The IHAT-GUT Iron Supplementation Trial in Rural Gambia: Barriers, Facilitators, and Benefits

Introduction: In most sub-Saharan African countries iron deficiency anaemia remains highly prevalent in children and this has not changed in the last 25 years. Supplementation with iron hydroxide adipate tartrate (IHAT) was being investigated in anaemic children in a phase two clinical trial (termed IHAT-GUT), conducted at the Medical Research Council Unit the Gambia at the London School of Hygiene and Tropical Medicine (LSHTM) (abbreviated as MRCG hereof). This qualitative study aimed to explore the personal perceptions of the trial staff in relation to conducting a clinical trial in such settings in order to highlight the health system specific needs and strengths in the rural, resource-poor setting of the Upper River Region in the Gambia. Methods: Individual interviews (n = 17) were conducted with local trial staff of the IHAT-GUT trial. Data were analysed using inductive thematic analysis. Results: Potential barriers and facilitators to conducting this clinical trial were identified at the patient, staff, and trial management levels. Several challenges, such as the rural location and cultural context, were identified but noted as not being long-term inhibitors. Participants believed the facilitators and benefits outnumbered the barriers, and included the impact on education and healthcare, the ambitious and knowledgeable locally recruited staff, and the local partnership. Conclusions: While facilitators and barriers were identified to conducting this clinical trial in a rural, resource-poor setting, the overall impact was perceived as beneficial, and this study is a useful example of community involvement and partnership for further health improvement programs. To effectively implement a nutrition intervention, the local health systems and context must be carefully considered through qualitative research beforehand.


INTRODUCTION Iron Deficiency and the IHAT-GUT Trial in The Gambia
At any given moment, more individuals suffer from iron deficiency (ID) than any other health problem, with an estimated 1.24 billion affected individuals worldwide 1 . ID is associated with multiple pathologies, including anaemia and defective organ function 2 . The prevalence of anaemia is five times higher in low-and middle-income countries (LMICs) than high-income countries, with ~30% of the world's population, and 43% of 6-59 months old children, being anaemic 1 3 . In Sub-Saharan Africa, 79% of children under six years are anaemic and iron deficiency anaemia (IDA) affects 58% of pre-school children 4 . As such, IDA is the largest international nutritional deficiency disorder and one of the five leading causes of global disease burden 1 .
Despite widespread supplementation schemes, ID prevalence has not changed much in LMICs over the last 25 years 5 . There is growing interest in developing novel nano-iron compounds or delivery systems for fortification and supplementation 2 6-9 . One proposed strategy is a targeted-release nano-iron formulation 10 . Iron hydroxide adipate tartrate (IHAT) and standard-of-care ferrous sulphate were tested in a randomised placebo-controlled double-blind clinical trial (acronym IHAT-GUT) conducted at the MRC Unit The Gambia at LSHTM (MRCG) 11 .

Nutrition Interventions in Varying Contexts and the Need for Qualitative Data
While integrating nutrition-specific interventions into health systems can be impactful for both health and nutrition outcomes, different countries will have specific delivery needs for implementation 12 . Until the barriers and facilitators of nutrition intervention trials are studied across various settings, there will be a lack of data to implement such interventions.
Additionally, LMICs remain underrepresented in research 13 . The Global Forum of Health Research termed the "10/90 gap" to exemplify that less than 10% of health research funds go towards problems affecting 90% of the population worldwide 14 , with a smaller percentage towards LMICs 15 16 . Evidently, clinical research is skewed, with more than 80% of clinical trials occurring in high-income countries [17][18][19] and only ~1% of drugs produced between 1975-2004 addressed LMIC issues 20 with research enhancement in LMICs being an efficient and beneficial way to correct this gap 14 .

