key: cord-0064137-oayljpnh
authors: Begg, Sara; Wright, Alexandra; Small, Graham; Abdoulaye, Diabate; Kisinza, William; Koudou, Benjamin; Moore, Sarah; Mosha, Franklin; Edi, Constant; Kirby, Matthew; Kija, Patrick; Malima, Robert; Moore, Jason; Bates, Imelda
title: Ripple effects of research capacity strengthening: a study of the effects of a project to support test facilities in three African countries towards Good Laboratory Practice certification
date: 2021-05-10
journal: Gates Open Res
DOI: 10.12688/gatesopenres.13190.2
sha: 81d886f0433bf56e2a37b2c03124543ee42c3c0e
doc_id: 64137
cord_uid: oayljpnh

Background: Strengthening capacity for public health research is essential to the generation of high-quality, reliable scientific data. This study focuses on a research capacity strengthening project supporting seven test facilities in Africa conducting studies on mosquito vector control products towards Good Laboratory Practice (GLP) certification. It captures the primary effects of the project on each facility’s research capacity, the secondary effects at the individual and institutional level, and the ripple effects that extend beyond the research system. The relationships between effects at different levels are identified and compared to an existing framework for the evaluation of research capacity strengthening initiatives. Methods: To capture the views of individuals engaged in the project at all levels within each facility, a maximum-variation purposive sampling strategy was used. This allowed triangulation between different data sources. Semi-structured interviews were conducted with individuals in three facilities and a combination of email and remote video-call interviews were conducted with individuals at two further facilities. Results: We found that, despite a focus of the GLP certification project at the institutional level, the project had effects also at individual (including enhanced motivation, furtherment of careers) and national/international levels (including development of regional expertise). In addition, we detected ripple effects of the project which extended beyond the research system. Conclusion: This study shows that research capacity strengthening interventions that are focussed on institutional level goals require actions also at individual and national/international levels. The effects of engagement at all three levels can be amplified by collaborative actions at the national/international level. These findings show that research capacity strengthening projects must develop plans that address and evaluate impact at all three levels. Capturing the ripple effects of investment in research capacity strengthening should also be planned for from the beginning of projects to support further engagement of all stakeholders.

Building research capacity in public health and related fields is essential to the generation of robust, innovative and locally relevant scientific data. When research staff are highly skilled and research infrastructure at institutions is strong, the evidence generated by these institutions can inform national policies, support progress towards population health goals and contribute to socioeconomic development [1] [2] [3] [4] . Research capacity strengthening is increasingly an area of focus for international development and global health partners and funding bodies 5,6 . With increasing investment of funds to support research capacity strengthening, there comes an increased need to evaluate the impact of this investment on data quality 7 . Test facilities are a key component of national research capacity. Attention is commonly focused on clinical diagnostic and research facilities, their role in diagnosis and support in disease and epidemiological surveys 8 . However, non-clinical and basic science facilities also have key roles to play in global health research 9 . This can include supporting entomological mapping surveys such as insecticide resistance mapping, generating scientific evidence that can inform the discovery of novel compounds for therapies, development of new products that may have uses in public health, including the control of vectors of diseases, and assessing the safety of these compounds and products before they are used. It is imperative, therefore, that such facilities are included in efforts to build health research capacity, given that not only are they vital for public health, but they also face many of the same challenges and gaps as the more widely researched clinical laboratories 10, 11 .

