A framework for tiered informed consent for health genomic research in Africa

Abstract

A generic framework for providing participant information and implementing a tiered consent process for health genomic research in Africa can help to harness global health benefits from sharing and meta-analysis of African genomic data while simultaneously respecting and upholding the autonomy and individual choices of African research participants.

Full text

comment
A framework for tiered informed consent for
health genomic research in Africa
A generic framework for providing participant information and implementing a tiered consent process for health
genomic research in Africa can help to harness global health benefits from sharing and meta-analysis of African
genomic data while simultaneously respecting and upholding the autonomy and individual choices of African
research participants.
Victoria Nembaware, Katherine Johnston, Alpha A. Diallo, Maritha J. Kotze, Alice Matimba,
Keymanthri Moodley, Godfrey B. Tangwa, Rispah Torrorey-Sawe and Nicki Tin
African human genome research is
advancing rapidly, owing to falling
sequencing costs and international
interest in African genomic data: the
diversity of African genomes can provide
novel insights into biological and etiological
mechanisms, thereby promoting diagnostic,
prognostic and therapeutic advances for
populations in Africa and the rest of the
world1,2. Conducting genomic research in
Africa can be logistically challenging3,4,
but equally challenging is recruiting
African participants—many of whom
have knowledge- and/or poverty-related
vulnerabilities5,6—while ensuring that they
are properly informed and truly consenting,
and that they retain their autonomy and
agency (examples in refs. 5,7–12).
Dynamic consent models, such as
those using ongoing engagement through
online media13,14, ensure autonomy and
choice for participants. However, these
models cannot be implemented in many
African environments, owing to suboptimal
internet and smartphone access, and poor
digital literacy. In broad-consent models,
participants consent to all future use of their
samples under the oversight of an access
committee, but this consent comes at the cost
of the autonomy and individual preferences
of participants10,15. Tiered informed consent
addresses these challenges by providing
detailed information about the intended
specimen/data use and storage, thus enabling
participants to individually select a level
of specimen and/or data sharing through
responses to specific questions10,16–18.
The launch of the H3Africa genotyping
chip, with 2.7 million African-specific
genomic variants19, has increased
opportunities for meta-analyses combining
multiple cohorts of African participants.
To undertake such studies with an ethical
mandate, research participants must
be properly informed, and their data/
sample-use preferences must be accurately
understood, faithfully recorded and
implemented with integrity. We propose a
framework for undertaking ethically sound,
tiered-informed-consent processes in Africa,
which provides a comprehensive guide on
compiling participant information and an
informed-consent template for capturing
each participant’s consent information and
mapping it to data-use ontologies.
Methods
The framework is derived from the authors’
combined experiences with informed
consent in Africa. A generic template for
participant information (Supplementary
Note) is provided, and validated translation
into the participants’ primary language(s) is
recommended. Framework tenets include
ease of understanding for field-workers and
participants, and practical administration in
a busy facility (such as clinics). We describe
the core components of tiered consent for
competent adult participants, although not
all elements may be appropriate for every
study. We provide recommendations for data
capture and standardization of participant
consent information.
What information should be provided
in the informed-consent documents?
Ethical research requires balancing benefits
and risks at the micro level for individuals,
the meso level for communities and the
macro level for populations. Although
ethics review boards consider all levels in
assessing a research study, an individual’s
right to decline or accept participation
remains paramount, and participants must
be provided with necessary and sufficient
information to support this right. We
present core concepts and highlight how
localized knowledge can be incorporated in
providing information to participants.
Information about genetics. African
colloquialisms often speak to an inherent
understanding of heredity. For example,
the Shona proverb “Mhembwe rudzi
inozvara mwana ane kazhumu” translates
to “The child of a duiker [small antelope]
is a duiker,” and its equivalent in English
is “Like father, like son.” Such local
expressions can be harnessed in explaining
heredity and genetic concepts by using
anecdotally accessible, emotionally neutral
examples such as height or facial similarity.
Caucasian-centric examples, such as eye
or hair color, are often inappropriate for
African participants.
Information about genetics and health
can be related to locally prevalent health
conditions, while always providing clarity
about complex risk factors to prevent
misunderstanding, anxiety or family
conflict—for example, an explanation that
genetic factors might influence susceptibility
to malaria, but environmental factors are the
chief drivers of becoming infected.
The focus of the study, and who is doing
the research. Clear, simple language
and local names for health conditions
can simplify explanation of the research
question. For example, “We want to
understand whether genes affect how likely
someone is to get sick from bilharzia”
is more accessible than “The primary
study objective is to elucidate the genetic
etiology of schistosomiasis” for an African
population exposed to schistosomes.
