e-Consent in Biomedical Research Registries: A GDPR-Compliant

Approach Explored in the Context of the Australasian Diabetes Data

Network

Zhe Wang

, Anthony Stell

, Jean Paul Vera Soto

and Richard O. Sinnott

School of Computing and Information Systems, The University of Melbourne, Melbourne, VIC 3010, Australia

Keywords:

e-Consent, Biomedical Registries, GDPR, ADDN, Type-1 Diabetes.

Abstract:

e-Consent - the digital capture of a patient’s consent to be involved in medical research - is a feature of

biomedical research that is becoming increasingly prevalent with the advance of digital technology to support

clinical/biomedical research and targeted registries. Although there have been many reviews of e-consent over

the past decade - evaluating aspects such as informed consent, engagement, comprehension and data security

- there remain unanswered questions about how e-consent ﬁts in the context of recent data legislation and

privacy demands such as the European General Data Protection Regulation (GDPR). This paper outlines key

aspects of e-consent in the context of GDPR and the speciﬁc demands placed on biomedical registries used

for diverse research objectives. We present a practical realisation of GDPR e-Consent in the context of the

Australasian Diabetes Data Network (ADDN) – the national type-1 diabetes registry for Australia.

1 INTRODUCTION

Informed consent is a foundational part of biomedi-

cal research ethics according to the Belmont Report

(1979)

. This introduced three key elements: infor-

mation, comprehension and voluntariness that should

be considered for informed consent to ensure the pro-

tection of human subjects in research. Such principles

have guided the evaluation of different consent pro-

cesses (Sugarman et al., 1998)(Del Carmen and Joffe,

2005). Traditionally, this process was dominated by

paper-based methods, where individuals physically

sign documents after discussions with their health-

care providers. However, with the advent of elec-

tronic medical records and advancements in digital

technology, the electronic format of informed con-

sent (e-Consent) is gaining attention. Review papers

of e-Consent have explored multiple domains includ-

ing general healthcare (Chimonas et al., 2023), sur-

gical procedures (Mirza et al., 2023) and biomed-

ical research (Cohen et al., 2023)(De Sutter et al.,

https://orcid.org/0000-0001-6054-6468

https://orcid.org/0000-0003-4819-9883

https://orcid.org/0000-0001-6345-5596

https://orcid.org/0000-0001-5998-222X

www.hhs.gov/ohrp/regulations-and-policy/belmont-

report/

2020)(Skelton et al., 2020). In healthcare and sur-

gical contexts, e-Consent mainly aims to streamline

administrative undertakings, elevate the standard of

care, and enrich the patient experience (Mirza et al.,

2023).

In the ﬁeld of biomedical research, different cate-

gories of registries are often used to store and process

electronic medical information extensively. There are

various types of registries including biobanks, clini-

cal trial registries, population registries, and targeted

disease registries. Biobanks are repositories that

store biological samples for use in research. These

are often managed according to professional stan-

dards (Hewitt and Watson, 2013). The NSW Health

Statewide Biobank

and Australian Health Biobank

(AHB)

are two such exemplar biobanks in Australia.

Biomedical samples are collected to advance future

research discoveries based on access to physical spec-

imens, e.g. for targeted genomic analysis. Clinical

Trial Registries such as the Australian New Zealand

Clinical Trial Registry (ANZCTR)

keeps track of

clinical trials being undertaken in diverse research ar-

eas. Disease Registries collect targeted data about

https://biobank.health.nsw.gov.au/

https://www.csiro.au/en/work-with-us/industries/healt

h/australian-health-biobank

https://anzctr.org.au/

Wang, Z., Stell, A., Soto, J. and Sinnott, R.

e-Consent in Biomedical Research Registries: A GDPR-Compliant Approach Explored in the Context of the Australasian Diabetes Data Network.

DOI: 10.5220/0012327200003657

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 17th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2024) - Volume 2, pages 45-55

ISBN: 978-989-758-688-0; ISSN: 2184-4305

individuals who have a speciﬁc disease together with

how they are being treated. The Australasian Dia-

betes Data Network (ADDN)

is a disease registry fo-

cused on collecting type-1 diabetes data from across

Australia and New Zealand. Unlike other registries,

ADDN directly re-uses existing health data from hos-

pitals and centres, i.e., as opposed to requiring man-

ual data entry to a separate registry. Such registries

are used to help researchers observe patterns, under-

stand diseases, and improve treatments and disease

management guidelines at scale. Integrating consent

mechanisms into biomedical registries is increasingly

seen as a pivotal step in ethical research (Win and

Fulcher, 2007).

