Ethical considerations for Biobanking and Use of Genomics Data in Africa: A narrative review

Background Biobanking and genomic research requires collection and storage of human tissue from study participants. From participants’ perspectives within the African context, this can be associated with fears and misgivings due to a myriad of factors including myths and mistrust of researchers. From the researchers angle ethical dilemmas may arise especially with consenting and sample reuse during storage. The aim of this paper was to explore these ethical considerations in the establishment and conduct of biobanking and biogenomic studies in Africa. Methods We conducted a narrative synthesis following a comprehensive search of nine (9) databases and grey literature. Ethical issues studied related to community knowledge and understanding of biobanking and genomic research, regulation, and governance of same by research ethics committees, enrolment of participants, types of informed consents, data collection, storage, usage and sharing as well as material transfer, returning results and benefit sharing. Results Of 2,663 title and abstracts screened, 94 full texts were retrieved and reviewed for eligibility. We included 12 studies (7 qualitative; 4 quantitative and one mixed methods). More education of study participants is needed, as well as appropriate community engagement processes that allow community confidence in enrolment into such studies. Competence of review and ethics committees (RECs) should be enhanced to adequately review and govern biobanking and genomic research in Africa. Biospecimen collection and storage is given in trust and participants expect confidentially of data and results generated. Most participants are comfortable with broad consent due to trust in researchers, though a few would like to be contacted for reconsenting in future studies, and this would depend on whether the new research is for good cause. Sharing data with external partners is welcome in some contexts but some research participants did not trust foreign researchers. Conclusion With these varying ethical considerations, we recommend that stakeholders, including research ethics committees, work together to adapt and use clearly defined ethical frameworks, guidelines, and policy documents to harmonize the establishment and running of biobanking and genomic research in Africa.


Abstract Background
Biobanking and genomic research requires collection and storage of human tissue from study participants. From participants' perspectives within the African context, this can be associated with fears and misgivings due to a myriad of factors including myths and mistrust of researchers. From the researchers angle ethical dilemmas may arise especially with consenting and sample reuse during storage. The aim of this paper was to explore these ethical considerations in the establishment and conduct of biobanking and biogenomic studies in Africa.

Methods
We conducted a narrative synthesis following a comprehensive search of nine (9) databases and grey literature. Ethical issues studied related to community knowledge and understanding of biobanking and genomic research, regulation, and governance of same by research ethics committees, enrolment of participants, types of informed consents, data collection, storage, usage and sharing as well as material transfer, returning results and bene t sharing.

Results
Of 2,663 title and abstracts screened, 94 full texts were retrieved and reviewed for eligibility. We included 12 studies (7 qualitative; 4 quantitative and one mixed methods). More education of study participants is needed, as well as appropriate community engagement processes that allow community con dence in enrolment into such studies. Competence of review and ethics committees (RECs) should be enhanced to adequately review and govern biobanking and genomic research in Africa. Biospecimen collection and storage is given in trust and participants expect con dentially of data and results generated. Most participants are comfortable with broad consent due to trust in researchers, though a few would like to be contacted for reconsenting in future studies, and this would depend on whether the new research is for good cause. Sharing data with external partners is welcome in some contexts but some research participants did not trust foreign researchers.

Conclusion
With these varying ethical considerations, we recommend that stakeholders, including research ethics committees, work together to adapt and use clearly de ned ethical frameworks, guidelines, and policy documents to harmonize the establishment and running of biobanking and genomic research in Africa.

Background
The era of high throughput sequencing technologies and the rapid growth in bioinformatic algorithms for the manipulation of genomic data has brought with it critical issues in bioethics that are worth considering in the acquisition, biobanking and analysis of genomic data (1). This calls for clear guidelines to govern the conduct of genomic research and the use of genomic data. Genomic data is a critical resource for the development of novel therapeutics (2). Sharing genomic data has become imperative for researchers, especially where family data with third parties are concerned For instance, cancer data needs to be shared to fast-track the search for novel therapeutics (3). There are serious concerns when it comes to sharing genomic data as it carries more information about the participants genealogy and associated risk factors to some diseases (4).
There are important ethical dilemmas when it comes to the genomic research in Africa, with seminal discourse on genomics and ethics in Africa (5). Participants in Africa must be informed of the use of their data and information generated from such research shared with research participants and their communities. In tackling ethical issues that impinge on genomics research, several attempts have been made at developing robust and carefully thought-out strategies in guiding the informed consent process in genomic research in Africa (6). Broad consent has been at the center of discussions on ethics in genomic research in Africa (6). Broad consent would provide ideal grounds for futuristic analyses of genome data to answer newer questions as they emerge. Although this might be debatable ethically, it assists in overcoming several bottlenecks that may arise at the population level interpretation. Experts in the Africa need to lead the development and implementation of ethical guidelines that govern such details of genomic research, including setting up biobanking and the use of genomic data in future as well as provide platforms for the continuous education on biogenomic research.
Our study therefore sought to describe the existing ethical considerations for biobanking and genomics research and data use in Africa. Our focus for this review relates to ethical considerations for biobanking and use and sharing of genomic data or stored specimen data generated from work in Africa. This includes knowledge, acceptance, and ethics of biobanking, generation of genomic data, from enrolment of participants (including consenting), sample collection, storage and transport, analysis of the data throughout the entire research process and even beyond for the length of time that the data is archived.