This study focuses on a research capacity strengthening project supporting seven test facilities in Africa towards full compliance with Organisation for Economic Co-operation and Development (OECD) principles of Good Laboratory Practice (GLP) 12 . These test facilities are all engaged in the evaluation of mosquito vector control products, including long-lasting insecticidal nets and indoor residual spraying formulations 13 . Each test facility consists of an insecticide testing facility (ITF), a molecular biology laboratory, experimental hut sites, an insectary, and animal houses. Data generated by these test facilities inform decision making at a national and international level, as these test facilities have historically conducted laboratory and field efficacy trials on vector control products for evaluation by the WHO Pesticide Evaluation Scheme (WHOPES) 14 which supported national programmes and other stakeholders in the selection and safe and judicious use of public health pesticides. With ever-mounting challenges related to increasing insecticide resistance and changes in vector profile and distribution due to climate change, there is a pressing need for innovative vector control products, tools and approaches. To support this, WHO has now transitioned the function for evaluating these products to the WHO Pre-Qualification Team Vector Control (WHO PQT-VC), to align the quality assurance of vector control products with existing prequalification processes within WHO 15 . Test facilities will now generate data on behalf of companies for the evaluation and prequalified listing of vector control products by WHO PQT-VC, which guides UN agencies, other international organizations and country-level procurement bodies on the procurement of products for malaria management and eradication 16 . Whilst test facilities are moving towards GLP certification, WHO PQT-VC can inspect data-generating facilities to ensure quality data. However, once sufficient test facilities have been granted GLP certification, WHO PQT-VC will require companies 'to develop a product dossier which includes data and information to support the safety, efficacy, and quality requirements appropriate to the product type and generated according to Good Laboratory Practices (GLP) and appropriate Quality Management System (QMS)' 17 . The conduct of studies compliant with GLP principles will ensure that data generated for product registration purposes are reliable, reproducible and auditable and will be recognised by scientists and regulatory authorities worldwide. Each test facility was supported towards GLP certification by the Innovative Vector Control Consortium (IVCC), with funding from the Bill & Melinda Gates Foundation being used to support the development and implementation of quality management systems, infrastructure improvements, facility inspections to identify and address nonconformances with GLP principles and staff training activities.

Research capacity strengthening has been defined as 'a process by which individuals, organisations, and society develop the ability to perform [research] functions effectively, efficiently and Amendments from Version 1 Methods We have clarified throughout that the effects described in this study are in the 5 test facilities involved in the study (described  in a new table, Table 1 ) and highlighted where national/ international level effects arose from the wider group of 7 test facilities. We have added detail on the three-levels of research capacity strengthening used as a framework, and on the purpose and approach to sampling in this study (maximum-variation purposive/purposeful sampling). We have added additional detail on the language of consent documentation (available in French in West Africa, Swahili in Tanzania).

We have reordered throughout so that content always follows the order of individual, institutional and national/international level. We have simplified Table 3 (formerly Table 2 ), with the original full table now available in supplementary materials. Figures 1-3 have been simplified into 2 figures: Figure 1 (effects at the three levels and additional "ripple" effects), and Figure 2 (illustration of the interrelated nature of the three levels in RCS efforts). We have clarified that the extent to which ripple effects can be directly attributed to the RCS efforts is inevitably difficult since within a 'real life' research context there will inevitably be other factors that influence these effects and have highlighted that unexpected/ripple effects have been previously observed in studies of individual level RCS efforts, and included additional references to support this. We have stated that further exploration could be an area for future work. We have added acknowledgement of the absence of quantitate data in this study.

We have simplified the text and removed some areas of repetition, particularly summaries of results within the discussion, and have addressed a few issues of grammar and spelling.

Any further responses from the reviewers can be found at the end of the article REVISED in a sustainable manner to define objectives and priorities, build sustainable institutions and bring solutions to key national problems' 18 . This definition highlights that research capacity strengthening happens at three levels: the individual level, the organisational or institutional level, and the societal or national/international level. In capacity strengthening, initiatives are often focused at one of these three levels 8, 19 , with programme goals and evaluation of programme success aligning directly with these levels. In this study, the described goal was at the institutional level -developing a QMS compliant with the principles of OECD GLP and being granted GLP certification. Despite an institutional-level goal, the interventions required to implement this system acted at individual, institutional, and national/international levels.