A recent review (Skelton et al., 2020) shed light

on the current e-Consent landscape by presenting a

workﬂow of the informed consent processes (shown

in Figure 1. The right-hand side of the workﬂow illus-

trates how digital informed consent typically works in

biomedical research. It comprises three elements of

informed consent: information, comprehension and

voluntariness which are supported by different digital

tools. However, the workﬂow is abstract and does not

consider the ﬁne-grained and evolving privacy regula-

tions of different countries. For example, in the U.S.,

the HIPAA (Health Insurance Portability and Ac-

countability Act)

was issued by the US Department

of Health and Human Services (HHS) in 1996. It in-

troduced HIPAA Authorization for Research

which

requires researchers and healthcare organisations to

obtain written authorisation from individuals before

using their health information for research purposes.

The European Union’s General GDPR (General Data

Protection Regulation)

, which was put into effect in

2018, offers the most advanced privacy legal frame-

work. It takes the protection of personal data, in-

cluding health-related information, to a much ﬁner-

grained and patient-oriented perspective. The consent

conditions for GDPR require that consent is freely

given, speciﬁc, informed, unambiguous, and easy to

withdraw (also known as the right to be forgotten).

GDPR is more structured and stringent compared to

HIPAA and empowers individuals in how “their data”

might be used or not as the case might be. In Aus-

tralia, the health data sharing landscape has largely

been shaped by the Privacy Act 1988 (The Act)

. This

is currently being amended to align with GDPR and

www.addn.org.au

https://www.hhs.gov/hipaa/index.html

https://privacyruleandresearch.nih.gov/authorization.a

https://eur-lex.europa.eu/legal-content/EN/TXT/?uri

=CELEX%3A32016R0679

https://www.legislation.gov.au/Details/C2014C00076

other international frameworks (Australian Govern-

ment, 2022).

In this context, biomedical research registries

need to design their electronic consent processes to be

aligned with stringent legal requirements while keep-

ing a delicate balance between individual privacy, re-

search safety, and broader public interest (Win and

Fulcher, 2007). Our work aims to augment the e-

Consent workﬂow (Figure 1) delineated by (Skelton

et al., 2020) with concrete examples of what it looks

like to capture and enforce e-Consent in a GDPR-

compliant manner in a speciﬁc biomedical registry:

the Australasian Diabetes Data Network (ADDN) that

is used for diverse research interests.

2 BACKGROUND

2.1 ADDN Background

Australia has one of the world’s highest rates of type-

1 diabetes (T1D). By September 2022, over 134,000

individuals with T1D were registered with the Na-

tional Diabetes Service Scheme (NDSS)

in Aus-

tralia. To better understand and manage this health

challenge, the Australasian Diabetes Data Network

(ADDN - www.addn.org.au) was launched. It was

funded by the Juvenile Diabetes Research Foundation

(JDRF – www.jdrf.org.au) in 2012. The ADDN reg-

istry consolidates longitudinal data from 59 diabetes

centres from across Australia and New Zealand, cap-

turing extensive records from more than 20,000 pa-

tients. This includes over 250,000 hospital visits.

ADDN’s primary mission is to collate T1D health

data from various centres into a single platform.

This uniﬁed database helps monitor long-term pa-

tient outcomes, advance T1D research, and enhance

clinical care across Australasia. The University of

Melbourne’s Melbourne eResearch Group (MeG–

www.eresearch.unimelb.edu.au) maintains and sup-

ports the ADDN registry.

Figure 2 shows an example of a subset of the pa-

tient data based on the ADDN schema. As noted,

ADDN re-uses existing health data from hospital sys-

tems. The health data is de-identiﬁed at source before

it is populated into the ADDN registry, a Unique Sub-

ject Identiﬁer (USI) is generated using the BioGrid

data linkage platform (www.biogrid.org.au). This

identiﬁer replaces personal identiﬁers whilst ensur-

ing that data remains traceable for clinical or research

https://www.ndss.com.au/about-the-ndss/diabetes-fac

ts-and-figures/

HEALTHINF 2024 - 17th International Conference on Health Informatics

Figure 1: Workﬂow of informed consent processes from (Skelton et al., 2020).

Figure 2: Example of patient data based on the ADDN

schema.

Figure 3: Example of a site-speciﬁc benchmarking report.

Figure 4: ADDN consent process diagram.

purposes without compromising the privacy of the in-

dividuals.

Figure 4 shows the basic process of the current

ADDN consent mechanism. Rather than seeking ex-

plicit permission, e.g., signing a consent form or tick-

ing a consent box from each individual (opt-in), pa-

tients are automatically included in the registry unless

they actively choose to opt-out. The opt-in/opt-out

consent status is captured by clinicians and/or nurses

who provide them with healthcare during hospital vis-

its. As part of this process, they are presented with all

of the details of the ADDN project and explanations

of the consent process. This consent capture has his-

torically been realised by a signed letter. If a partici-

pant chooses not to opt-out after being provided with

this information, the date on which an individual de-

cides not to opt-out (and thereby passively provides

consent) is documented. This is shown as ”dateO-

fAddnConsent” in Figure 2.