Clinical Nutrition Trials in LMICs
Given the state of healthcare and disease prevalence in LMICs, clinical trials are well received and participant recruitment is often easier than a health secure country 13 21 . The ethical argument remains that medicines targeting conditions highly prevalent in LMICs should be tested in those populations without transferring data from high-income settings.
Interventions often have the largest impact medically in LMICs and benefit from involving local staff [22][23][24] .
Difficulties of conducting clinical trials in LMICs stem from limitations in obtaining informed consent, ethical compensation mechanisms, poor health infrastructure, socio-economic and cultural differences 21 , and lack of education amongst study participants 25 26 . Additional barriers are limited research governance, funding, logistics, commercial ability, infrastructure, research materials, overall research capacity, and unsupportive administrative and government systems 15 16 25 27-29 40 . A recent review found that, service delivery and health workforce were well-integrated, but governance, information systems, finance and supplies and technology were less well-integrated 12 .
The aims of this study was to: qualitatively explore IHAT-GUT trial staff perceptions of barriers and facilitators to conducting this clinical trial to highlight the health system specific needs and strengths in a rural and resource-poor settings.

IHAT-GUT
IHAT-GUT was conducted on children with anaemia between the age of six and 35 months, living in The Upper River Region (URR) of The Gambia 30 . The children were enrolled in the trial for 113 days, within which they underwent supplementation for 85 days, with weekly study visits to test haemoglobin (Hb) levels and malaria status, and three study timepoints included venous blood collection. Further information about IHAT-GUT study protocol is provided in the protocol paper 30 .

Study Setting
We describe the involvement of research staff in the iron supplementation trial IHAT- GUT 30 which was under the governance of MRCG (Figure 1). The Gambia is the smallest and most densely populated country in West Africa, with about 2.28 million inhabitants, of which roughly one million (48.6%) live below the national poverty line 31 . Islam is the predominant religion, polygamy is widely practiced, and families live in multigenerational compounds within villages 31 . The Gambia is subject to bimodal weather conditions having a "wet" (June All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted September 30, 2020. . to October) and "dry" (November to May) season. The seasonal rain determines the farming practices at that time of year, lending to extreme variations in seasonal diets and fluctuating levels of malnutrition 32 .

Study Participants
Individual-interviews were conducted with 17 IHAT-GUT local trial staff (Table 1). Trial staff were purposively sampled to ensure insights from varying job types. To maintain anonymity, participants are identified via job title only. was developed and used to facilitate discussions. Interviews, with informed consent, were audio recorded and later transcribed verbatim. Photographs were taken during data collection with informed consent. Interviews continued until no additional insights were gained (i.e. data saturation 33 ).

Data Analysis
Transcripts were analysed using inductive thematic analysis to explore patterns in the data 34 . Braun and Clarke's six phases of thematic analysis were used ( Table 2) 34 . The transcripts were read several times (data familiarisation). Notes were made in relation to significant/interesting comments made by interviewees. The transcripts were coded and memos written. A code represented a feature of the data that the researcher found interesting and the memo was a summary of the interesting findings 34 . A list of codes was constructed and connections between them sought to develop provisional themes (repeated patterned responses within data sets) and sub-themes 34 . When all transcripts were analysed, a final list of themes and sub-themes was created.
An inductive approach was used whereby data analysis was data driven so that participant's views took precedence over the interviewer's previous knowledge or beliefs 34 . That said, previous knowledge will, to some extent, influence the research. Therefore, to ensure rigour, transcripts were analysed by the second author (LMD) using the same procedure.
Discrepancies were discussed and jointly altered. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Ethical Considerations
Ethical approval was obtained from The Gambia Government/MRC Joint Ethics Committee (REF: L2018.25). Written and verbal informed consent for interviews and photographs was obtained from each participant. It was made clear that refusal to be involved would be confidential and would not affect their work; no participants refused to be interviewed.

RESULTS
A distinction was made between barriers and facilitators at three levels: study participant, trial-staff, and trial-management.

Barriers
The barriers included community factors and low incentivisation (participant level); motivation (staff level); and country context (trial-management level) ( Figure 2). Illustrative quotes from each sub-theme of the barriers can be found in Tables 3.