However, information accuracy supports
transparency, and researchers should not
confuse straightforward language with
incomplete information. During recruitment
of controls, researchers can explain that
comparing samples from people with and
without an illness can help understand
what contributes to getting the illness. Local
researchers and institutions should always
be named as the primary contact, to ensure
that they are familiar, un-intimidating,
identifiable and contactable for participants.
NATURE GENETICS | www.nature.com/naturegenetics

comment
What will you be asked to provide or do
in this study?. A brief, simple explanation
of exactly what will be requested in terms
of visits, data and sample collection, and
common-use estimates of collection
volumes can inform participants about the
collection of blood or saliva. For example,
referring to “about two teaspoons” of blood
is more readily understood by participants
than “10 ml.”
What are the potential risks and benefits
of this study?. There are often no direct
benefits to participants in genome research;
in such cases, the lack of benefits should
be honestly stated. However, researchers
should also recognize and respect that
African participants may also value altruistic
behavior and contributing to the well-being
of others or the advancement of science,
regardless of their circumstances20,21.
Research studies may be misunderstood by
participants as an offer of additional health
care or an opportunity for cure, and if such
benefits will not be provided, this must be
explicitly stated.
Communicating risks is complicated
by the unknowable nature of future risks
associated with genomic data: the rapidly
evolving landscape makes cataloging future
possible uses impossible. Existing risks
include re-identification of individuals and
exposure of personal health information in
the event of data breach or inappropriate
data reuse22–24, as well as stigmatization
of families, communities or ancestral
groups. Another risk to be communicated
to participants is discovery of information
about the participant’s health or information
that might negatively affect family members
or local community members who did not
necessarily consent to the study. Describing
the risks alongside clear, practical plans for
risk mitigation can reassure participants that
risks have been appropriately identified and
planned for.
Remuneration for costs incurred by study
participation, or refreshments provided,
should be detailed separately to avoid
confusion with ‘study benefits’.
Privacy protection, and data- and
specimen-protection protocols. The
processes and infrastructure in place to
protect data and specimens should be briefly
outlined to assure participants that privacy
protection is in place. Sample storage
locations and security measures should
be detailed—including the geographical
locations of storage and use, what data
will be generated from the samples, and
how and by whom those data will be used
further. Information can be provided
about the committees that will oversee
access and the plans for sharing aggregated
or group-level data with collaborators,
scientific journals, international
collaborators or online platforms.
Return of results. General study findings
can be communicated to participants in
a context-appropriate process involving
posters and pamphlets distributed in
recruitment facilities, cell-phone messaging
or website updates. Community meetings
with the researchers to provide feedback on
findings should occur only if participant
confidentiality can be protected, especially
for studies on sensitive health issues. The
return of findings from genome-related
research requires approval by a research
ethics committee and medical specialists
who can determine whether the findings
(primary or incidental) meet criteria for
return to individual participants and are
actionable. The consent process should
support the participants’ right not to know
certain results, whether actionable or not
actionable, within their current context.
Furthermore, the process of disclosing
research results should involve professionals
with the appropriate expertise, and a
summary of this planned process should be
communicated to participants.
For the initial study for which
participants are recruited, participant
information should detail whether
individual results will be made available; if
so, the process for returning results should
be clearly communicated. For example,
participant information might state that
no results will be returned to individuals
because the findings will not be sufficient to
provide accurate health-related information;
that a doctor will provide results during a
clinical consultation with the participant; or
that an individual report will be provided by
appropriately registered medical scientists.
This section must also inform participants of
what action will be taken if a communicable
disease is identified, including how the
participant will be informed, plans for
linkage to care and the process for infectious
diseases that must legally be reported to a
central/national registry.
Information should describe how study
findings will be shared with the participant
community, for example through a project
website (with provided URLs) and/or
newsletters by e-mail or hard copy. The
intended publication of results within the
research sector can also be described.
Who can be contacted with questions or
concerns, and how to withdraw consent.
Names and contact information must
be provided for participant questions or
concerns. In addition, contact details for
an oversight body independent of the study
researchers, such as an ethics review board,
which can address concerns impartially,
should be provided. Clear instructions on
an easy process to withdraw from the study
must be provided, with reassurance that
withdrawal will not affect access to standard
health care if recruitment takes place in a
health facility.