The purpose of this study was to capture both the primary effects of the GLP certification project on each institution's research capacity, the secondary effects at the individual and institutional level, and any ripple effects beyond the research system. The relationships between effects at different levels are identified. These effects are compared to an existing framework for the evaluation of research capacity strengthening initiatives, to identify new areas for future laboratory capacity strengthening programmes to consider when developing and evaluating their interventions. In addition, we saw ripple effects of the project beyond research capacity strengthening for both individuals within each facility and into the community surrounding them. Participants were informed about the research using participant information sheets 20 . Written consent was obtained from each participant prior to undertaking an interview. For individuals in Francophone countries, all consent documentation (participant information sheets and consent forms) was provided in French. In Tanzania, consent documentation was provided optionally in both English and Swahili. All individuals were offered onsite translation into an alternative local language; however, this was not required for any interview participants.

This study encompasses five test facilities engaged in the testing of novel vector control products for the purpose of supporting malaria control programmes in Tanzania, Côte D'Ivoire and Burkina Faso. These test facilities are have all received investment and support from IVCC to achieve GLP certification, and are part of a wider programme of support for seven test facilities. Throughout the results and discussion below, findings relate to these five test facilities, although there are references to the benefits of being part of a group of seven institutions. The five test facilities (Table 1 ) included in this study encompass a diverse array of contexts. Kilimanjaro Christian Medical University College, Pan-African Malaria Vector Research Consortium (KCMUCo-PAMVERC), Tanzania, provides crucial information on how GLP certification can be achieved, being the first insecticide testing facility in Africa to do so. Comparison between East and West African contexts was facilitated through inclusion of Centre Suisse de Recherches Scientifques en Côte D'Ivoire (CSRS) and Institut de Recherche en Sciences de la Santé (IRSS), Burkina Faso. Comparison between government and non-government test facilities was facilitated through inclusion of National Institute for Medical Research (NIMR) Amani Centre, Tanzania and Ifakara Health Institute (IHI), Tanzania. These contrasting test facilities enhanced our ability to identify both direct and indirect effects of investments in developing a QMS. Generalisability of findings was assessed through using these facilities to compare effects of investment in QMS in a diverse range of contexts, including different national policy contexts and government/non-government supported test facilities.

To capture the views of individuals who had exposure to the GLP certification process at all levels of these test facilities, a maximum-variation purposive sampling strategy was used 21 . This sampling method intentionally seeks to capture a wide range of views, to identify important shared patterns and points of contrast or conflict. For the purpose of this study, the key dimension of variation was role within the test facility, in recognition that this will have determined both which aspects of the GLP certification process individuals were involved with, and the tasks and duties required of them. Sampling included those who hold key roles within a test facility, as determined by a case-study conducted on the first test facility to achieve GLP certification, KCMUCo-PAMVERC 22 , as well as multiple representatives at each organisational level of the facility. This allowed triangulation between different data sources to determine the trustworthiness of findings. Test facility organograms were used to identify relevant participants, with guidance from stakeholders at IVCC and GLP project managers.

Semi-structured interviews were conducted with individual staff members involved in the GLP process in three test facilities: KCMUCo-PAMVERC, NIMR Amani Centre, and CSRS. The interview topic guide 20 was developed based on previous studies of laboratory capacity strengthening 8 , with additional questions derived from findings from a case study of the GLP certification process at PAMVERC-KCMUCo 22 . One overarching question was specifically related to perceived effects of the project. However, due to the semi-structured nature of the interview, interview participants reflected on the effect of the project throughout the interview. Specific questions asked from the topic guide were matched to the roles and responsibilities of the interviewee. Interviews were audio-recorded and transcribed in full. All interviews were conducted in person, in a private room or office, by two researchers, one of whom had a technical understanding of GLP requirements in insecticide testing facilities and the other having systems evaluation experience. Whilst the lead researcher spoke basic French and Swahili, for interview participants who preferred to undertake the interview in a language other than English, a trusted colleague or research student sat in on the interview to aid with translation.

