Health Policy

Decentralised clinical trials: ethical opportunities and

challenges
Effy Vayena*, Alessandro Blasimme*, Jeremy Sugarman

Fuelled by adaptations to clinical trial implementation during the COVID-19 pandemic, decentralised clinical trials Lancet Digit Health 2023;
are burgeoning. Decentralised clinical trials involve many digital tools to facilitate research without physical contact 5: e390–94

between research teams and participants at various stages, such as recruitment, enrolment, informed consent, Published Online
April 25, 2023
administering study interventions, obtaining patient-reported outcome measures, and safety monitoring. These tools
https://doi.org/10.1016/
can provide ways of ensuring participants’ safety and research integrity, while sometimes reducing participant burden S2589-7500(23)00052-3
and trial cost. Research sponsors and investigators are interested in expanding the use of decentralised clinical trials. *Contributed equally
The US Food and Drug Administration and other regulators worldwide have issued guidance on how to implement Health Ethics and Policy Lab,
such adaptations. However, there has been little focus on the distinct ethical challenges these trials pose. In this Department of Health Sciences
Health Policy report, which is informed by both traditional research ethics and digital ethics frameworks, we group and Technology, Federal
Institute of Technology–ETH
the related ethical issues under three areas requiring increased ethical vigilance: participants’ safety and rights,
Zurich, Zürich, Switzerland
scientific validity, and ethics oversight. Our aim is to describe these issues, offer practical means of addressing them, (E Vayena PhD,
and prompt the delineation of ethical standards for decentralised trials. A Blasimme PhD); Berman
Institute of Bioethics,
Background However, the type and nature of such methods is rapidly Department of Medicine, and
Department of Health Policy
The COVID-19 pandemic and the public health measures expanding, as digital tools (eg, electronic consent, virtual and Management, Johns
implemented to contain it considerably impacted clinical consultations, the use of digital reporting platforms, Hopkins University, Baltimore,
research.1,2 Ongoing and new clinical trials were adapted by digitally acquired endpoints, apps, and wearable MD, USA (J Sugarman MD)
decentralising essential elements, such as enrolment, technologies) prove increasingly able to replace elements Correspondence to:
administration of study interventions, and safety of traditional clinical trials.12 Prof Effy Vayena, Health Ethics
and Policy Lab, Department of
monitoring. This change involved a wide range of The growing interest in DCTs by sponsors and regulators Health Sciences and Technology,
approaches, including the use of remote methods for is justified for several reasons. DCTs can promote greater Federal Institute of Technology–
consent and monitoring, and sending medications, inclusivity, diversity, and equitable access to research ETH Zurich, Zürich 8092,
Switzerland
equipment, or study staff to partici­pants’ homes to enable participation by removing the obstacles asso­ciated with the
[email protected]
study continuation. need for proximity to clinical sites or mobility.13 Beyond
Several regulatory agencies consequently issued facilitating access and partici­ pation, DCTs can ease
guidance on conducting clinical trials remotely during a recruitment, decrease delays, enhance participant
pandemic. The US Food and Drug Administration (FDA) retention, and be less costly.12 Never­ theless, DCTs also
emphasised the importance of evaluating risks and involve distinct ethical challenges. In this Health Policy
benefits of continuing or discontinuing a clinical trial, report, we identify and analyse such challenges and offer
noting that in some circumstances patient safety might be practical recommen­dations for addressing them.
best preserved by study continuation. The European
Medicines Agency similarly recommended a risk-based Ethical considerations for DCTs
approach to study monitoring, focusing on the need for Our analysis draws on both clinical research ethics and
data essential to participant safety, rights, and wellbeing. digital ethics frameworks, which are premised on sub­
The Danish Medicines Agency not only issued guidance stantive principles (eg, respect for persons, beneficence,
on clinical trials during the pan­ demic, but also on and justice)14,15 and incorporate procedural values (eg,
decentralised trials in general, stating that decentralisation ensuring responsiveness, sound research design, and
is a feature of clinical research that is bound to continue. transparency).16,17 Although elaborating on the conceptual
The Swiss and the Swedish agencies for therapeutic foundations for these frameworks is beyond the scope of
products issued separate guidance on decentralised clinical this Health Policy report, we use these frameworks to
trials (DCTs) unrelated to the COVID-19 pandemic (table). help identify and evaluate the ethical issues associated
Pharmaceutical companies and contract research organ­ with DCTs. We describe three broad areas requiring
isations are adopting these decentralised approaches, and increased ethical vigilance: participants’ safety and
scholarship on the benefits of the digital components of protection of their rights, scientific validity, and ethical
clinical trials is emerging.11 The multitude of terms used to oversight mechanisms.
describe these types of trials (eg, decentralised, remote,
digital, virtual, and teletrials) indicates both the evolving Advancing participants’ safety and protecting
nature of such clinical trials, and the multiplicity of ways in their rights
which decentralisation can be implemented in research. Advancing participants’ safety and protecting their rights
DCTs (the term that seems to be favoured by regulators) includes consideration of physical safety, privacy,
tend to mix conventional and digitally facilitated methods. informed consent, and wellbeing.