After recording the ”dateOfAddnConsent”, pa-

tient data is transmitted to the ADDN registry by cen-

tres. This occurs twice per year. This action indi-

cates “by default” that the data is authorised for use

in subsequent research studies. Such downstream use

of the data is unbeknownst to the patients. This is

not aligned with GDPR, however. As part of ADDN,

centres receive a site-speciﬁc benchmarking report as

shown in Figure 4 illustrating how their centre com-

pares to other centres across a range of metrics, e.g.,

average HbA1c for patients at their centre compared

to other centres for example).

Centres and researchers more generally also have

the opportunity to propose studies using the ag-

gregated T1D data from ADDN. Projects undergo

an ADDN-speciﬁc approval process decided by an

ADDN Study Group

. However, patients are not in-

formed about these projects, hence there is a lack of

patient awareness regarding research initiatives using

their data.

2.2 Background to Consent

Based on a systematic review conducted by (de Man

et al., 2023), it was found that opt-out procedures tend

to yield higher consent rates and result in more repre-

sentative participant samples when compared to opt-

in procedures. This is because opt-out procedures do

not require individuals to take proactive steps to pro-

vide consent. However, it may raise ethical concerns

about personal autonomy when participants are un-

https://www.addn.org.au/governance

e-Consent in Biomedical Research Registries: A GDPR-Compliant Approach Explored in the Context of the Australasian Diabetes Data

Network

aware of their participation and the subsequent, down-

stream use of their data. (Williams et al., 2015) dis-

cuss the legality of opt-out consent, which can de-

crease potential participation due to unwarranted mis-

conceptions and associated risks thereby impacting

the validity of research outcomes.

Additionally (de Man et al., 2023) identiﬁed that

opt-in studies utilizing broad consent tend to achieve

higher consent rates compared to study-speciﬁc con-

sent projects. Study-speciﬁc consent has been criti-

cised for potentially causing consent fatigue (Dankar

et al., 2020)(Ploug and Holm, 2015)(Holm and Ploug,

2017), since participants receive high volumes of con-

sent requests so that the choice will become routinised

and/or cause them to refuse or withdraw consent due

to the volume of requests (Dankar et al., 2020)(Holm

and Ploug, 2017).

However, it is important to note that while broad

consent offers advantages in terms of a lower admin-

istrative burden, (Haas et al., 2021) identiﬁed that

broad consent models raise privacy issues regarding

personal data, potentially resulting in reduced partici-

pation rates in studies. This reduction in participation

rates could also lead to participants feeling as though

they have less control over their data, ultimately giv-

ing rise to trust and ethical concerns (Mamo et al.,

2020).

The one-off opt-out consent model provided to

ADDN patients remains in place unless patients

proactively contact the ADDN project manager to

make changes to their registry status. Patients have

the option to leave the information collected so far

but not permit further collection (partial opt-out) or

request deletion of all information collected (full opt-

out). This approach lacks ongoing patient engage-

ment and involvement in dynamic decision-making.

When centres send their data to the ADDN registry,

patients are not informed about it, nor do they have ac-

cess to the data being collected on them, or the oppor-

tunity to express their preferences regarding the data

use in speciﬁc downstream studies.

This lack of ongoing interaction between patients

and the broader research community and the use of

one-off static consent overlooks the dynamic nature

of biomedical research for several reasons. For ex-

ample, the notion of “personal data” is not static

and data can be easily repurposed (Ausloos, 2012).

Some biomedical research registries such as longitu-

dinal studies require the ongoing collection of bio-

logical samples and health-related records (Lee and

Lee, 2022). The future usage of data is often not

static following initial data collection (Kaye et al.,

2015)(Mamo et al., 2020), particularly in rapidly

evolving ﬁelds like biotechnology. As a result, par-

ticipant preferences may evolve with changing values

and aspirations (Mascalzoni et al., 2022) and legality

across different jurisdictions in cross-border research

may evolve.

The existing opt-out, broad, one-off consent

framework for ADDN gives rise to many issues, es-

pecially with regard to frameworks such as GDPR.

When assessed against the three pillars of the Bel-

mont Report

— information, comprehension, and

voluntariness—the current mechanism is predomi-

nantly aligned with the left-hand side of the work-

ﬂow depicted in Figure 1., as shown in the consent

process diagram (Figure 4), patients receive informa-

tion through a physical patient letter, comprehension

is facilitated by discussions between the clinician and

patients during regular visits, and the exercise of vol-

untariness is limited to opt-out options.

These challenges are further compounded by re-

search demands such as linking ADDN data with

external datasets such as the Australian Institute of

Health and Welfare’s

national death index, the Aus-

tralian medical beneﬁts scheme

, and the Australian

pharmaceutical beneﬁts scheme

. To facilitate this,

introducing more personal identiﬁcation into ADDN

is essential to avoid exacerbating technical, ethical,

and legal issues.