Questions to be asked for each
component of the tiered consent
We propose a series of questions that each
define a tier of the consent process. The
first question defines inclusion in the initial
study, and a response of ‘yes’ is required to
proceed. The questions that then follow are
designed to be freestanding, and each can be
independently agreed to or declined, so that
participants define a particular combination
of data/biospecimen uses with which they
feel comfortable.
Question 1: Agreement for collection of
data/biospecimens for the primary study.
“Do you agree for us to collect this saliva/
blood sample and your health information
for this study that we have described on
how genes might affect [specific health
phenotype]?”
This question defines consent to
participate in the current (primary) study
for which participants are being recruited.
Often in health research, primary studies
examine a specific disease, for example,
tuberculosis or hypertension.
Question 2: Agreement for secondary
data/biospecimen use for other studies on
the same phenotype/health condition. “Do
you agree for us to use your genetic sample
together with your health information in
other studies in the future on the effects of
genes on [specific health phenotype]?”
This question provides an opportunity
for individuals to contribute to furthering
research on a specific disease that they
might feel particularly strongly about
because of their personal experience,
without committing to general research use
in other types of health research.
Question 3: Agreement for general
secondary data/biospecimen use in other
unrelated studies. “Do you agree for us to
use your genetic sample together with your
health information for other studies in the
future to study the effects of genes on other
conditions or biological processes?”
This option enables general re-use of data
and samples in future studies that are not yet
defined or known, while providing the option
for those who are not comfortable with wider
re-use of their data/sample to clearly define
NATURE GENETICS | www.nature.com/naturegenetics

comment
a boundary of health or biological function
research for secondary use.
Question 4: Agreement for inclusion in
aggregated data (for example, genome
summary data) for the study. “Sometimes
researchers combine the genetic information
from everyone in the study and provide
a summary of genetic data for the whole
group. Do you agree for us to use your
information when providing combined
information about the whole research group
(x total individuals in this study)?”
A recent policy change by the US
National Institutes of Health25 about open
sharing of genomic summary data from
studies has prompted this consideration.
The risk of stigmatization or of
discrimination against ancestral groups
is substantial in Africa, where the genetic
distance between subgroups can be
large, and communities can be small and
easily identified. Both historically and
recently, the term ‘ethnicity’ has been a
sensitive or volatile identifier. We believe,
therefore, that consent should be obtained
before individuals’ data are released
within an aggregate dataset. Providing
the total number of participants can give
an indication of the likelihood of being
re-identified.
Question 5: Agreement for re-contact for
follow-on studies. “Sometimes, what we
find from a study like this might lead to
new studies being done in the future.
Can other researchers contact you in the
future to invite you to take part in other
research studies?”
Re-contact of individuals can be
difficult, owing to the high geographic
mobility of some African populations,
shared or transitory cell-phone ownership,
informal residential addresses and limited
modes of contact. Re-contact outside
the consented study violates the
privacy of participants and constitutes
inappropriate or even illegal secondary
use of personal information unless consent
has been specifically given for such
re-contact. However, some individuals
may agree to be contacted about other
studies, particularly to facilitate access to
specialist health care.
Question 6: Agreement for return of
defined genetic findings from the current
study. “In this study, we hope to identify
genetic factors that mean that someone is
more likely to [have an outcome, such as
susceptibility to a disease]. If someone has
this genetic factor, there is [no treatment/we
recommend treatment x]. If we find during
this study that you have this kind of genetic
factor, would you like us to tell you this
information?”
When genomic studies investigate specific
genetic factors affecting health, consent for
return of results is easier to delineate. For non-
African settings, recommendations suggest
that participants be asked whether they wish
to receive clearly identified, actionable and/
or non-actionable genetic findings from the
study26. The practical implications, including
treatment recommendations, can be defined
in advance for the participants, who can use
contextualized information and personal
preferences to decide which results they wish
to receive. Challenges arise around limited
health care access and whether participants
can access or afford a recommended
intervention. Researchers and ethics review
boards must address these questions
within the study context to maximize
benefits through diagnostic and therapeutic
information while minimizing emotional
and societal harm in cases in which such
information cannot improve health outcomes.
Question 7: Agreement for return of
unanticipated genetic findings from future
studies. “Sometimes what we find from our
research might include new information
about your health. Would you like us to
contact you again if we believe we have
new information that may directly affect
your health—if there is some kind of
action or treatment that might be able to
help you with the health issue (yes/no) and
if there is no kind of action or treatment that
might be able to help you with the health
issue (yes/no)?”