Methods
We conducted a narrative review following the preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 checklist (7).

Information sources and search
A medical librarian (DV), trained in systematic reviews, conducted the literature search on September 30, 2020 PubMed/Medline, Embase (Ovid), Cochrane Library (Ovid), Global Health (Ovid), APA PsycInfo (Ovid), Cumulative Index to Nursing and Allied Health (CINAHL), Web of Science, Biosis Citation Index (BCI), and Scielo/Lilacs were the databases searched for bibliographic citations. The World Health Organization and Google were searched for relevant grey literature not found in bibliographic databases. The following journals were searched electronically to ensure articles were not missed through the database search: Bioethics, Bio-preservation & Biobanking, BMC Genomics, BMD Medical Genomics, BMC Medical Ethics, Developing World Bioethics, Genomic Medicine, Human Genomics, and the Journal of Medical Ethics. Search terms included the following keywords: 'biobank', 'biobanking', 'biological specimen banks', 'biomedical research', 'specimen handling', 'genomics', 'research ethics', 'ethics', 'ethical', 'ethics research', 'research ethics committee' 'Africa', and 'African'. The detailed search strategy is in Additional le S1. All citations were managed through EndNote and uploaded to Rayyan for systematic review management.

Eligibility criteria
All primary research study designs were eligible for inclusion, including experimental and non-experimental studies. We included studies employing both quantitative and qualitative evidence from peer reviewed journals, spanning a maximum of 20 years (2000-2020). Except for project protocols, reviews, commentaries and reports/conference proceedings, all study designs were included. It focused on research work conducted in Africa, even if data was stored or analysed outside the region. Included articles were all peer-reviewed, written in English and contained the pre-de ned domain and determinant and were primary data. Articles in other language but whose abstracts were available also in English were also reviewed for eligibility. Articles were excluded when they did not match the domains and the determinant we de ned or were reports of proceedings or secondary analysis. Reviews and editorials were excluded, but individual studies identi ed in such reviews and editorials were assessed for their relevance and eligibility.

Study selection
All duplicates were manually removed using Endnote. Screening based on title and abstract was done independently by three reviewers for the database searches (MAC, JA and DV). Any discrepancies or disagreements between the reviewers were discussed amongst reviewers until a consensus was reached. Where necessary, full text was assessed for clarity. The authors had access to full text of all included papers.
Data extraction and synthesis MAC and JA independently conducted data extraction from the included papers, with no blinding to the journal or author details, using a standardised data extraction form based on PRISMA-P guidelines. First, data on the overview of the characteristics of the included studies was extracted. This included data on the variables author (year), study design, setting (country, population, sample size), study objective, speci c genomic issue studied (general biobanking, genomic data), speci c body tissue(s) mentioned, speci c ethical issue studied, and main ndings were extracted (Tables 1 & 2).
The quality assessment was conducted by MAC and JA using the mixed methods appraisal tool (MMAT) version 2018. The tool was used because the included studies employed quantitative, qualitative, and mixed methods. Quality of qualitative, quantitative, and mixed methods papers were assessed separately using the relevant sections of the tool.

Overview of Included Studies
A total of 3,517 papers were identi ed in the database and grey literature searches. After removal of duplicates using Endnote, 2,535 articles remained and were screened by title and abstract. This resulted in exclusion of 2,461 articles, leaving 74 articles to be screened for eligibility. Another 62 articles were further excluded due to the following reasons: eight duplicates, 44 reviews, reports, or commentary, one study protocol, and nine were not studies from Africa. A total of twelve (12) articles were included in our study, four quantitative and seven qualitative and one mixed methods study. Figure 1 illustrates the study screening and selection process. Four studies (33.3%) were conducted in South Africa, and the rest were conducted in Ghana (2), Nigeria, Gambia, Uganda, Egypt, Botswana, and Kenya.