Given the shift to more rigorous privacy regu-

lations in Australia, it is imperative to design the

consent workﬂow with GDPR compliance in mind,

to offer participants better control over downstream

use of their personal data. The existing ADDN

consent model, as represented on the left-hand side

of the workﬂow in Figure 1, is incompatible with

GDPR standards. In the subsequent sections, we

delve deeper into the speciﬁc conditions of GDPR

consent and highlight gaps in the current framework

and how to augment the right-hand side of the work-

ﬂow (e-consent) to better capture consent in a GDPR-

compliant manner for biomedical research registries

such as ADDN.

2.3 The GDPR Context

Article 4.7 and 4.8 of GDPR provides deﬁnitions for

two crucial roles within the data processing ecosys-

tem. The ’Data Controller’ is the entity responsible

for determining the ’why’ and ’how’ of processing

personal data. In the case of a biomedical research

registry, for example, the institution overseeing the

www.hhs.gov/ohrp/regulations-and-policy/belmont-r

eport/

https://www.aihw.gov.au/

www.mbsonline.gov.au

www.pbs.gov.au

HEALTHINF 2024 - 17th International Conference on Health Informatics

registry’s operations would be considered the Data

Controller, as it decides the purposes and methods for

collecting and storing biological data.

The ’Data Processor’ is the entity performing ac-

tual actions on behalf of the Data Controller. In a

biomedical research registry scenario, an IT or data

management company contracted by the project to

process associated data would typically assume the

role of Data Processor.

GDPR broadly deﬁnes a ’data subject’ as any liv-

ing individual whose personal data is collected, held,

or processed by a particular organisation.

In the context of ADDN, the roles of data con-

troller and data processor are delineated among var-

ious collaborating entities. The direction and pur-

pose of ADDN operations are primarily shaped by

the ADDN governance team. This team comprises

independent external investigators, ADDN investiga-

tors, representatives from JDRF

, and a group of pa-

tient advisors. Meanwhile, the actual processing of

patient data — transferring it from hospital systems

and ensuring its alignment with ADDN’s governance

policies — is managed by the software developers of

the Melbourne eResearch Group (MeG). Within the

GDPR’s framework, every living patient registered

within ADDN falls under the deﬁnition of a ’data sub-

ject’.

According to Article 7 of GDPR, consent only be-

comes legally valid when it satisﬁes the ﬁve condi-

tions listed below:

“(1) Freely Given - the data subjects must not be

cornered into agreeing, noting that the imbalance be-

tween the data subject and controller can often mak-

ing unencumbered consent difﬁcult, e.g., patients may

feel obliged or have concerns that the treatments they

receive may be inferior if they do not agree. Further-

more, each use of personal data should be given sep-

arate consent.

(2) Speciﬁc – the consent must be collected for

certain agreed activities or purposes unless explicitly

identiﬁed as “general” research.

(3) Informed - the data subject must fully under-

stand the implications of consent before making a de-

cision. This includes an understanding of data pro-

cessing activities, their purpose and any associated

risks or consequences.

(4) Unambiguous – it should be immediately

clear whether a data subject has consented. Consent

under GDPR cannot be implied or assumed, rather ex-

plicit opt-in consent is required.

(5) Withdrawal – individuals can withdraw their

consent at any time, and this withdrawal should be

made as easy as obtaining the original consent. This

www.jdrf.org.au

should result in the removal/deletion of their data

from the registry.”

Based on the GDPR’s ﬁve consent conditions, an

ideal consent process in the realm of biomedical re-

search should evolve into a patient-centric, contin-

uous, opt-in, and dynamic engagement mechanism.

This not only maintains GDPR compliance but also

grants individual’s autonomy over use of their data.

2.4 Australian Privacy Law

In Australia, the health data sharing landscape has

predominantly been shaped by the Privacy Act of

1988 (The Act)

. Recently in September 2023, the

Australian Government unveiled its response

to the

Privacy Act Review Report (Feb 2023)

. This is

the culmination of two years of extensive consulta-

tion and review resulting in the release of Issues Paper

(Oct 2020)

and Discussion Paper (Oct 2021)

In this response, the government’s stance on pri-

vacy has been clariﬁed, showing alignment with nu-

merous proposals designed to enhance privacy safe-

guards for individuals.

One of the key points has been the area of con-

sent. Aiming to alleviate burdens on individuals and

avoid consent fatigue, the government has given in-

principle agreement to Proposal 11.1 which seeks to

reﬁne the deﬁnition of consent, emphasising that it

must be voluntary, informed, current, speciﬁc, and

unambiguous. Moreover, with Principle 11.3, indi-

viduals are given the clear empowerment to withdraw

their consent, insisting that the withdrawal process

should be as straightforward as giving the consent

process.

These developments in Australian privacy stan-

dards align with the consent conditions of the GDPR,

underscoring their global relevance and prominence.

