative demonstrates the value of addressing both research and translational opportunities as part of a coherent entity.

Realising the opportunities for improving genomic data governance in both clinical and research settings requires that agreed principles for human genomic data sharing be adapted to local settings. This is a key aim of the Australian LINEAGE Study [64]. International guidelines for sharing human genomic data broadly share the principles of respect for individuals, including rights to privacy and self-determination, and collective interests, such as fairness, equity, solidarity and reciprocity [64]. But how those are realised depends on the local settings in which they are operationalised. For example, the European Genomic Data Infrastructure project, a federated approach to sharing genomic and clinical data, was enabled by collaboration between EU member states towards collective healthcare goals [65]. In contrast, the UK Biobank developed its database through a more centralised structure that drew on the altruism of volunteers, leaving some groups in the UK under-represented, such as those from ‘more deprived areas compared with the UK population’ more generally [66]. How clinical and research data governance are reconciled in practice depend on these contextual factors even where common values may prevail. It is essential to consider such contextual factors in the development of a governance framework, a claim also made by the WHO in outline of principles for using, accessing, and sharing human genomic data [67].

In Australia, expanding the use and sharing of genomic data in both clinical and research settings will also require an approach suited to Australian cultural, historical, and regulatory circumstances. As Stark et al. note, reflecting on the Australian Genomics initiative, the underrepresentation of diverse groups, including Aboriginal and Torres Strait Islander Peoples, has led to ‘significant investment’ in engagement strategies to include those groups [32]. Genetic research with Aboriginal and Torres Strait Islander Peoples is particularly marred by a settler-colonial history of excluding those peoples from the control and downstream benefits of their bodies, tissue samples, and genomic data [64, 68]. More broadly, how the interests of a diverse population are recognised in a data governance framework will be important for its legitimacy and acceptance. Health systems are not ‘cultureless’ but rather shaped by dominant cultural norms that privilege certain forms of knowledge and communication, having the potential to impact trust in those systems [69]. Governance of genomic data needs to accommodate the views of underrepresented and diverse groups, in order to articulate a national model that respects and engages with all national communities and avoids the undermining of cultural beliefs, discrimination, or culturally inappropriatepractices.

Limitations

This systematic review is limited by using two databases (Scopus and PubMed) which, although extensive, may not have captured all relevant research papers. Similarly, any work not published in a peer-reviewed journal will have been excluded from this review, although we sought to capture some of this grey literature in the typology-development phase. This review’s geographical limitation to the Australian setting (in order to inform Australian governance) limits the applicability of findings to other jurisdictions to some extent, but the high-level findings are likely of international relevance. Further, identification of opportunities for national genomic data governance frameworks requires a degree of jurisdictional specificity given the variation in legal instruments and institutions governing genomic data in different countries. This review has focused on papers that draw on empirical research and so may have overlooked ethical or legal analysis that did not discuss empirical research.

Conclusion

The development and implementation of a national genomic data governance framework is an ambitious and complex undertaking. Such a project is especially challenging in a field shaped by ever-changing technologies, data management, regulatory rules and procedures, social and political values, and dynamic connections between genomic data stakeholders. Our systematic review has assisted in this process through categorising and assessing the opportunities for developing a national genomic data governance framework that are represented in current research.

In terms of the opportunities for enacting an Australian national data governance framework, we found that jurisdictionally relevant research focuses heavily on individual consent as the primary mechanism for protecting data subjects, and that clinical and research opportunities are not always aligned. The development of trustworthy governance within a national genomic data governance framework should be premised on public discussion of ‘consent’ and what role a national approach to genomic data should play in healthcare and research. Koenig’s representative participatory model of the Community Advisory Board, which aims to enable data subjects to consent to be governed, as opposed to achieving informed consent for every data use, is valuable in placing the focus on the methods by which trust is fostered. Further advancement of a national genomic data governance framework in Australia and abroad will need to account for both individual and communal interests that must be acknowledged to achieve trustworthy data governance.

Author contributions

FC, RM, AP, and MEV jointly conceptualised and conducted the literature review. MO, MR, JH, and AJN contributed to the intellectual development of the data presentation and analysis. Authors have jointly contributed to all sections of the manuscript.

Funding

The “Ethical governance for clinical and genomic data” project, now known as LINEAGE: Law, Sociology and Ethics in Data Governance for Genomics, is funded by the Commonwealth of Australia, Medical Research Future Fund, 2021 Genomics Health Futures Mission Grant, grant MRF2015531.

Data availability

The availability of the full data supporting the findings of this study is subject to restrictions due to the copyright of the included papers. The quotes analysed during this study are included in this published article. Additional data are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.