Figure 1here
Narrative synthesis We found studies that described perspectives of researchers, previous and prospective study participants and or their relatives or care givers, on ethical issues related to biobanking and genomic data collection, usage and sharing in Africa.
We grouped our results under the following subheadings based on our ndings from the included studies, rst presenting any quantitative data, followed by the qualitative data: Biobanking -knowledge, establishment, participation, regulation, and governance Participant engagement and enrolment Preferred consent models Sample collection and storage, sample reuse Con dentiality

Return of results
Sample export and bene t sharing

200
To explore the issues of sample collection, storage, export and future use of stored biological samples. Standard operating procedures (SOPs) for sample collection and Material transfer agreements (MTAs) for shipment of samples, were developed and complied with by all sites.
The project has enormous biospecimen resource for future genomic research stored with the H3Africa biorepository, in compliance with H3Africa guidelines.

Frequent visits and effective communication channels established.
Data and biospecimen from the project are freely and widely available within the framework of safeguarding participant safety and con dentiality.

Governance of Biobank
Blood or tissue samples Informed consent, Con dentiality.

Community participation (bene ts and barriers)
There was scanty knowledge about biobanking among Egyptian patients.
Many had positive attitude towards sample donation with no religious or cultural barriers. There was no signi cant relationship with willingness to donate samples and sociodemographic factors.
Most believed there were bene ts in participation but had concerns with sharing their samples across borders or with pharmaceutical companies. They wanted their information to remain private and con dential. Law enforcement agents must have access to their data when necessary. Rumours play a key role in anxiety. Although trial recruitment was initially high in the village, some families refused screening when rumours started spreading that the trial team was taking too much blood.
Women are more vulnerable to weakness after blood donation.
Some thought blood taken for research would be donated for therapy.
They believed they had increased nutritional need after donating samples to replace lost blood.
Concerns about 'loss of blood' were equated to loss of strength and lack of good food. Loss of blood can lead to fainting and weakness which will not allow them to work on their farms. They generally had limited understanding of genetic research and were not in favour of "for-pro t" research.
They demonstrated a great deal of trust in researchers and reuse of samples. However, they indicated that certain types of re-uses were more acceptable than others. Education of subjects on genomic studies was better during community enrolment than hospital enrolment. It was di cult explaining genomic research in local languages.
Participants and staff seeking consent were less aware of the methodologies employed during genomic research and their implications, such as the breadth of data generated and the potential for future secondary research.
Community engagement processes involved fewer women, so meetings were intentionally organised for women groups. Guidelines and Research Ethics Committees (RECs) need to stratify risk accordingly and governance processes and structures must be exible.
While broad general consent was preferred, tiered consent was thought to be more consistent with respect for autonomy and building trust.
Material Transfer Agreements (MTAs) were often lacking when bio samples were exported, and this was perceived to impact negatively on trust. Authentic community engagement would help to build trust. Participants emphasised the importance of sample export, storage, and reuse, and acknowledged the existence of some structures governing these research practices.
There is the need for several practical ethical concerns to be addressed to ensure high standards of practice and to maintain public con dence in international research collaborations. These concerns relate to obtaining culturally appropriate consent for sample export and reuse, understanding cultural sensitivities around the use of blood samples, facilitating a degree of local control of samples and sustainable scienti c capacity building. Half of study participants supported use of broad consent; a quarter supported restricted consent while the remaining quarter were in favour of tiered consent.
Most support shipment of their samples to other countries for further research, but they prefer those collaborations to be done only with competent, ethical researchers and they would like to receive feedback about such projects. Participants emphasized the need to ensure that donated samples were not used for research that contradicts their religious beliefs.
Majority of respondents preferred health care as a bene t from participation, particularly for any unexpected condition that may be discovered during the research instead of nancial compensation. Most respondents were willing to share their samples with local and international researchers for future research.
They were also willing to have samples used for any research.
Barchi et al, Belief systems about bodily integrity and strong national identity in the construct of bene ts may be at odds with initiatives that involve foreign biorepositories.
There is lack of understanding among patients and providers about the use of biospecimen in clinical care and research; reuse of biospecimen, particularly issues of consent, ownership, and decision-making.
Several respondents were okay with broad consent at the time of sample collection but believe individuals do not fully understand such choice.
Others had concerns about export of specimens and loss of control over reuse and potential bene ts were also areas of concern; and felt need for regulatory guidance and IRB-member training.