Implications for returning results become
more nuanced and difficult to communicate
for secondary use of data/biospecimens,
because the nature of future genetic or
health findings is currently unknown. We
propose that a distinction should be made
between actionable and non-actionable
results, to accommodate individuals
who prefer to know nothing about their
etiological genetic background; those
who wish to know only information with
available interventions; or those who wish
to know all genetic components to their
health, both actionable or unactionable.
The principal investigators, data-access
committee and ethics review board must
ensure that secondary-analysis protocols
include plans for the return of results
and assessments of appropriate available
interventions in a consultative, informed
and supported manner.
How to store informed-consent choices
Common practice in Africa is to capture
participant consent in hard copy, without
sharing the consent with secondary users.
Digitalization of informed consent usually
consists of an e-consent signature or scanned
copy of the signed consent form in a non-
machine-readable format. These practices
do not support tiered consent, which
captures multiple accessible, query-able and
actionable choices for individuals. Consent
interpretation and automated selection of
data/samples for secondary use can also
be challenging, owing to unstandardized
and heterogeneous consent questions;
moreover, different consent standards
between fields, such as clinical and genomic
research, also limit cross-disciplinary data
sharing27. Ontologies such as the Informed
Consent Ontology (ICO)28 and Data Usage
Ontology (DUO)27,29 provide standardized
terminology and systems to semantically
label samples and data with their usage
restrictions: ICO captures the process
of obtaining informed consent, whereas
DUO describes consent and governance
categories, coding 19 primary and secondary
data-use cases. Tools, templates and software
can assist with practical implementation
of DUO30 and capture metadata of consent
information, restrictions and requirements
for a study, but they do not capture
individual consent choices30. DUO27 lacks
some key consent codes for return of
actionable and nonactionable, anticipated
and unanticipated findings, aggregated data
use and consent for participant re-contact
after the study. However, consent ontologies
and coding are likely to evolve to meet such
requirements.
Simple, low-cost, systematic strategies
are needed to capture, store, share and
take action on individuals’ tiered consent
choices through commonly used platforms
(for example REDCap31). Each informed-
consent question requires binary response
options as checkbox items; the case-report
form should be versioned; and any consent
changes should be exported to relevant
laboratories, biobanks, data repositories or
data analysts for action to be taken if sample
destruction or data deletion is required. An
example of data-use consent information
and corresponding DUO codes for a sickle-
cell disease genetic study is shown in Box 1.
Each tier of consent is coded as a binary
variable (yes/no, 0/1), with the date of
consent. When individuals modify their
consent information, a database structure
can allow for the addition of new consent
data with a combined, unique date-
study ID key. A ‘current’ flag facilitates
identification of most recent consent data
(Table 1). The binary matrix design enables
simple, intuitive data capture of consent
information, and combinations of binary
values are subsequently mapped to consent
codes from DUO or other ontologies as they
NATURE GENETICS | www.nature.com/naturegenetics

comment
become available or are updated. Future
work to improve data-capture fidelity may
include image processing to automatically
capture and code responses from signed
consent documents. A consent ‘metadata’
record can also describe the type of
informed-consent questions asked during
that study.
Discussion
This framework is a practical guide for
preparing informed-consent documents
for consultation with people considering
participation in a genetic or genomic
research study, assuming that informational
tools including pamphlets, videos and flip
charts are also used, and a community-
engagement process precedes and continues
during the study. Although beyond the
scope of this publication, community
engagement before human genomic research
in Africa is paramount15,32.
Individual informed consent addresses
the micro level; however, at the meso
and macro levels, families, communities,
religious groups, ancestral groups and
populations are all affected by genomic
data from individuals32,33, particularly
in Africa, where populations are highly
diverse and often genetically unique
and re-identifiable, and ancestry-based
perceptions have previously fueled life-
threatening discord. Consent processes
must by necessity be situated in this
broader context34. Sufficient time should be
given between providing information
to potential participants and the enrolment
visit, to respect family and communal
decision-making; in addition, providing
a recap of consent options in multiple visits
is advisable.
While compiling the framework, we
identified issues requiring further exploration
regarding the recommendations for
implementation of tiered informed consent
in Africa. These include the following.
Return of results. Incidental or unforeseen
results from secondary data analysis
are challenging to address during tiered
informed consent and must be considered
within the context of each study. This
area has a high risk of unforeseeable
harm to participants, families and
communities, and ongoing research into
the return of results and how to define
‘actionable results’ in Africa and low- and
middle-income countries (LMICs) in
general is essential.
Informed consent during times of crisis.