A combination of email and remote video-call interviews were conducted with individual staff members involved in the GLP process at two other test facilities, IRSS and IHI. This was necessitated by restrictions on travel and reduced working hours following the COVID-19 pandemic, which resulted in significant disruption from March 2019. The overarching questions asked in these interviews were retained from the semi-structured interview guide used for in-person interviews. Follow-up questions, where relevant, were conducted via video-call or email.

A framework analysis 23 was used to identify themes emerging from the interview transcripts following the five-step process of familiarization, identification of thematic framework, indexing, charting and mapping/interpretation. The framework identified was the Research Capacity Strengthening evaluation framework developed by Khisa et al., from African Population and Health Research Center, Nairobi, Kenya and Centre for Capacity Research, Liverpool School of Tropical Medicine, UK 24 . This framework delineates the identified and envisioned effect of research capacity strengthening initiatives at the individual, institutional, and national/international level, developed from a review of the research capacity strengthening literature and refined in consultation with research capacity strengthening funders, implementers, managers and evaluators ( Table 2) .

This framework's conceptualisation of research capacity strengthening initiatives happening at three levels, individual, institutional, and national/international, is rooted in the understanding that while these three levels have different foci, they are interconnected, with interventions at one level both influencing and being influenced by factors at other levels 25 . Broadly speaking, at the individual level the focus is typically on the development of researchers and teams, at the institutional level the focus is on development of systems and processes within university departments or other organizations/institutions, and at the national/international level the focus is on influencing structural factors including policy, regulation and research networks 25,26 .

Following familiarisation with the interview data, further themes were identified and incorporated into the framework, while retaining the individual, institution, and societal level structure. All interview transcripts were indexed using NVivo software version 11 (QSR International).

A total of 65 members of staff from five test facilities participated in this study. 66 were approached to take part, with one declining to take part. Of these staff, 16 were laboratory/ insectary technicians or attendants, 17 were from non-scientific administration/information technology positions, 22 were from scientific middle-management positions, and 11 were from scientific senior management positions. 49 were male and 16 were female. Anonymised identifiers have been used for quotes from transcripts, highlighting the role of the interview participant but not the test facility they are connected to. These are presented in supplementary materials (Effects of GLP project.tab) and referenced by section in the text. Table 3 summarises themes as they relate to the individual, institutional and national/international levels, and two illustrative quotes for each theme are presented. Where relevant, illustrative quotes are from individuals in differing roles.

From the interviews, the research capacity strengthening effect of the programme at the all three level was consistently identified, despite the project's focus on the institutional level. At the individual level these effects were related to the training delivered as part of the GLP project, but there was also a positive relationship between the institutional level effects of improved research environment (both physical and administrative) and individual level motivation and job satisfaction. Recognition of research leadership/ esteem (IND2) I think the team as well would be happy to see the products which have been evaluated here and found to be effective as seen in the market… I'm saying this because I've been involved in evaluating a number of these products. When I go out there even in other missions and I found those products in the market, it's a great feeling and I can tell the story.

Yes, because we are now professional. Professional in everything--when you do something and you see the results you think, "Yes, I've done it." It's a feeling of professionalism. Technician

Career trajectory (IND3) I must admit that in the government system, we don't train these people that much. Here, the system was good for scientists and technicians but not for supportive staff. With GLP, at least they're now considered. 

Nationally/ internationally competitive research and grants (INS3) That is one of the success that we had. Also, the other issue is that we managed to attract some clients, looking for our technical support and the evaluation of their products. For instance, for the phase one evaluation of products--Since the inception of the workshop in Liverpool, we had about three-phase one studies.

We can give data that is trustworthy since it is collected in a defined standard and by using well maintained and validated equipment.

Most of all, the output of good quality data from research is for the benefit of the whole community i.e. when we say a certain product is efficacious then it's really so. This then means protection of the whole public.

Research environment infrastructure (INS4a) I think that before when you worked in different projects the infrastructure was sometimes not adapted to the entomology. Now the infrastructure is there, when you go to the insectarium and to the lab you see there is new materials. There are meeting rooms, and the archive office. Also, in [Field Site] there is a new building. We see that there is some evolution.