Moodley et al, (2014)
Use, storage, and export of biosamples Blood Informed Consent, Con dentiality, Ownership Export, Most participants were supportive of providing samples for research, but had serious concerns about future use, bene t sharing and export of samples.
While researchers view the provision of biosamples as a donation, participants believe that they still have ownership rights and are therefore in favour of bene t sharing.
Almost half of the participants expressed a desire to be recontacted for consent for future use of their samples.
25% of participants had concerns about sample export. Tables 1 & 2Here Biobanking -knowledge, establishment, participation, regulation, and governance Knowledge on biobanking was associated with higher education and maleness (8), and there was di culty in explaining genomics in local languages (9). In one study, researchers found innovative ways to explain essential parts of the term genomics, with parents relating genetics to hereditary characteristics, making it easier to understand (10). In another study, participants recommended strengthening understanding of biological research with more engagement of the community, for example, organizing an open day for community to visit laboratories to see the conduct of research. This will help increase knowledge, awareness, and participation in genomic research (11).
The role of biospecimens in clinical research is often misunderstood. Giving samples for clinical care is considered an act for one's own wellbeing -leading to treatment for cure (12). When participants provide biospecimen in the hospital context, they do so, expecting a cure for their condition. Some believe that collection of some special specimen e.g. spinal uid, worsens health and may hasten death. Due to some of these beliefs, during extraction or amputation, patients ask for their limbs or tooth to take home.
Central to setting up biobanks is the role of review and ethics committees (RECs) (10,13,14). Most participants in one study recommended the need to have discussions about the process of reviewing research applications (10). In one study that assessed the competence of research ethics committees (REC) to review biobank projects, researchers distinguished between study speci c biobank and biorepository, and noted that RECs must understand heterogeneity of biobanks. It emphasized the need for RECs to stratify their reviews according to risks related to volume and types of specimens since this is important for decision-making and governance of biobanks (14). The study mentioned that REC members sometimes lack the expertise to review protocols expertly, and that national audit of biobanks and their governance structures is required. International standards and best practices should be followed during biobanking as was done in the SIREN project (10). Some of these best practices were but not limited to, obtaining individual consent from each participant, development of standard operating procedures (SOPs) for managing samples which were adopted by laboratories at all sites, standardization of sample collection and processing, periodic visits to study sites, effective communication, having material transfer agreements (MTA) for all sample shipments, development of guidelines for accessing biospecimen, hands-on training for staff, regular consultations and consortium meetings, quality control/ assurance measures and use of laboratory information management systems.
With respect to participation in biobanks or genomic research, one study reported 89% (352/396) of respondents believed that study participants have rights with respect to what is done with their tissues, and that these should apply even to anonymized samples (15). Another study reported between 79% (202/259) -85% (221/259) of participants willing to donate samples for biobank, some on condition that samples are anonymized, and 53% (151/259) were willing to donate speci cally for genetic related research (8). However, some papers reported refusal of participation or challenges in getting participants (10)(11)(12)16). In one study, as many as 62.8% (149/237) potential participants could not be screened for inclusion in a study due to lack of consent by heads of households (16). Some reasons given for refusal to participate include the fact that blood-taking depletes life-force and body strength which affects health of participants and their capacity to work. Another reason was that it is dangerous for pregnant women and women in general to give blood since they are vulnerable. Some were unable to distinguish between blood sampling and blood donation, believing that some of the sampled blood will be donated or sold to others. Others feared that more blood will be taken than is needed, and that taking blood is associated with hospitalization, with subsequent need for transfusion which has nancial implications for them. Other challenges were identi ed as barriers to participation.
Barchi reports that cultural traditions, norms, and beliefs around human tissue and its meaning compelled participants to compare donation of human biological material to practice of male circumcision, in which case the tissue may be sold to others for use. Others also believe that their samples may be stolen by others from the health workers for magic to harm or bring bene t to others (12). One study also identi ed local cultural sensitivities around the use of blood samples, with apprehension about blood more than other samples (such as urine, stool, saliva), and export and storage of the samples. Concerns such as pain for children and the volume of sample causing harm to sick children and making them weaker were also reported. There were also issues of mistrust with rumours of researchers selling some of the blood (due to the idea of transport of samples and transfusion) (11).
Regarding operational management and sustainability of biobanks, some studies reported challenges (10,14). Researchers mentioned RECs having issues with their objectives, informed consent and documentation as well as the need for additional ethical reviews for new studies or lack of plans to deal with community harm/ bene ts, as challenges (10). One study highlights infrastructural and security needs, especially power interruption which affects freezers, the quality of samples and retrieval systems (14). Lost samples are a violation of promises made to participants. Sustainability of biobanks requires stable and continuous funding to avoid wasted samples which undermines the trust of participants. The study also mentions that multiple levels of governance are imperative in biobanking because biobanks usually involve different institutions and multi-tiered governance systems with varying legal and policy frameworks (14). In one study, some participants (researchers) were critical of existing regulatory systems, indicating that they are not clear on biobanking, do not have a proper de nition of tissue, are often written for therapeutic biobanks and not research biobanks (11). The study discussed the importance of local capacity building and effective research governance, recognizing the point that while international collaboration is important for scienti c research, it works well on mutual trust, transparency and respect and scienti c leadership (11). The use of more protective measures and assurances that allay concerns are needed. Capacity building should be focused on technology and infrastructure training and retaining local personnel with requisite skills to contribute to the conduct analysis and reporting of research locally. The paper concludes that effective research governance structures must be central to the gatekeeping role of RECs, and, institutional and national guidelines should be in place to govern research practice, with RECs serving as Trustees of research samples.