Community Factors
Many viewed specific cultural factors as barriers to successful running of this trials, specifically community hierarchies and education. These issues posed challenges with community sensitisation for the staff, but it was noted that these barriers were easily overcome if approached properly. Regarding hierarchies, the Local PI noted that to run the trial fluidly it was key to: "Pass the messages smoothly, so everyone can understand. If your

Low Incentivisation
State Enrolled Nurse 2 maintained that offering free medication and healthcare was an incentive to take part in the trial, but moving forward researchers need to consider incentives beyond the child's healthcare: "When we do the venous bleeding, we give them [the mothers] 50 Dalasis [to buy breakfast], but I would love for it to be more than that. We take blood from these children, so we need to make life easy for them. Maybe they are on the drug that doesn't do anything. The mothers sacrifice a lot for this project, so we need to give them something back". Staff All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Motivation
As summarised by the Research Clinician, there is always challenges when working within teams: "There will always be challenges. That you should expect. The biggest challenge is working with individuals and managing individuals. Everyone has negative qualities, I have them". Workload was also noted as a barrier by both the laboratory and data teams, as those departments "tend to be oversubscribed" (Statistician). The Analytical Project Manager commented that the workload became unmanageable for the laboratory staff at some points.

Staff motivation was difficult for those with managerial responsibilities (See figure in
Supplementary Data: S1): "Although this is not the first time I've manned a group like this, it's not easy" (Field Coordinator; managed thirty employees).

Country Context
The remote setting and harsh weather conditions required excess planning and coordination: "Resources are always a problem in developing countries" (Statistician). preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Worker noted funding needs such as giving "[phone] credit to call for the Field Workers and mothers" so they could afford to make trial related calls.

Facilitators & Benefits
The facilitators included healthcare, incentivisation and receptive communities (participant level); staff characteristics and education enhancement (staff level); and the local partnership (trial-management level) (Figure 4). Illustrative quotes from each sub-theme of the facilitators can be found in Tables 4.

Healthcare
The study was conducted in the North Bank of the URR where the poorest Gambian communities reside and the double-burden of malnutrition and infection is highest 35 . All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Another advantage for trial participants was that healthcare was fully covered during the trial.
The Nurse Coordinator found that local communities embraced these interventions: "These studies excite them. They want them to continue. We deal with the health, the management, the transportation."

Incentivisation
The Data Manager mentioned that: "You need to bring a social impact, so the participants feel valued [rather] than just coming to do what you want and not giving the mothers and children something". A successful incentive was the provision of Yandi juice, which made supplement administration easier and more enjoyable for participants: "Giving the Yandi attracts the children because they want to take the drugs. When they see the Field Workers, they get excited" (State Enrolled Nurse 2). The Nurse Coordinator also mentioned how: "Sometimes the kids are running for the Fields Workers because they are excited for the juice" and advised to use this approach in the future.

Receptive Communities
Staff found the communities were receptive towards the study. The Research Clinician, who is not Gambian himself, found: "The Gambians, they are remarkable people. They are the most amazing, welcoming people. It's a very friendly environment to work in. It's a researchfriendly country". State Enrolled Nurse 2 highlighted that URR, being a low resource setting, facilitated this: "The advantage of the project being run here is due to low-income earners, so having these projects is a big deal…They are very cooperative. In an urban area it would have been harder, but in a rural area they are excited". The Data Manager echoed this sentiment and believed a similar trial in the UK would be more challenging.

Staff Characteristics
It was very helpful that the staff going into the communities to conduct research were from a similar cultural background and could speak local languages. The Senior Field Worker All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. When we go into our own communities, they are accepting". Likewise, Scientific Officer 2 was able to complete her master's because of MRCG.

Local Partnership
MRCG's research unit has been in The Gambia for 70 years, giving them respect in the country for their long-standing establishment, advances in healthcare and facilities for epidemiological studies and clinical trials. MRCG's upstanding reputation with The Gambian All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted September 30, 2020. The Analytical Project Manager was grateful for the high-quality facilities, which allowed the project to run smoothly ( Figure 5). While LMICs were noted as lacking commercial ability, infrastructure, materials and overall research capacity, elsewhere, this was not seen in the present study 15 16 27 . Running clinical trials through MRCG appeared as a facilitator, while unsupportive administrative systems were reported to be a barrier for clinical trials in LMICs 27 28 . In terms of "human capacity," a recurrent theme in this data was the experienced MRCG staff and investment in training. In contrast, Ross et al. 33 found that in developed countries, lack of staff and adequate training posed barriers. Likewise, in our study staff motivation was noted as occasionally posing a barrier, which is concurrent with the literature 27 36 .