www.thelancet.com/digital-health Vol 5 June 2023 e390

 Health Policy

Regulatory body Title Year

Australia3 The Commonweath Department of Health National principles for teletrials in Australia 2020
Canada4 Health Canada Management of clinical trials during the COVID-19 pandemic: notice to clinical trial 2023
sponsors
Denmark5 Danish Medicines Agency The Danish Medicines Agency’s guidance on the implementation of decentralised 2021
elements in clinical trials with medicinal products
European European Medicines Agency Guidance on the management of clinical trials during the COVID-19 (coronavirus) 2022
Commission6 pandemic
Singapore7 Health Sciences Authority Guidance on the conduct of clinical trials in relation to the COVID-19 situation 2020
Sweden8 Swedish Medical Products Agency Decentralised clinical trials 2021
Switzerland9 Swissmedic and Swissethics Position paper by Swissmedic and swissethics on decentralized clinical trials (DCTs) 2022
with medicinal products
USA10 Food and Drug Administration Conduct of clinical trials of medical products during the COVID-19 public health 2021
emergency–guidance for industry, investigators, and Institutional Review Boards

Table: Regulatory guidance for decentralised clinical trials

Physical safety have substantial data security risks as well, distributed

In conventional research settings, study personnel (eg, networks might increase system vulnerabilities. Appro­
physicians and nurses) help ensure safe handling and priate safeguards should be adopted to protect
administration of investigational products, which is more participants’ privacy. Privacy impact assessment18 can be
challenging in DCTs. DCT protocols should specify how useful to identify technical vulnerabilities and tailor
safety and risk mitigation conditions for the delivery, safeguards to the kind of data collected, especially for
storage, use, disposal, and return of medicinal products participants from social groups facing specific privacy-
and devices are met. If DCTs require decentralised related risks such as discrimination and stigmatisation
collection of biological specimens from participants, due to social perceptions about certain health conditions
investigators and sponsors should ensure appropriate or behaviours.
hygienic conditions and that the risks linked to the Pseudonymisation and anonymisation—together with
sampling procedure do not affect participants’ safety. To data minimisation and privacy-preserving tech­ nolo­
mitigate such risks, self-collection might have to be gies19,20—can offer additional safeguards in DCTs.
restricted to simple, non-invasive procedures, or be done Although adhering to the data minimisation principle in
by specialised personnel. With respect to both product DCTs might be challenging, efforts should be made to
handling and biological samples, participants should be explicitly justify the types of data collected and how they
provided with comprehensive instructions, for instance, will be used. Privacy-by-design and privacy-by-default
through dedicated apps or websites, with user-friendly approaches21 should be followed to prevent risks linked to
communication tools, such as visual aids, infographics, excessive data collection and unauthorised data uses. Data
and videos. protection safeguards should be adopted across all
Participants’ safety is also dependent on careful technical components and data processing activities, and
monitoring for adverse reactions. DCTs generally offer the most stringent privacy-preserving setting should be
fewer opportunities for direct interaction between adopted.
participants and study personnel than con­ ventional Special consideration should be given to the protection
trials. However, this restricted physical contact should of identifiable personal data and information. When
not affect the robustness of safety monitoring. Therefore, digital devices are used to verify the identity of enrolled
participants in DCTs could be given access to a digital participants, monitor intended use of medicinal
platform to register adverse reactions, receive medical products, or communicate adverse reactions, state-of-the-
advice in real time, and automatically trigger a request art encryption should be used to minimise the risk of
for in-person visits at home or a nearby clinical centre in breaches.22 Such safeguards should be clearly described
case of severe adverse reactions. It might be feasible to in the study protocol to be assessed by research ethics
analyse safety issues more continuously than in committees (eg, Institutional Review Boards and their
conventional trials, and possibly protect participants equivalents).
more effectively. Privacy issues might also arise not directly for the
participants but for others living or closely interacting
Privacy with them. For example, family members or caregivers
The use of digital tools and data infrastructures to might be present during study-related activities and
support DCTs entails data protection challenges. Trials virtual interactions. Investigators should be aware of
using wearables, apps, and web-based interactions might such risks and, depending on the type of activity or
increase cybersecurity risks. Although conventional trials circumstances, consider mitigation strategies.23,24