Engagement and enrolment of study participants
Two studies identi ed the importance of stakeholder (community and patient) engagement for biobanking and genomic research (9,14). One discusses a model of engagement based on traditional practices that had been established in their study community and followed by all researchers. This involved consultations with the gatekeepers of the community (chiefs and elders), followed by community durbars with the wider community (9). Enrolment of cases in the hospital also involved a two-step process of consenting, rst at the time of admission, and subsequently a second consent was sought based on eligibility criteria. Enrolment of community participants involves a traditional multilevel process which engages heads of compound and household to seek their permission, and then parental consent. The second study also mentions community engagement as a priority to build trust and encourages that it should be undertaken extensively using community newspapers, educational videos advocacy groups that represent trust, and encouraging community feedback (14).

Preferred consent models
One quantitative study reported on who should give consent in studies involving children (15). In the study which had researchers as respondents, 84% (317/396) of them agreed that parental consent is enough to store child genetic sample for children who cannot assent because they do not understand the nature of the research, while 92% (364/396) thought once the child understands he/she should give assent. A total of 45% (178/396) of participants believed children between 16-18 years, can understand implications of storage of samples for future research, and that they should have the right to withdraw from the study once they reach age 18 years.
On type of consenting, different models were preferred from the included studies. In one quantitative study, 66% (262/396) of participants preferred broad consent, and this was irrespective of whether participants had ever participated in biobank research (15). A couple of qualitative studies show preference for one-time broad consent (14,17,18). One such study which describes broad, tiered, and dynamic consent models, indicated that most researchers preferred one-time broad consent because participants saw reconsenting as impractical and resource intensive, and some found detailed consenting not to be sustainable for clinician researchers (14). Community researchers, however, found broad consenting unacceptable to their clients and prefer reconsenting whenever necessary. A qualitative study reported that half of respondents supported broad consent, saying they donated samples because they had faith in the research (18). One-time consent was also perceived as good if all future re-use is listed from the onset, otherwise previously undescribed use should seek reconsent (17). In a study in which 58% (230/396) of participants did not think reconsent is necessary, reconsent was deemed necessary in a variety of scenarios such as investigating unrelated condition (65%, 257/396)), or when researchers want to add other genetic measures to study (67%, 267/396) (15).
Another listed that although blanket consent may be given, participants may not fully appreciate the risks and potential bene ts, especially from vulnerable populations (12). Consent for reuse is mostly preferred, after IRBs approve the reuse.
In one study, researchers found it di cult explaining future uses of blood at the point of collection, and acknowledged that requirements of full disclosure cannot be assured in broad consent (11). It argues that though reconsenting is di cult, it should be done through community engagement, moving away from blanket consent since it limits acceptability of research and defers future consenting to local ethics committee. This is corroborated in another study where most participants (82%, 324/396) thought RECs approval was enough if re-consent was not possible (15). The participants were evenly split on whether reconsent is necessary to share de-identi ed samples with another investigator. A tiered consent model was described to include speci c and broad consent which is enabled and facilitated by technological advancement, while dynamic consent and participation, facilitated by constant information sharing through technology was also preferred by some in one study (15,17). One study found 49.5% (99/200) of participants wanted to be contacted for future use of their samples, even if an IRB approves the new use. They didn't think RECs can consent on their behalf. The rest were comfortable with broad consent to allow reuse of samples (13).
Researchers were also concerned about the actual information on the informed consent form (ICF), saying it is researcher and legal -focused but may not be important for the patient/ community. One study describes timing of consent, with parents satis ed with timing of consent and two-steps process of consenting (9).
Sample ownership and storage, sample reuse Four papers described considerations around storage of biospecimen (12,13,17,19). In one quantitative study, 95% (335/353) of study participants were willing to have their child's sample stored with a code linked to a patient identi er (19). In two studies, most study participants were unconcerned about storage of their samples for future use, believing the samples were no longer theirs once it's given (13,17). Only a few believed they still were owners of the samples, with 12% (24/200) wanting a reason to grant permission of sample storage (13). Participants wanted their samples to be stored securely and used for the purpose for which they were taken, being accessible only to research staff (17). In one study, many participants felt that specimen belonged to donors, who should retain rights to them including storage, and retrieval if needed, possibly with gradual transfer of ownership to researchers (12). Some had concerns about storing specimen beyond what is used for care; nding it worrisome that some sample is kept for another purpose. The study recommends the need for a regulatory framework to safeguard storage and duration of same, and the need to build capacity in-country for storage, safekeeping, and research, recommending that specimens belong to Botswana/government and should not be exported (12).
Multiple uses of biospecimen and data is common with genomic research and some studies reported on use and sharing of the biogenetic specimen. In one quantitative study, few (4%, 14/353) participants feared researchers might sell their samples or use them for other nonresearch related purposes (19). Most (97%, 343/353) wanted to know about future studies prospectively, 85% (300/353) were willing for their samples to be used for HIV studies and 81% for any disease. The majority (95%, 335/353) were willing to share their samples with researchers in Kenya and Tanzania as well as UK and USA. Living in peri urban areas was associated with being more likely to believe study samples would be used for research purposes only and wanting information about studies (19). A study in Ghana found that the community does not appreciate data sharing, and that having a policy allows data to be shared with external researchers (9). Sharing and reuse of data or samples were acceptable by participants if there was a clear data release policy, if new research was for a good cause or would come up with new health solutions and if revenue generated from the research will be shared with them (9). Good cause refers to current and future community bene ts from the study, academic and institutional bene t, and career advancement. Commercial purposes were not considered as good cause (17).
In respect of access to stored samples, priority should be given to local researchers to access samples to bene t Botswana. Some families wanted their samples to be kept identi able so that they can reconsent when needed for reuse. It should also be possible to link specimen to new discoveries relevant to health. Some international need for reuse may not align with local needs (13).