DISCUSSION
Materials can be difficult to resource in LMICs and for projects to run smoothly, advanced planning is needed. However, a lack of resources in LMICs actually increases the need for sound research to prioritise these limitations 37 . This barrier of planning and preparation not only encompasses the materials needed, but also appropriate government approvals, which was previously noted 27 36 . MRCG's long-standing establishment and good government relations ease this potential barrier. Overall, the participants showed a clear awareness of the country contexts' barriers and facilitators, and the implications these have for the trial. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted September 30, 2020. . The overall impact for the country's healthcare can be beneficial and outweighs the barriers described. Our findings are in-line with previous research, noting the need for research in such communities 13 21 . Clinical interventions in LMICs have the largest impact in decreasing childhood mortality rates 13 21 . The high infection burden setting of rural Gambia was the ideal setting for IHAT-GUT. If the drug were to be tested in a high-income, low-infection burden country, its effects might not be translatable to the country of target, as seen in other clinical trials 38 39 .
The ease in recruiting patients was noted as a facilitator in running clinical trials in LMICs like The Gambia. LMICs offer an attractive setting for clinical trials as there are often higher incidence rates of the issue, therefore lending to shorter periods of participant recruitment 21 .
Recruitment time can be five to ten times quicker in LMICs than in the United States or Europe 40 . One participant highlighted that he did not think mothers in developed countries would be happy to enrol their child in a study, but in The Gambia the mothers are willing.
Ross et al. 33 similarly reported that consenting to partake in clinical trials was a challenge in developed countries.
Thanks to the local partnership and strong establishment of MRCG in The Gambia, the country is well-equipped for incoming projects due to receptive communities, ambitious and knowledgeable locally recruited staff, and the research facilities and governance offered by MRCG. Locally recruited staff working in their own communities made communities receptive to the trial due to increased trust. It was previously reported that it was beneficial for clinical trials to utilise their local workforce because it allowed for the use of local knowledge 22 . This made the trials more responsive to the country's needs and more effective in influencing policy 24 . Likewise, the staff were aware that passing knowledge through the family hierarchies may facilitate participant recruitment. Preliminary community sensitisation that allowed information to be passed through the appropriate village hierarchy in The Gambia has been successful in the past 23 .
Another highlighted facilitator in the trial, beyond just participant healthcare, was participant transportation to and from the clinic where the interventions took place. Mobility has been seen as a barrier for non-participation in a clinical trial in The Gambia 29 25 .

LIMITATIONS
A few limitations of this study warrant noting. Data was collected by a young white woman interviewing mostly older men. Broom et al. 41 found heightened "professionalism" and selfcredentialing by men when interviewed by a woman. Likewise, it has been noted that gender All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted September 30, 2020. . https://doi.org/10.1101/2020.09.30.20199992 doi: medRxiv preprint in research warrants more attention, especially in the context of women interviewing men, such as men's assertions of gender identities and gender hierarchy 42 . Additionally, the PI (IS) worked closely with the PI of IHAT-GUT (DIAP) and an association of the two women being colleagues may have influenced participant answers due to concern around socially desirable responding in high-stakes situations where participants aim to make a good impression 43 . Lastly, IHAT-GUT was noted to be well run, using different monitoring, training and consenting systems that have not been used by a study in those communities before.
Therefore, facilitators may have been more prominent 27 , making it harder to relate to other clinical trials.

CONCLUSION
This study highlighted the barriers and facilitators to conducting this clinical trial in a rural and resource poor setting. We reported that the staff were proud of their high calibre of work and ambitious to continue making an impact on the country's education and healthcare levels. These findings highlight how qualitative research identifies the value that staff find clinical trials add to their lives and enables the continuous embrace of clinical research. This study highlighted the value of creating local partnerships in research and for future health programs.
For future clinical trials to be effective in rural and resources poor settings, the cultural context must be carefully considered. Specifically, researchers should devote substantial time to engaging with the community to gain insight into pre-existing beliefs, knowledge, and awareness levels of the population, as well as the social structures at play. Understanding the social context is integral for development and success of clinical trials aimed at creating effective healthcare systems in LMICs. A trial in rural and resource-poor settings can be conducted to the same international standards applied to clinical research in high-income countries.
All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.