The recent Ebola crisis in West Africa
highlighted issues in using biological
samples toward the common good and
in informed-consent processes in times
of crisis35–37. We must explore waiver of
informed consent in times of crisis in which
minimum use of samples/data for public
good might override individual rights, or
individuals might be too ill to consent:
within the tiered-consent model, levels of
consent that might be acceptably waived and
those that should remain non-negotiable
may be identified.
Consent for vulnerable populations. We
reiterate that these consent guidelines are
intended for competent and autonomous
adult participants. Particularly in LMICs
and in Africa, many participants may be
vulnerable, including those with limited
access to health care and socioeconomic
resources, children and adolescents,
disenfranchised women, persecuted
ancestral groups and people marginalized
or criminalized because of their sexual
orientation and/or gender identity. For
studies involving potentially vulnerable
participants, we recommend that specialist
advice be sought to ensure an appropriate
informed-consent process.
Intended commercialization of study
findings. Intended commercialization of
findings provides a special case for consent,
and benefit-sharing agreements with
participants require research to address
the potential for coercion or inducement
into participation with promise of financial
rewards, community consultation around
appropriate avenues to return benefits, and
social constructs and community pressure to
participate that might arise from promises
of community-level benefits with sufficient
participation rates.
Legislative constraints for consent
processes. Protection of participant data
and confidentiality is increasingly protected
by new legislation, and consent processes
must comply with national laws and
regulations. In South Africa, for example,
the Protection of Personal Information Act
may affect whether broad consent can be
legally obtained from participants. Local
legal advice is essential to ensuring that local
legislation is respected.
Consent to studies of population
origins. We have intentionally omitted
secondary-use consent for studies of
ancestry or population origins in this
framework, because of the complexity of
risks and future-use cases for such research.
Recruiting health-research participants
entails approaching ill people who may have
personal motivations to participate in health
research. We propose that population-
diversity genome-research consent should
be addressed separately to avoid a ‘bait-
and-switch’ approach in which vulnerable
participants consent to health research but
provide consent to population-diversity
research as a secondary mechanism without
necessarily understanding the full risks or
implications of such research.
We present this framework as a starting
point for implementing tiered informed
consent in Africa, providing a generic
Box 1 | Illustration of the coding of tiered-consent forms for two participants being
recruited into a sickle-cell disease genetic study
Participant 1
1.1. Do you agree for us to use your
genetic sample together with your health
information for this study on the eects
of genes on sickle-cell disease?
Answer: Yes
1.2. Do you agree for us to use your
genetic sample together with your health
information for other studies in the future
on the eects of genes on sickle-cell
disease? Answer: Yes
1.3. Do you agree for us to use your
genetic sample together with your health
information for other studies in the
future to study the eects of genes on
other conditions or biological processes?
Answer: No
DUO requirements/restriction
description: Disease-specic research and
clinical care
DUO requirement code: DS-(XX)(CC)
Participant 2
1.1. Do you agree for us to use your
genetic sample together with your health
information for this study on the eects of
genes on sickle-cell disease? Answer: Yes
1.2: Do you agree for us to use your
genetic sample together with your health
information for other studies in the future
on the eects of genes on sickle-cell
disease? Answer: Yes
1.3: Do you agree for us to use your
genetic sample together with your health
information for other studies in the
future to study the eects of genes on
other conditions or biological processes?
Answer: Yes
DUO requirements/restriction description:
Use of the data limited to health/medical/
biomedical research but not population
origin/ancestry
DUO requirement code: HMB(CC)
NATURE GENETICS | www.nature.com/naturegenetics

comment
Table 1 | Example of data coding and capture for tiered-consent questions, with retrospective mapping to DUO
Captured and stored binary variables from the informed-consent process, showing capture of response to each tier Retrospective mapping to
current ontologies; can be
re-mapped as ontologies are
updated or replacedc
Study
ID Date of
consent Version
consent
document
Is
current Use for
primary
study of
health
condition
Use for
other
studies
on this
condition
Use for
other
studies on
health or
general
biology
Use of
aggregate
data for
entire study
Re-
contact
for future
studies
Return of
specific,
described
genetic
results from
primary
study
Return of
undefined,
actionable
results from
other studies
Return of
undefined,
non-
actionable
results from
other studies
DUO consent
description DUO code
ID_1 2010-10-20 1 1 0 NA NA NA NA NA NA NA No consent
given
ID_2 2010-10-20 1 0 1 1 0 0a1 1 1 1 Use of data
must be related
to [disease]
DS-[XX](CC)
ID_2b2010-10-25 1 1 1 0 0 0 0 1 0 0 Use of data
is limited to
use within
an approved
project
PS
ID_3 2010-10-20 1 1 1 1 0 1 1 1 1 1 Use of data
must be related
to [disease]
DS-[XX](CC)
ID_4 2010-10-20 1 1 1 1 1 1 0 1 1 0 Use of data
is limited to
health/medical/
biomedical
research but
not population
origin/ancestry
HMB(CC)
a Currently, aggregate data use, re-contact and return of results are not coded in DUO. b Shows change in consent for individual ID_2 and use of the ‘current’ 0/1 flag. c Fields are retrospectively mapped to DUO by using stored variables and can be updated. NA,
not applicable.