e391 www.thelancet.com/digital-health Vol 5 June 2023

 Health Policy

Informed consent Ensuring scientific validity

DCTs face several ethical and practical challenges related The scientific value of a clinical trial is a precondition to
to the process of obtaining informed consent. First, justify the enrolment of participants.15 Therefore, a
electronic consent approaches must incorporate measures decentralised approach should not be chosen solely
to verify the identity of the person giving consent. Second, because it is more convenient or less expensive than an
remotely ensuring that consent is voluntary and free from in-person study. DCTs, like conventional trials, should
undue influence or coercion can be difficult. Third is the address important evidence gaps and be designed
challenge of providing comprehensive information about rigorously, prospectively registered, and compre­
the research in an easy to understand manner by use of hensively reported. Inclusion criteria should be clearly
digital technologies. Nonetheless, electronic consent defined and their clinical relevance to the study aims
designed appropriately can promote improved under­ must be justified. Study endpoints must be clinically
standing and meaningful autonomous decision making, meaningful to enable the robust assessment of the
despite no physical interaction between study personnel safety and efficacy of the planned intervention. Investi­
and participants.25 gators should thus justify the choice of study endpoints
Outside the research context, important information and related measurements if they propose to replace
such as online privacy notices are notoriously ineffective endpoints typically assessed with physical examinations
at informing users and enabling them to make informed in conventional trials. The choice of study endpoints
decisions.26 This type of outcome should be avoided in (including digitally acquired ones) should be justified
the case of electronic consent for DCTs. Emerging as scientifically valid alternatives to endpoints that
standards about the most appropriate ways to use novel would be adopted in the context of conventional clinical
approaches to consent (ie, interaction design and choice trials.
architecture approaches) can make electronic consent a Adverse events might be more easily reported in
preferred choice to fulfil the ethical aims of informed decentralised than in conventional trials. Measures
consent—conveying the necessary information to taken to minimise the biases and inconsistencies
enable free and autonomous decision making regarding potentially introduced by self-reporting should be clearly
enrol­ment.27,28 Finally, an additional concern for described in the study protocol for appropriate ethical
electronic consent is its legal permissibility and accept­ assessment. The adoption of digital devices and an
ability in various jurisdictions. Although several auditable adverse event reporting (digital) platform
countries have issued clear guidance regarding the following specific standards could facilitate safety
legality of electronic informed consent (eg, the UK, assessment and improve accountability of sponsors and
Belgium, and the USA), others only accept a wet ink or investigators. Reporting standards that can be adapted to
in-person signature. DCTs are emerging.31
DCTs can reduce some barriers to participation, such
Wellbeing as distance from a clinical site or absence of
Although digital tools offer convenient means of transportation. In a DCT, however, self-selection bias is
communication between study teams and participants, a risk because of the influence of the digital divide on
excessive reminders and requests can be unnecessarily the composition of the trial population. Individuals
intrusive. To promote participants’ wellbeing, the who might meet inclusion criteria could feel
number of digitally mediated interactions in DCTs discouraged to enrol due to unfamiliarity with the
should be kept to a minimum. Qualitative studies digital tools that typically have a key role in DCTs, or
regarding DCTs show that participants can feel due to restricted access to adequate internet
overburdened by many technologies and devices.29 Due connectivity. However, digital proficiency assessment
consideration should also be given to the subtle tools (eg, DigiComp2.132) can allow investigators and
intrusiveness of data collection practices taking place in research ethics committees to predict the level of digital
the background, such as through wearables and other proficiency needed for participation and to take
connected devices that might leave participants with the corrective or mitigating actions if necessary. Systematic
impression of being constantly surveilled.30 exclusion of participants from some demographic
In conventional research, participants generally get groups from clinical research results in what has been
financially reimbursed for the expenses they incur due to termed health data poverty33—impairing what we know
trial participation, such as transportation costs. Direct about the safety and efficacy of new drugs for
costs of trial participation are arguably reduced in DCTs. populations not included in clinical trials, which
However, as participants will be asked to complete deprives people from the excluded groups of the
several actions related to the study on their own, a shift in benefits of biomedical knowledge.
labour might occur from the study personnel to the
participants themselves. To avoid exploitation of Enhanced ethics oversight
participants’ time and resources, an assessment of direct The issues identified in this Health Policy report (panel)
and indirect costs for participants should be done. and their respective mitigation strategies should be