Given these evolving legal landscapes, it is paramount

for Australasia-based research initiatives such as

ADDN, to improve their consent mechanisms. By

aligning with these contemporary standards, not only

will they be adhering to domestic regulations but

this will also ensure compatibility with international

norms, driven by GDPR.

To realise this vision, a digital platform — poten-

www.legislation.gov.au/Details/C2014C00076

www.ag.gov.au/rights-and-protections/publications/

government-response-privacy-act-review-report

www.ag.gov.au/integrity/consultations/review-priva

cy-act-1988

www.ag.gov.au/rights-and-protections/publications/re

view-privacy-act-1988-cth-issues-paper

https://consultations.ag.gov.au/rights-and-protections

/privacyact-review-discussion-paper/

e-Consent in Biomedical Research Registries: A GDPR-Compliant Approach Explored in the Context of the Australasian Diabetes Data

Network

Figure 5: Augmented e-Consent workﬂow.

Figure 6: High-level architecture of ADDN Consent app.

tially in the form of a web or mobile app — is essen-

tial. This app should not only provide patients with an

interactive interface but also grant them direct access

to their data. They could then comprehend their in-

formation, and based on this, decide if they consent to

its use in speciﬁc research projects on an ongoing ba-

sis. The traditional consent process represented in the

left-hand side of (Skelton et al., 2020)’s workﬂow for

e-consent in Figure 1 would never support such dy-

namic and evolving systems. Furthermore, different

from the right-hand side of Figure 1, this model goes

beyond just digital signatures or list of digital media.

Rather it is about designing a continuous, informed,

and interactive relationship between the patient and

the ongoing use of their data.

3 E-CONSENT WORKFLOW

As discussed above, we have reﬁned the workﬂow

presented by (Skelton et al., 2020) as shown in Fig-

ure 1. The augmented tree diagram in Figure 5 un-

derscores the foundational pillars of GDPR consent,

essential for an e-Consent platform. Extended from

the “information, comprehension and voluntariness”

requirements, each branch of the tree represents one

of the main GDPR consent conditions and the critical

requirements and features that e-Consent platforms

should integrate to ensure GDPR compliance. This il-

lustration offers practical implementation suggestions

suitable for real-world applications. Guided by our

augmented workﬂow, the subsequent sections detail

the speciﬁc requirements of the e-Consent platform

in the ADDN context.

4 ADDN IMPLEMENTATION

4.1 Overview

To improve the current consent process as described

in section 2, we introduce the ADDN eConsent mo-

bile app, as detailed in (Wang et al., 2022)(Wang

et al., ). Figure 6 shows the high-level architecture

of the application. At the heart of the backend sys-

tem lies the Authorisation Service JWT (JSON Web

Tokens). This service controls the authentication pro-

cesses, ensuring that only authorized users and ser-

vices can access the required data.

Building on this foundation, the platform sup-

ports a Python-based API Service. This service

serves as a conduit between the Authorisation Ser-

vice and the underlying data store, which is realised

as a NoSQL MongoDB. The MongoDB database pro-

vides the scalability and ﬂexibility needed to store

vast amounts of data pertaining to users, consent data,

HEALTHINF 2024 - 17th International Conference on Health Informatics

Figure 7: Screenshot of ADDN Consent Mobile Application.

Figure 8: (Left) Screenshot of web application: create a research study (right) Screenshots of web application: Consent

statistics (a), create a user (b) and upload user’s data.

and details of research studies. The Python API Ser-

vice handles the communication between the database

and the client-side applications. In addition, its ar-

chitecture allows connections with external databases,

which can be a valuable feature in future iterations.

For the front end, there are two clients: the Re-

act Web Application (Figure 8) is primarily designed

for the ADDN administrative team clinicians, and the

React Native Mobile Application is tailored for the

end-users and is where the actual consent processes

take place.

During a patient’s regular visit, the clinician will

explain the ADDN project and the usage of the

ADDN consent app. Once the patient agrees to use

the app, the clinician will prepare it for the patient, as

shown in Figure 8(b), with an activation code being

generated for security log-in.

When an ADDN research study is approved, the

administrator can dispatch consent tasks to target

groups as shown in Figure 8 (left) and a notiﬁcation

indicating a new consent task will be sent to the tar-

get users via their mobile application. They can also

monitor the progress of different consents across var-

ious research studies Figure 8 (right(a)). The inter-

face ensures that admins can keep track of all ongo-

ing activities and make decisions swiftly. At the same

time, patients are fully aware of the projects that are

requesting access to and use of their data and can ac-

cept or refuse such requests on an ongoing basis.

4.2 The Consent App

In this section, we use a table (shown in the Ap-

pendix) to outline the speciﬁc consent requirements

associated with the ﬁve GDPR conditions based on

the augmented e-Consent workﬂow presented in Sec-

tion 3. These requirements inform the design of a

compliant and functional e-Consent system.