NATURE GENETICS | www.nature.com/naturegenetics

comment
example of participant information and
tiered-consent questions to be adapted for
individual contexts in Africa, other LMICs
and beyond. This example is intended as
an illustration of ways to address different
scenarios for participant information and
consent questions. We welcome dialog and
recommendations for improvements to this
framework to benefit African participants in
the future.
Reporting Summary. Further information
on research design is available in the Nature
Research Reporting Summary linked to
this article. ❐
Victoria Nembaware1,13, Katherine Johnston2,13,
Alpha A. Diallo3,4, Maritha J. Kotze5,6,
Alice Matimba7, Keymanthri Moodley8,
Godfrey B. Tangwa9, Rispah Torrorey-Sawe5,10
and Nicki Tin 2,11,12*
1Division of Human Genetics, University of Cape
Town, Cape Town, South Africa. 2Department of
Integrative Biomedical Sciences, Computational
Biology Division, University of Cape Town, Cape
Town, South Africa. 3Ministry of Health, Conakry,
Guinea. 4Chair of Public Health, University of
Conakry, Conakry, Guinea. 5Department of
Pathology, Division of Chemical Pathology, Faculty
of Medicine and Health Sciences, Stellenbosch
University, Stellenbosch, South Africa. 6Tygerberg
and National Health Laboratory Service, Tygerberg
Hospital, Parow, South Africa. 7Wellcome Genome
Campus Advanced Courses and Scientic
Conferences, Hinxton, UK. 8Centre for Medical Ethics
& Law, Department of Medicine, Faculty of Health
Sciences, Stellenbosch University, Stellenbosch, South
Africa. 9University of Yaounde 1 and Cameroon
Bioethics Initiative (CAMBIN), Yaounde, Cameroon.
10Immunology Department, School of Medicine,
College of Health Sciences, Moi University, Kesses,
Kenya. 11Wellcome Centre for Infectious Disease
Research in Africa, Institute for Infectious Disease
and Molecular Medicine, University of Cape Town,
Cape Town, South Africa. 12Centre for Infectious
Disease Epidemiology and Research, Public Health
and Family Medicine, University of Cape Town,
Cape Town, South Africa. 13ese authors contributed
equally: Victoria Nembaware, Katherine Johnston.
*e-mail: nicki.tin@uct.ac.za
Published: xx xx xxxx
https://doi.org/10.1038/s41588-019-0520-x
References
1. Popejoy, A. B. & Fullerton, S. M. Nature 538, 161–164 (2016).
2. Nordling, L. Na t. News 544, 20 (2017).
3. Adebamowo, S. N. etal. Glob. Health Action 11, 1419033 (2018).
4. Tin, N. Per. Med. 11, 187–196 (2014).
5. Wright, G. E. B., Koornhof, P. G. J., Adeyemo, A. A. & Tin, N.
BMC Med. Ethics 14, 21 (2013).
6. Paré Toe, L. etal. PLoS One 8, e80800 (2013).
7. Hofmann, B. J. Med. Ethics 35, 125–129 (2009).
8. Moodley, K., Sibanda, N., February, K. & Rossouw, T. BMC Med.
Ethics 15, 4 (2014).
9. Schroeder, D., Cook, J., Hirsch, F., Fenet, S. & Muthuswamy,
V. Ethics Dumping: Case Studies from North-South Research
Collaborations (Springer, 2017).
10. Tin, N. BMJ Glob. Health 3, e001249 (2018).
11. Campbell, M. M. etal. PLoS One 12, e0188466 (2017).
12. Munung, N. S. etal. J. Med. Ethics 42, 132–137 (2016).
13. Budin-Ljøsne, I. etal. BMC Med. Ethics 18, 4 (2017).
14. Kaye, J. etal. Eur. J. Hum. Genet. 23, 141–146 (2015).
15. Igbe, M. A. & Adebamowo, C. A. BMC Med. Ethics 13, 27 (2012).
16. Moodley, K. & Singh, S. BMC Med. Ethics 17, 57 (2016).
17. Kim, H. etal. J. Am. Med. Inform. Assoc. 24, 380–387 (2017).
18. Wright, G. E. B., Adeyemo, A. A. & Tin, N. Hum. Genomics 8,
18 (2014).