Con dentiality
One quantitative study reported that most participants (91%) in biobanks want researchers to maintain privacy and con dentiality of donor information (8), with 70% believing this will be done. More than half (64%, 166/259) believe data collected will not be used for other reasons without their consent, with 72% believing that law enforcement agencies can have access to their data when necessary.

Return of results
Two studies reported on return of results to study participants. Fifty-ve percent (142/259) of participants in one study wanted results of tests conducted on their samples to be put in their medical records and 93% (240/259) of them want to be contacted if their results show any risk (8). In another study, the majority wanted to be informed about their individual results to know their health status as well as bene t from new discovery. Feedback to study participants should be preceded by counselling. A few did not want to know test results because of fear (18).

Sample export and bene t sharing
Some studies reported on concern over sample exports and who eventually gains from these samples (11,13,18). Some challenges were the loss of control over use of samples and data once transferred, local researchers' inability to account for exported or shared samples or data, possible use for other purposes and analysis unknown to local researchers, fear of use for rituals, lack of recognition of authorship of local researchers in future work using samples, and others taking credit without acknowledging researchers or community. Participants did not want their samples to be taken to the UK, USA, Europe, and Israel which is an enemy of Muslims (13). Due to concerns about sample exports, local researchers should ensure local control of samples during and after transfer and scienti c collaborations should be supported by mutual agreements, including material transfer agreements (11). Participants (laymen) in one study were willing to share their specimen and data with commercial and non-commercial entities (18). A few wanted to be contacted by any new researchers, concerned that if samples were sent outside Nigeria, ndings will be used to discriminate against Nigerians or used for something against their religion. Also, collaboration should be with only competent institutions/researchers and feedback would be required (18). In a qualitative study, strong opinions were expressed on export and material transfer agreements (14). In this study, researchers insisted that patients must be told at the onset that their samples will be sent abroad with contracts. It is important to prevent unilateral transfer of samples out of Africa and to retain intellectual rights here. The lack of a national MTAs and export permits was of concern. Local collaborators and participants exhibit a lot of trust, but sometimes external partners are not respectful and are surprised that a MTA is asked for. The H3 Africa project biobank will provide specimens to researchers anywhere with an approved protocol. Movement of samples within Africa could also be problematic as some countries do not have MTAs (14). In the study in Ghana, research ethics committees were also concerned about exported samples due to inability to control what is transported (11). MTAs are now coming up, and projects should build local capacity to analyse samples to reduce export. Possibly a local research institution should be established to take responsibility and update research ethics committees on overseas analysis, to make external partners accountable.
Participants also believe that research collaboration should be with only competent institutions/researchers, providing feedback (18).
In one study, 25% (64/259) of participants believed that donors should be compensated nancially (8). In the same study 50% (130/259) believed the sample does not belong to donor once it is given. Another study assessed the risk-bene t ratios of biobanks and emphasized that individual bene t is distinguished from public health bene t (14). Most participants were con dent about scienti c and clinical bene ts of biobanks but felt individual potential bene ts are lost due to anonymization of study samples. Long and short-term bene ts must also be distinguished, realizing that biobanking is for future bene t and must be done in such a way that retrieving data is not jeopardized. Some risks identi ed included over-researched communities, with different project groups competing for the same participants in the community; commercialization where samples are sold to pharmaceutical companies; infectious disease samples being a risk to biobanks and patients and so the need for robust infection control. Stigma associated with genetics and genomics making participation by some patients di cult was also mentioned. In a Ghanaian study, participants believed that research is for the bene t of participants, and that participants unmet needs and parents' expectations of free medical care for their children is a motivation for their participation in such research, and other studies in the community had offered such bene ts (9).
On potential bene ts and bene t sharing, IRB members in Botswana believe subjects should continue to be paid for their time in the study, and any bene ts from their stored samples -monetary, intellectual property rights, new treatment -should be shared with the participants, communities, and the nation at large (12). They were worried that if samples are stored outside, they would be denied these bene ts. They also agreed that not adequately acknowledging research subjects and lack of feedback to subjects (including results), affects others' participation in research later. Some participants (40%, 79/200) mind if pro t is generated from work with their samples and 43% of these want to share in the pro t. Some 20% (39/200) do not mind if pro t is generated for a good cause (13).