www.thelancet.com/digital-health Vol 5 June 2023 e392

 Health Policy

Panel: Questions to consider in decentralised clinical trials

Several specific issues across different ethics domains need to • Have privacy-preserving approaches such as data
be addressed to advance participants’ safety, protect their minimisation, privacy-by-design, and privacy-by-default been
rights, and ensure scientific validity. taken into account?
• Have appropriate measures been taken to minimise the
Physical safety
intrusiveness of data collection and communication activities?
• Is it safe to deliver the investigational product or intervention
remotely? Informed consent
• Can the investigational product or intervention be self- • Is the electronic consent process based on accepted
administered? standards?
• Has comprehensive safety-related information been • What is the process of identity verification of electronic
provided to participants? consent?
• Is there timely online help available for participants? • What measures are in place to ascertain voluntary
• Can specimens be collected safely by non-specialised participation?
personnel? • Are electronic consents and signatures accepted in the
• Can specimens be stored safely by the participants? jurisdiction of the trial?
• Is there sufficient information and assistance regarding • Is the information provided to potential participants through
specimen and data collection available to participants? informed consent comprehensive and understandable?
• Are wearable devices, apps, and other digital equipment used
Wellbeing
by participants in the study validated for clinical use?
• Does the study rely on reasonable and justifiable amounts of
• Is there a system in place for efficient monitoring of adverse
active engagement of research participants (eg, data
reactions?
collection tasks)?
• Do participants have easy access to medical consultation in
• Have direct and indirect costs of participation been assessed?
case of adverse reactions?
• Will direct and indirect costs of participation be
Privacy compensated?
• Do digital devices such as wearables and apps used by
Scientific validity
participants for the purposes of the study carry specific data
• Are inclusion criteria and study endpoints scientifically sound
security risks?
compared with those that would be adopted in conventional
• Are stopping rules and other harm mitigation mechanisms in
trials?
place in case of data breaches or other technical failure
• Has the risk of self-reporting bias for adverse events been
affecting data security?
assessed?
• Has a privacy impact assessment been done?
• Has a digital proficiency assessment been done?
• Have data encryption approaches been considered in relation
• Has the risk of self-selection bias been assessed?
to actual privacy risks?

included in risk–benefit assessments determining the should consider that participant safety includes protection
ethical acceptability of DCTs. Investigators and research of their informational privacy and make use of the tools
ethics committees will have to increase their vigilance that are available for assessing information privacy risks
when identifying the novel risks raised in DCTs. Despite in the digital environment. Privacy impact assessments
the promise of DCTs to address many of the deficits in and digital proficiency assessment tools can assist with
conventional clinical trials, not all trials will be suitable identifying the hidden risks including those arising from
for such an approach. Therefore, the choice of the persisting digital divide. Stopping rules and
decentralising requires careful consideration (panel). contingency plans should cover data breaches and other
Concerns have been raised about the ability of invest­ possible technical issues. Second, the safeguards in place
igators and research ethics committees to competently to help ensure that DCTs are done ethically should be
evaluate research projects involving big data analytics transparently reported. Open reporting about such
and other digital elements.34 Specifically, the types of safeguards is particularly useful in this evolving area.36
risks that digital approaches might pose for participants’ Finally, independent oversight bodies, such as research
informational privacy, their communities, and society ethics committees, and Data and Safety Monitoring
might be difficult to recognise and consider in Boards, should ensure they have adequate expertise in
conventional risk–benefit assessments.35 digital health and DCTs. The guidance issued by drug
To ensure that DCTs receive the necessary ethical regulators can serve as a basis for clarifying some of the
scrutiny, we recommend a three-pronged approach. First, ethical criteria for DCTs, but additional deliberation by
those engaged in DCTs (including investigators, sponsors, national ethics bodies about these issues would be
research institutions, and contract research organisations) welcome.