The e-Consent app offers the ”My Records” dash-

board, where patients can access and review their clin-

ical data. As illustrated in Figure 9, patients can re-

trieve their visit and medication data directly on their

devices. This provides them with direct access to the

data captured about them that exists within the ADDN

registry. This not only reinforces the principle of data

accessibility but also serves as an added incentive for

patients to engage with the app. For example, they

can track key health indicators such as their HBA1c

levels over time or their BMI.

It is noted that to comply with the freely given

e-Consent in Biomedical Research Registries: A GDPR-Compliant Approach Explored in the Context of the Australasian Diabetes Data

Network

Figure 9: Augmented e-Consent workﬂow.

condition of GDPR consent, - for those patients who

may opt against using the app - they can still re-

quest their data via traditional means through their

clinicians. Furthermore, for those who use the app

but decide not to provide speciﬁc consents to spe-

ciﬁc research project requests to use their data, the

”My Records” dashboard remains accessible, ensur-

ing they aren’t disadvantaged or penalised for their

choices.

The app itself requires a unique token to be gener-

ated on the server before it can be used. This is used

for several purposes: to activate the mobile applica-

tion; to identify the end user mobile application (and

hence the anonymised patient) so that they can access

and see their data and get notiﬁcations of studies re-

lated to the use of their data, and to encrypt the data

sent between the mobile application and the server.

It is important to note that all of the data within the

ADDN data registry has been anonymised at source.

The mobile application also has no uniquely identify-

ing data that is kept. The web and mobile applications

have been developed based on privacy by design prin-

ciples. There is no need to know the speciﬁc individ-

ual details. Instead, all patients are associated with a

unique and system-generated identiﬁer on the server.

5 DISCUSSION

Our augmented workﬂow of e-Consent and the con-

crete example in the ADDN context offers practical

guidance for other biomedical research registries aim-

ing to comply with strict health data access regula-

tions. However, some limitations must be acknowl-

edged. There is no standard implementation for the

consent application due to the multifaceted nature of

biomedical research demands.

Data controllers still need to consider the speciﬁc

requirements of their research registry when deter-

mining the most appropriate consent process. Two

key factors to take into account are altruism and the

stage of the research. For instance, the disease reg-

istry RUDY project (Rare UK Diseases of bone, joints

and blood vessels) (Teare et al., 2017) project has

demonstrated that patients with extensive experience

of their disease can become active partners in re-

search. (Garrison et al., 2016)(Spencer and Patel,

2019) also found that altruistic beneﬁts of sharing

health-related data sometimes outweighs the associ-

ated risks, leading participants to prefer broader con-

sent or even full access (Wallace and Miola, 2021) in

their efforts to contribute to society. The maturity of

a given research registry and the level of trust estab-

lished with its participants should also inﬂuence the

introduction of new consent mechanisms. As high-

lighted by (Wallace and Miola, 2021) for mature reg-

istries where participants have developed trust in the

project, introducing a new consent system may intro-

duce risks and potentially jeopardize the relationship

with participants instead of providing beneﬁts. There-

fore, the timing and manner of rolling out novel con-

sent procedures need to be considered. Moreover, not

all research registries might have the technological in-

frastructure or expertise to deploy such advanced e-

Consent platforms. Indeed, this is one of the key fac-

tors that controllers should take into account.

Furthermore, data protection regulations are con-

tinuously evolving. In the Australian context, while

our work is based on the latest government response

, the Privacy Act 1988 is still under review. The ﬁnal

revised version has yet to be released. As such, there

could be further discrepancies between the ﬁnalized

version of this Act and GDPR with regards to consent

conditions. Such disparities may necessitate further

adjustments in the future.

6 CONCLUSIONS

In the evolving landscape of biomedical research,

the move towards electronic informed consent (e-

HEALTHINF 2024 - 17th International Conference on Health Informatics

Consent) is inevitable. The up-to-date review of e-

consent by (Skelton et al., 2020) presented a work-

ﬂow of the informed consent processes, but this was

more theoretical than practical and did not consider

evolving privacy mandates and the longitudinal na-

ture of research data and evolving research demands.

In this paper we explore the complexities of aligning

e-Consent mechanisms with stringent global data pro-

tection regulations, notably GDPR.

Through the exploration of the biomedical re-

search registry Australasian Diabetes Data Network’s

(ADDN)’s consent app, we illuminate practical im-

plementations that adhere to such regulations while

enhancing the patient experience.

We also presented an improved GDPR-compliant

consent workﬂow in biomedical research settings.

This provides guidance to other biomedical research

registries attempting to navigate the complexities of

GDPR-compliant e-consent implementations.

We note that the mobile application is undergoing

advanced testing and will be rolled out as part of the

ADDN project in due course. The adoption, use and

feedback of the application will be explored in down-

stream work.

ACKNOWLEDGEMENTS

The authors would like to thank the ADDN project

partners for the ongoing work and support.