Quality of Included Papers
In all, we included 12 papers, comprising of seven (7) qualitative studies, four (4) cross sectional studies and one (1) mixed methods study.
For the qualitative studies, all except two of them had clear research questions and the collected data addressed the questions; for all studies, the qualitative approaches and the data collection methods employed were appropriate to answer the questions. The results, and interpretations of same, were adequately derived from the data collected, and there was coherence between the data sources, collection, analysis, and interpretation.
For the cross-sectional studies, 75% of the studies had clear research questions. For all the studies, the collected data addressed the questions, and the sampling strategy was relevant to address the question. Three-quarters of the studies had a representative sample, and all studies used appropriate measurements. The risk of non-response bias for all the studies was not low for 50% of studies and unclear for one study. Statistical analysis was appropriate for all the studies.
The only mixed methods paper did not have a clear research question, and thus it is not clear if the data collected addresses the question(s). Table 3 provides details of the quality assessment.

Discussion
Our synthesis of the twelve included studies focused on the ethical considerations of biobanking and genomics research and data use in Africa. This is an important area that needs more exploration because of the paucity of such research in the sub-region and the strong in uence of external collaborators in the eld. Ethical issues surrounding research and speci cally biobanking and genomic research are crucial since informed participation as well as human subject protection must always be assured (20). All stakeholders in the process must bring their perspectives and experiences to bear in progressively shaping the course of how future research are set up and implemented.
The ethical considerations gleaned in this review are corroborated ndings in a previous review (21), and highlight the need to increase awareness and knowledge about biobanking especially in the context of Africa where there many myths surrounding other people having access to one's biological tissue. Potential study participants need to be well informed about the purpose of biobank-based genomic studies so that their participation will be well informed. Stakeholder engagement and proper community entry that leverages the existing governance structures of communities and their cultural practices have been found to be appropriate and makes communities more receptive when these are adequately carried out with prior planning (5,10,14,(22)(23)(24). Community durbars for example, allow for wider participation and afford researchers the opportunity to explain the research to the community, eliminate myths and doubts about the usefulness of genomic research and biobanking to the community, so that they are more likely to partake in such research.
Biobank-based genomic studies should be regulated by both local and national regulatory mechanisms that ensure that participant autonomy, privacy and safety are not compromised in any way. Research ethics committees are very central to this process. They are expected to be adequately trained and have expert knowledge to make them effective in addressing all the potential ethical issues that could arise with different studies and prescribe guidance to researchers (14). They should also be positioned to monitor such research in a consistent manner that will assure public con dence in them as well as for biobanks to serve their intended purposes. There is evidence that regulation of biobanking in the African context is weak, often allowing unethical practices (25). National level biobanks and genomic data governance structures, that are well equipped to ensure periodic audits of biobanks and biorepositories are required, especially due to international collaboration usually involved in genomic studies in Africa. These governance structures should provide the required oversight responsibility and ensure that biobanks and genomic data are not misused. This strategy would by and large build con dence in research participants and their communities to enable them trust that their samples would be used for their intended purposes.
On the issue of preferred consent models, several models were spelt out by various studies. However, broad consent was arguably the most preferred, with studies reporting the need to be supplemented with reconsenting wherever possible (15). Where reconsenting would be impossible, RECs approval su ces. The choice for broad consent is based on trust imposed in researchers and possibly RECs to protect participants' interest throughout the project. Speci c future use of biospecimen may not be known by researchers and may be di cult to explain to participants at onset of study. Participants however, had opinions about what a good cause is for which in future their samples can be used. It is important that RECs and researchers work in sync to ensure that biospecimen are used for approved research purposes, in order not to jeopardize community trust.
Research results, especially biogenomic data often contains personalized data and issues of con dentiality and return of results arise. Our review shows that participants care about their privacy and trust researchers to ensure con dentiality of their personal information (8). Any identi ed risk should be communicated to them in the best way possible. Taken together, it behooves researchers to maintain con dentiality and act responsibly and be judicious in sharing information on risk factors revealed from biobanks and genomic data.