e393 www.thelancet.com/digital-health Vol 5 June 2023

 Health Policy

Conclusions 13 Doroshow JH, Prindiville S, McCaskill-Stevens W, Mooney M,

DCTs will continue to evolve and are likely to become a Loehrer PJ. COVID-19, social justice, and clinical cancer research.
J Natl Cancer Inst 2021; 113: 1281–84.
common approach to clinical research. To ensure that this 14 National Commission for the Protection of Human Subjects of
evolution results in robust scientific evidence, while Biomedical and Behavioral Research. The Belmont Report: ethical
protecting the rights and wellbeing of research participants, principles and guidelines for the protection of human subjects of
research. Washington, DC: US Government Printing Office, 1978.
adherence to clear ethical standards is essential. 15 Emanuel EJWD, Wendler D, Grady C. What makes clinical research
Contributors ethical? JAMA 2000; 283: 2701–11.
EV designed the study. EV and AB conducted the normative and policy 16 Vayena E, Blasimme A. Health research with big data: time for
analysis and wrote the first draft. All authors revised the draft and systemic oversight. J Law Med Ethics 2018; 46: 119–29.
approved the final manuscript. 17 Floridi L. Translating principles into practices of digital ethics:
five risks of being unethical. Philos Technol 2019; 32: 185–93.
Declaration of interests
18 Clarke R. Privacy impact assessment: its origins and development.
JS is a member of Merck KGaA’s Ethics Advisory Panel and Stem Cell Comput Law Secur Rep 2009; 25: 123–35.
Research Oversight Committee, IQVIA’s Ethics Advisory Panel, Aspen
19 Tene O, Polonetsky J. Privacy in the age of big data: a time for big
Neurosciences Clinical Advisory Panel, and Merck Data Monitoring decisions. 2012. https://www.stanfordlawreview.org/online/privacy-
Committee; and a consultant to Biogen. EV is a member of IQVIA’s Ethics paradox-privacy-and-big-data/(accessed March 30, 2023).
Advisory Panel and a member of Merck KGaA’s Digital Ethics Advisory 20 Froelicher D, Troncoso-Pastoriza JR, Raisaro JL, et al. Truly privacy-
Panel. EV co-chaired the WHO expert group on Ethics and Governance for preserving federated analytics for precision medicine with
AI in Health. This work was supported by the Swiss National Science multiparty homomorphic encryption. Nat Commun 2021; 12: 1–10.
Foundation, NRP77 research grant number 407740_187356. 21 Cavoukian A. Privacy by design: the 7 foundational principles.
Acknowledgments Inf Priv Comm Ont Can 2009; 5: 12.
We would like to thank Shannon Hubbs for her support and assistance 22 Silva BM, Rodrigues JJ, Canelo F, Lopes IC, Zhou L. A data
in the literature review and manuscript editing connected to this work. encryption solution for mobile health apps in cooperation
environments. J Med Internet Res 2013; 15: e66.
References 23 Malin B, Loukides G, Benitez K, Clayton EW. Identifiability in
1 Banks MA. Core concept: in the wake of COVID-19, decentralized biobanks: models, measures, and mitigation strategies. Hum Genet
clinical trials move to center stage. Proc Natl Acad Sci USA 2021; 2011; 130: 383–92.
118: 1–4. 24 Altman RB, Clayton EW, Kohane IS, Malin BA, Roden DM.
2 van Dorn A. COVID-19 and readjusting clinical trials. Lancet 2020; Data re-identification: societal safeguards. Science 2013;
396: 523–24. 339: 1032–33.
3 The Commonwealth Department of Health. National principles for 25 Nebeker C, Gholami M, Kareem D, Kim E. Applying a digital health
clinical trials, including teletrials in Australia. 2020. https://www. checklist and readability tools to improve informed consent for
health.gov.au/sites/default/files/documents/2021/03/national- digital health research. Front Digit Health 2021; 3: 690901.
principles-for-teletrials-in-australia.pdf (accessed March 30, 2023). 26 Schaub F, Balebako R, Cranor LF. Designing effective privacy
4 Health Canada. Management of clinical trials during the COVID-19 notices and controls. IEEE Internet Comput 2017; 21: 70–77.
pandemic: notice to clinical trial sponsors. 2023. https://www. 27 US Department of Health and Human Services. Use of electronic