REFERENCES

Ausloos, J. (2012). The ’right to be forgotten’ - worth re-

membering? Computer Law and Security Review,

28(2):143–152.

Australian Government (2022). Privacy act review - discus-

sion paper. Webpage. Accessed on 2023-11-29.

Chimonas, S., Lipitz-Snyderman, A., Matsoukas, K., and

Kuperman, G. (2023). Electronic consent in clinical

care: an international scoping review. BMJ Health &

Care Informatics, 30(1).

Cohen, E., Byrom, B., Becher, A., J

ornt

en-Karlsson, M.,

and Mackenzie, A. K. (2023). Comparative effective-

ness of econsent: Systematic review. Journal of med-

ical Internet research, 25.

Dankar, F. K., Gergely, M., Malin, B., Badji, R., Dankar,

S. K., and Shuaib, K. (2020). Dynamic-informed con-

sent: A potential solution for ethical dilemmas in pop-

ulation sequencing initiatives. Comput Struct Biotech-

nol J, 18:913–921.

de Man, Y., Wieland-Jorna, Y., Torensma, B., de Wit, K.,

Francke, A. L., Oosterveld-Vlug, M. G., and Verheij,

R. A. (2023). Opt-in and opt-out consent procedures

for the reuse of routinely recorded health data in scien-

tiﬁc research and their consequences for consent rate

and consent bias: Systematic review. Journal of med-

ical Internet research, 25.

De Sutter, E., Zac¸e, D., Boccia, S., Di Pietro, M. L., Geerts,

D., Borry, P., and Huys, I. (2020). Implementa-

tion of electronic informed consent in biomedical re-

search and stakeholders’ perspectives: systematic re-

view. Journal of medical Internet research, 22(10).

Del Carmen, M. G. and Joffe, S. (2005). Informed consent

for medical treatment and research: a review. The on-

cologist, 10(8):636–641.

Garrison, N. A., Sathe, N. A., Antommaria, A. H. M.,

Holm, I. A., Sanderson, S. C., Smith, M. E.,

McPheeters, M. L., and Clayton, E. W. (2016). A sys-

tematic literature review of individuals’ perspectives

on broad consent and data sharing in the united states.

Genetics in Medicine, 18(7):663–671.

Haas, M. A., Teare, H., Prictor, M., Ceregra, G., Vidgen,

M. E., Bunker, D., Kaye, J., and Boughtwood, T.

(2021). ’ctrl’: an online, dynamic consent and partic-

ipant engagement platform working towards solving

the complexities of consent in genomic research. Eur

J Hum Genet, 29(4):687–698.

Hewitt, R. and Watson, P. (2013). Deﬁning biobank. Biop-

reservation and biobanking, 11(5):309–315.

Holm, S. and Ploug, T. (2017). Big data and health re-

search—the governance challenges in a mixed data

economy. Journal of bioethical inquiry, 14:515–525.

Kaye, J., Whitley, E. A., Lund, D., Morrison, M., Teare, H.,

and Melham, K. (2015). Dynamic consent: a patient

interface for twenty-ﬁrst-century research networks.

Eur J Hum Genet, 23(2):141–146.

Lee, H. and Lee, U. (2022). Toward dynamic consent

for privacy-aware pervasive health and well-being: A

scoping review and research directions. IEEE Perva-

sive Computing.

Mamo, N., Martin, G. M., Desira, M., Ellul, B., and Ebe-

jer, J. P. (2020). Dwarna: a blockchain solution for

dynamic consent in biobanking. Eur J Hum Genet,

28(5):609–626.

Mascalzoni, D., Melotti, R., Pattaro, C., Pramstaller, P. P.,

ogele, M., De Grandi, A., and Biasiotto, R. (2022).

Ten years of dynamic consent in the chris study: in-

formed consent as a dynamic process. Eur J Hum

Genet, 30(12):1391–1397.

Mirza, A. B., Khoja, A. K., Ali, F., El-Sheikh, M., Bibi-

Shahid, A., Trindade, J., Rocos, B., Grahovac, G.,

Bull, J., and Montgomery, A. (2023). The use of e-

consent in surgery and application to neurosurgery:

a systematic review and meta-analysis. Acta Neu-

rochirurgica, pages 1–32.

Ploug, T. and Holm, S. (2015). Meta consent: a ﬂexible

and autonomous way of obtaining informed consent

for secondary research. BMJ, 350:h2146.

Skelton, E., Drey, N., Rutherford, M., Ayers, S., and Mala-

mateniou, C. (2020). Electronic consenting for con-

ducting research remotely: A review of current prac-

tice and key recommendations for using econsent-

e-Consent in Biomedical Research Registries: A GDPR-Compliant Approach Explored in the Context of the Australasian Diabetes Data

Network

ing. International Journal of Medical Informatics,

143:104271–104271.