19. H3Africa consortium array available soon. Illumina. https://www.
illumina.com/company/news-center/feature-articles/h3africa-
consortium-array-available-soon-.html (2016).
20. Browne, J. L. etal. Trop. Med. Int. Health 24, 264–279 (2018).
21. Ogunrin, O., Woolfall, K., Gabbay, M. & Frith, L. PLoS One 13,
e0195171 (2018).
22. Rothstein, M. A. Am. J. Bioeth. 10, 3–11 (2010).
23. Erlich, Y., Shor, T., Pe’er, I. & Carmi, S. Science 362, 690–694 (2018).
24. Gymrek, M., McGuire, A. L., Golan, D., Halperin, E. & Erlich, Y.
Science 339, 321–324 (2013).
25. National Institutes of Health. Update to NIH Management of
Genomic Summary Results Access Notice no. NOT-OD-19-023
https://grants.nih.gov/grants/guide/notice-les/NOT-OD-19-023.
html (2019).
26. National Heart, Lung, and Blood Institute working group. etal.
Circ. Cardiovasc. Genet. 3, 574–580 (2010).
27. Dyke, S. O. M. etal. PLoS Genet. 12, e1005772 (2016).
28. Lin, Y. etal. Development of a BFO-based informed consent
ontology. in Proc. 5th International Conference on Biomedical
Ontology (ICBO, 2014).
29. Genomic data toolkit. Global Alliance for Genomics and Health
https://www.ga4gh.org/genomic-data-toolkit/ (2019).
30. Woolley, J. P. etal. NPJ Genom. Med. 3, 17 (2018).
31. Harris, P. A. etal. J. Biomed. Inform. 42, 377–381 (2009).
32. Folayan, M. O., Oyedeji, K. S. & Fatusi, O. A. Dev. World Bioeth.
15, 1–7 (2015).
33. Kengne-Ouafo, J. A. etal. BMC Public Health 14,
1026 (2014).
34. Moodley, K. & Beyer, C. Biopreserv. Biobank. https://doi.
org/10.1089/bio.2018.0136 (2019).
35. Freudenthal, E. & Hecketsweiler, C. Où sont les échantillons
sanguins infectés par Ebola? Le Monde https://www.lemonde.
fr/afrique/article/2019/01/22/ebola-l-utilisation-opaque-des-
echantillons-sanguins_5412571_3212.html (22 January 2019).
36. Freudenthal, E. Ebola’s lost blood: row over samples own out of
Africa as ‘big pharma’ set to cash in. e Te le gra ph https://www.
telegraph.co.uk/global-health/science-and-disease/ebolas-lost-
blood-row-samples-own-africa-big-pharma-set-cash/
(6 February 2019).
37. Community mistrust worsening DR Congo Ebola outbreak:
study. Al Jazeera https://www.aljazeera.com/news/2019/03/
community-mistrust-worsening-dr-congo-ebola-outbreak-
study-190328062938247.html (28 March 2019).
Acknowledgements
N.T. is supported by the Wellcome Trust (203135/Z/16/Z)
and the National Institutes of Health (awards H3ABioNet
R01HD080465 and B-Positive U24HG006941). K.M.
is funded by The National Human Genome Research
Institute of the National Institutes of Health under award
number U01HG008222. K.J. is funded by National
Institutes of Health H3ABioNet award R01HD080465.
The research of M.J.K. is supported by the Strategic
Health Innovation Partnerships Unit of the South African
Medical Research Council, with funds received from
the Department of Science and Technology (Research
grant number S003665), the Cancer Association of
South Africa and the South African BioDesign Initiative
of the Department of Science and Technology and
the Technology Innovation Agency. R.T.-S. of the Moi
University/Teaching and Referral Hospital in Kenya
received a postdoctoral fellowship from Stellenbosch
University. V.N. is supported by the National Heart, Lung,
And Blood Institute of the National Institutes of Health
under award numbers U24HL135600 and 5U24HL135881.
The content is solely the responsibility of the authors and
does not necessarily represent the official views of the
National Institutes of Health.
Author contributions
N.T. conceptualized the framework and wrote the
manuscript outline. N.T. and all authors developed the
content of the framework and completed and reviewed the
manuscript content.