canada.ca/en/health-canada/services/drugs-health-products/drug- informed consent in clinical trials. 2016. https://www.hhs.gov/
products/announcements/management-clinical-trials-during-covid- ohrp/news/announcements-and-news-releases/2016/use-electronic-
19-pandemic.html (accessed March 30, 2023). informed-consent-clinical-trials/index.html (accessed Oct 6, 2022).
5 Danish Medicines Agency. The Danish Medicines Agency’s 28 Wilbanks J. Design issues in e-consent. J Law Med Ethics 2018;
guidance on the implementation of decentralised elements in 46: 110–18.
clinical trials with medicinal products. 2021. https://
29 Coyle J, Rogers A, Copland R, De Paoli G, MacDonald TM,
laegemiddelstyrelsen.dk/en/news/2021/guidance-on-the-
Mackenzie IS. Learning from remote decentralised clinical trial
implementation-of-decentralised-elements-in-clinical-trials-with-
experiences: a qualitative analysis of interviews with trial personnel,
medicinal-products-is-now-available/~/media/5A96356760ED408C
patient representatives and other stakeholders. Br J Clin Pharmacol
BFA9F85784543B53.ashx (accessed March 30, 2023).
2022; 88: 1031–42.
6 European Commission. Guidance on the management of clinical
30 Nebeker C, Bartlett Ellis RJ, Torous J. Development of a decision-
trials during the COVID-19 (coronavirus) pandemic. 2022. https://
making checklist tool to support technology selection in digital
health.ec.europa.eu/system/files/2022-02/guidanceclinicaltrials_
health research. Transl Behav Med 2020; 10: 1004–15.
covid19_en_1.pdf (accessed March 30, 2023).
31 Orkin AM, Gill PJ, Ghersi D, et al. Guidelines for reporting trial
7 Health Sciences Authority. Guidance on the conduct of clinical
protocols and completed trials modified due to the COVID-19
trials in relation to the COVID-19 situation. 2020. https://www.hsa.
pandemic and other extenuating circumstances: the CONSERVE
gov.sg/docs/default-source/hprg-io-ctb/hsa_ctb_covid-19_guidance_
2021 statement. JAMA 2021; 326: 257–65.
for_clinical_trials_29jul2020.pdf (accessed March 30, 2023).
32 Carretero Gomez S, Vuorikari R, Punie Y. DigComp 2.1: the digital
8 Swedish Medical Products Agency. Decentralised clinical trials. 2021.
competence framework for citizens with eight proficiency levels and
https://www.lakemedelsverket.se/en/permission-approval-and-
examples of use, EUR 28558 EN. Luxembourg: Publications Office
control/clinical-trials/medicinal-products-for-human-use/
of the European Union, 2017.
decentralised-and-virtual-interventional-clinical-trials (accessed
March 30, 2023). 33 Ibrahim H, Liu X, Zariffa N, Morris AD, Denniston AK. Health
data poverty: an assailable barrier to equitable digital health care.
9 Swissmedic. Position paper by Swissmedic and swissethics on
Lancet Digit Health 2021; 3: e260–65.
decentralized clinical trials (DCTs) with medicinal products. 2022.
https://www.swissmedic.ch/swissmedic/en/home/ 34 Ferretti A, Ienca M, Sheehan M, et al. Ethics review of big data
humanarzneimittel/clinical-trials/clinical-trials-on-medicinal- research: what should stay and what should be reformed?
products/publikationen.html (accessed March 30, 2023). BMC Med Ethics 2021; 22: 51.
10 US Food and Drug Administration. Conduct of clinical trials of 35 Cohen IG, Mello MM. Big data, big tech, and protecting patient
medical products during the COVID-19 public health emergency. privacy. JAMA 2019; 322: 1141–42.
2021. https://www.fda.gov/media/136238/download (accessed 36 Sounderajah V, McCradden MD, Liu X, et al. Ethics methods are
March 30, 2023). required as part of reporting guidelines for artificial intelligence in
11 ADVARRA. Ethical issues in the design and review of decentralized healthcare. Nat Mach Intell 2022; 4: 31–37.
research. 2022. https://info.advarra.com/ethical-issues-in-design-of-
decentralized-clinical-trials.html (accessed March 30, 2023). Copyright © 2023 The Author(s). Published by Elsevier Ltd. This is an
Open Access article under the CC BY-NC-ND 4.0 license.
12 Inan OT, Tenaerts P, Prindiville SA, et al. Digitizing clinical trials.
NPJ Digit Med 2020; 3: 101.

www.thelancet.com/digital-health Vol 5 June 2023 e394