The existence of international collaboration and partnerships in research involving biobanks and genomic data requires that data is sometimes shared with external partners. Ethics committees should have clear guidance for researchers regrading this process. Research protocols should at the very onset have plans on MTAs, allowing study participants to be informed if their samples will be shared with other countries, and how samples would be used by collaborators. This will allay the fears of samples being used for other ritual purposes in foreign countries (14). Unfortunately, as reported in one study, sometimes RECs have expressed worries about their inability to control the use of exported samples once they leave the sample origin country (11). Irrespective of well laid out MTAs, some collaborators were reported to have used samples from biobanks and genomic data for other purposes without recourse to the sample origin RECs. This may not be a disadvantage to only participants but also to the collaborating researchers. Data and results generated by such research including adverse ndings on risk factors, may not reach the study participants and communities. The researchers in the local context may also lose career bene ts that should have accrued to them since they have been involved in the setting up of the original studies. Often, one reason for data export is lack of local capacity to analyze the samples, although African biorepositories have been attested to be able to collect, process, store, and ship biospecimens of good quality (25). One way of addressing this issue, beyond strict enforcement of MTAs, will be to improve the capacity of researchers in Africa and provide the necessary logistics including reagents and equipment, so that most analysis of biospecimens would be done locally (26). Other times the export is necessitated because the partner institutions also have their own larger biorepositories (27).
Biobanks and genomic data may provide public bene ts as information obtained from these studies could be bene cial to pharmaceutical companies. However, there may not be direct individual bene ts. Issues of individual compensation have been encountered though most individual participants advocate the public good (8).
Unfortunately, although one scoping review found as many as thirty-six guidelines or policy documents for biobanking or genomic studies within the African context, our included studies did not report participants referring to any of these guidelines (28). Such frameworks must necessarily empower local scientists based in Africa to spearhead genomic research and biobanking in the jurisdiction. Continuous capacity building in bioethics, data analyses and bioinformatics, with these guidelines as references, would be required. National RECs must also work together in this endeavor of harmonizing ethical reviews at all levels and across institutions or organizations. This will help to harmonize existing guidelines or policy documents. In some settings, it is commonly known that even at national levels RECs work independently of each other, with sometimes researchers required to get approvals from multiple RECs before they carry out their work.
There is an urgent need to establish robust ethics frameworks and governance systems that will ensure that biobanking and genomic research in Africa is anchored on systems that allow participants derive optimum bene ts from our shared data. A best practice guideline such as the H3 Africa ethical framework on biobanking and genomic research is highly recommended (25). It draws on existing policy documents and empowers African researchers and communities, educating them on their rights and demand greater control over sample collection, storage and usage, and also deals with rules of engagement for collaborating and funding non-African institutions that they work with (29).

Strengths and Limitations
This narrative synthesis reviewed literature from a wide range of databases using a systematic approach, and we believe it provides comprehensive data from relevant publications based on our scope. We however envisage some methodological limitations with our work. First, we admit that the African region produces other language publications such as French and Portuguese, and our restriction to publications in English language may be a limitation. We however believe our search was very comprehensive with inclusion of current literature through up to 2020. Lastly due to the nature of biobanking and the evident lack of knowledge about its signi cance among communities, it is possible that participants in these studies may not represent the true populations, with some segments of the populations being under-reported in studies.

Conclusion
Biobanking and genomic studies are a real need in Africa and are increasing numbers, despite the poor knowledge levels of communities on the subject. Related to this are ethical considerations related to setting up and participation in biobanks as well as data use and sharing. Research ethics committees, researchers (including international collaborators), communities and individuals have roles to play in the areas of human subject protection and con dentiality, education of participants, securing appropriate informed consent, regulation and control of data storage, usage, export of biospecimen and data sharing especially for future studies. Expertise of RECs in Africa should be built for e cient regulation and governance of biobanks and genomic research. More stakeholder engagement is recommended to clearly de ne the way forward in addressing these ethical issues through a best practice guideline that is acceptable to most if not all stakeholders.

Competing interests
The authors declare they have no competing interests with respect to this manuscript.