Competing interests
M.K. is a non-executive director and shareholder of
Gknowmix (Pty) Ltd.
Additional information
Supplementary information is available for this paper at
https://doi.org/10.1038/s41588-019-0520-x.
NATURE GENETICS | www.nature.com/naturegenetics

1
nature research | reporting summary October 2018
Corresponding author(s): Prof. Nicki Tiffin
Last updated by author(s): Sep 5, 2019
Reporting Summary
Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Policy Checklist.
Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested
A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient)
AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)
For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.
Software and code
Policy information about availability of computer code
Data collection No data were collected
Data analysis No data analysis was performed
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers.
We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.
Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A list of figures that have associated raw data
- A description of any restrictions on data availability
N/A
Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf

2
nature research | reporting summary October 2018
Life sciences study design
All studies must disclose on these points even when the disclosure is negative.
Sample size N/A
Data exclusions N/A
Replication N/A
Randomization N/A
Blinding N/A
Reporting for specific materials, systems and methods
We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.
Materials & experimental systems
n/a Involved in the study
Antibodies
Eukaryotic cell lines
Palaeontology
Animals and other organisms
Human research participants
Clinical data
Methods
n/a Involved in the study
ChIP-seq
Flow cytometry
MRI-based neuroimaging
Antibodies
Antibodies used N/A
Validation N/A
Eukaryotic cell lines
Policy information about cell lines
Cell line source(s) N/A
Authentication N/A
Mycoplasma contamination N/A
Commonly misidentified lines
(See ICLAC register)
N/A
Palaeontology
Specimen provenance N/A
Specimen deposition N/A
Dating methods N/A
Tick this box to confirm that the raw and calibrated dates are available in the paper or in Supplementary Information.

3
nature research | reporting summary October 2018
Animals and other organisms
Policy information about studies involving animals; ARRIVE guidelines recommended for reporting animal research
Laboratory animals N/A
Wild animals N/A
Field-collected samples N/A
Ethics oversight N/A
Note that full information on the approval of the study protocol must also be provided in the manuscript.
Human research participants
Policy information about studies involving human research participants
Population characteristics N/A
Recruitment N/A
Ethics oversight N/A
Note that full information on the approval of the study protocol must also be provided in the manuscript.
Clinical data
Policy information about clinical studies
All manuscripts should comply with the ICMJE guidelines for publication of clinical research and a completed CONSORT checklist must be included with all submissions.
Clinical trial registration N/A
Study protocol N/A
Data collection N/A
Outcomes N/A
ChIP-seq
Data deposition
Confirm that both raw and final processed data have been deposited in a public database such as GEO.
Confirm that you have deposited or provided access to graph files (e.g. BED files) for the called peaks.
Data access links
May remain private before publication.
N/A
Files in database submission N/A
Genome browser session
(e.g. UCSC)
N/A
Methodology
Replicates N/A
Sequencing depth N/A
Antibodies N/A
Peak calling parameters N/A
Data quality N/A
Software N/A

4
nature research | reporting summary October 2018
Flow Cytometry
Plots
Confirm that:
The axis labels state the marker and fluorochrome used (e.g. CD4-FITC).
The axis scales are clearly visible. Include numbers along axes only for bottom left plot of group (a 'group' is an analysis of identical markers).
All plots are contour plots with outliers or pseudocolor plots.
A numerical value for number of cells or percentage (with statistics) is provided.
Methodology
Sample preparation N/A
Instrument N/A
Software N/A
Cell population abundance N/A
Gating strategy N/A
Tick this box to confirm that a figure exemplifying the gating strategy is provided in the Supplementary Information.
Magnetic resonance imaging
Experimental design
Design type N/A
Design specifications N/A
Behavioral performance measures N/A
Acquisition
Imaging type(s) N/A
Field strength N/A
Sequence & imaging parameters N/A
Area of acquisition N/A
Diffusion MRI Used Not used
Preprocessing
Preprocessing software N/A
Normalization N/A
Normalization template N/A
Noise and artifact removal N/A
Volume censoring N/A
Statistical modeling & inference
Model type and settings N/A
Effect(s) tested N/A
Specify type of analysis: Whole brain ROI-based Both

5
nature research | reporting summary October 2018
Statistic type for inference
(See Eklund et al. 2016)
N/A
Correction N/A
Models & analysis
n/a Involved in the study
Functional and/or effective connectivity
Graph analysis
Multivariate modeling or predictive analysis
Functional and/or effective connectivity N/A
Graph analysis N/A
Multivariate modeling and predictive analysis N/A