Spencer, A. and Patel, S. (2019). Applying the data pro-

tection act 2018 and general data protection regula-

tion principles in healthcare settings. Nursing Man-

agement, 26(1).

Sugarman, J., McCrory, D. C., and Hubal, R. C. (1998).

Getting meaningful informed consent from older

adults: a structured literature review of empirical re-

search. Journal of the American Geriatrics Society,

46(4):517–524.

Teare, H. J. A., Hogg, J., Kaye, J., Luqmani, R., Rush, E.,

Turner, A., Watts, L., Williams, M., and Javaid, M. K.

(2017). The rudy study: using digital technologies

to enable a research partnership. Eur J Hum Genet,

25(7):816–822.

Wallace, S. E. and Miola, J. (2021). Adding dynamic con-

sent to a longitudinal cohort study: A qualitative study

of exceed participant perspectives. BMC Med Ethics,

22(1):12.

Wang, Z., Stell, A., Sinnott, R. O., and Group, A. S. (2022).

The impact of general data protection regulation on

the australasian type-1 diabetes platform. pages 21–

23 July 2022.

Wang, Z., Stell, A., Sinnott, R. O., and The Addn Study, G.

A gdpr-compliant dynamic consent mobile applica-

tion for the australasian type-1 diabetes data network.

Healthcare (Basel), 11(4).

Williams, H., Spencer, K., Sanders, C., Lund, D., Whitley,

E. A., Kaye, J., and Dixon, W. G. (2015). Dynamic

consent: A possible solution to improve patient con-

ﬁdence and trust in how electronic patient records are

used in medical research. JMIR Med Inform, 3(1):e3.

Win, K. T. and Fulcher, J. A. (2007). Consent mechanisms

for electronic health record systems: a simple yet un-

resolved issue. Journal of Medical Systems, 31:91–96.

HEALTHINF 2024 - 17th International Conference on Health Informatics

APPENDIX

Table 1: Speciﬁc consent requirements associated with the ﬁve GDPR conditions based on the augmented e-Consent workﬂow.

Consent Condition Requirement App Interface Example

Freely Given Consent should be given

without any pressure, en-

suring no imbalance be-

tween the data subject and

controller.

The ﬁrst screenshot (a) of

the ADDN app shows a

dashboard with a list of

studies. When new stud-

ies are approved by the

ADDN study group, they

are displayed here for tar-

geted participants. Each

study’s entry acts as an in-

vitation, not a command.

The BMI study appears on

the dashboard without any

highlighting or prioritiza-

tion, ensuring users don’t

feel compelled to partici-

pate.

Speciﬁc Consent must be collected

for distinct, predeﬁned

purposes.

By selecting a study from

the dashboard, like the

BMI study, users are taken

to (b), which offers de-

tailed information about

the speciﬁc study, ensuring

the user knows precisely

what they are consenting

to.

The BMI study informa-

tion clearly outlines the

speciﬁc goals and pur-

poses of the study, and

details how the speciﬁc

data will be collected and

utilised.

Informed Data subjects should fully

understand the data pro-

cessing activities and any

associated implications.

(b) offers study details and

provides a direct link to the

(c), allowing users to view

the exact data records that

will be used for the study.

For the BMI study, users

can view their BMI data,

ensuring they’re com-

pletely aware of what

information is being used.

Unambiguous & Opt-in It should be crystal clear

whether a user has given

their consent. GDPR de-

mands an explicit opt-in

system.

Within the detailed study

page (b), users have the ex-

plicit choice to ’Consent’

or ’Withdraw’. Only an

active action (like pressing

’Consent’) will register as

the user giving their per-

mission.

If a user decides to con-

sent to the BMI study,

the user has to press the

’Consent’ button; other-

wise, the consent is not

given, and ADDN can’t

use the data.

Ease of Withdrawal Withdrawing consent

should be as simple as giv-

ing it. Upon withdrawal,

their data should be re-

moved from the study.

The ’Withdraw’ option on

the detailed study page (a)

ensures that users can pull

back their consent at any

time, and it will be as easy

as giving the consent.

If a user initially agrees to

the BMI study but later de-

cides against it, they can

simply press ’Withdraw’,

and their BMI data will not

be used, but they can still

consent to other studies.

Right to be Forgotten Users should have the

power to request that all

their data be deleted from

the ADDN platform per-

manently, reﬂecting the

GDPR’s ”Right to be For-

gotten.”

A ’Delete My Data’ but-

ton (d) lets users remove

all their data from ADDN,

withdrawing from all stud-

ies and permanently eras-

ing their presence on the

platform.

If a user decides to exit the

ADDN platform entirely,

they can press ’Delete

My Data’, wiping out all

their records and simulta-

neously revoking consent

for all studies they had pre-

viously agreed to.

e-Consent in Biomedical Research Registries: A GDPR-Compliant Approach Explored in the Context of the Australasian Diabetes Data

Network