You can see now the condition of working for every one of us-scientists, the technicians, the assistants--has been very drastically improved. Research impact and user engagement (NAT3) As [Test Facility], I think the main benefit is to provide -we normally say that if you don't have data, you don't have the right to speak -to provide data which will actually influence the policy change for vector control in the country.

The Ministry of Health are anxious because the [Umbrella Institution] is the technical arm of the Ministry of Health, so they are looking forward to make sure they have a very competent technical arm that can provide good advice related to vector control and evaluation of new vector control tools.

Effect of investment in surrounding community (NAT4) Those people, they really liked the project. The same people from the environment were very happy to be involved in part of the work. They said, "It's useful for us. Our kids stay there. They live there." It was easy to recruit people for working there in the site. Very easy.

The renovation activity which has been done here benefitted the entire community of [Region] . The businesspeople who we are working with, the retailer shops where we were getting materials for the construction, the people here who worked here, we all benefited from this. That money went to them, actually, we had to buy materials and to construct. We had good relationship with them during the entire process. Even there at the villages, the way we renovated our buildings, the way we are taking care of the road to reach there… because of the projects that keeps the volunteers, they get some incentives, things like that, it has been very nice. They got the benefit out of it.

Further institutional level effects encompass sustainable provision of training, and enhanced capacity to deliver competitive research, i.e. GLP-compliant studies. At the national/international level, networks between institutions were developed, which further strengthened individual test facilities (institutions) as inter-facility learning was made possible.

Whilst the project was focused on the institution level, important effects were identified at the individual level. These included extensive training, strengthening of career prospects, furtherment of careers, structured working practices and enhanced work motivation. The research environment was also strengthened through improvements in the procurement processes in some test facilities, and to IT infrastructure across all test facilities. Streamlined procurement processes included the implementation of quality management system practices initiated by the GLP project, in particular in the widespread use of SOPs. This simplified processes and made transfer of work responsibilities more seamless. IT infrastructure improvements were relevant across GLP and non-GLP studies, improving processes for accessing and storing study data, managing results in preparation for scientific reports and publications, and improving communication between staff within the test facility through more widespread use of email and installation of internal telephone systems.

National/international level effects At the national and international level, identified effects included sharing of best practices within consortia and linked institutions, and the development of regional expertise related to data management and quality assurance.

Test facilities saw increased support from national level institutions, including increased investment in infrastructure. This was often coupled with the expectation that they would now act as national centres of excellence, both as a model of best practice and as a provider of training in entomology and relevant SOPs. Increased engagement with research outputs at the national decision-making level was anticipated as the next stage of this enhanced relationship with national level institutions, alongside a belief that this would raise policymakers' expectations of the test facilities' performance. (Quotes: NAT1 and NAT2)) At a national and international level, the opportunity to meet and share experiences with collaborating test facilities allowed best practice to be shared throughout the network, although this was not always fully realised as test facilities sought to strike a balance between collaboration and retaining a competitive advantage as a provider of product testing services. For construction and renovation of infrastructure, best practice was shared between test facilities that were geographically close together, because the requirements for buildings were the same and because travelling to these test facilities to see the buildings in person was easier. Data management and quality assurance expertise that was developed by individuals in test facilities further along the path to GLP certification was also disseminated through the network. This was done formally through the project network, via training workshops and shared resources such as SOPs, and informally as these individuals acted in consultancy roles both within and outside of the institutions collaborating in the programme. Involvement in this network also raised the profile of individual test facilities, allowing these facilities to attract new studies and collaboratorsincluding both GLP and non-GLP studies. (Quotes: NAT3)

Non-research capacity strengthening "ripple" effects Ripple effects of the project beyond research capacity strengthening were widely reported for both individuals and the community surrounding the institutions. At the individual level, these were particularly focused on the transfer of skills developed through training and new practices associated with GLP to home lives. This was particularly true in test facilities that had broad and inclusive training programmes. Here, individuals noted how they had applied time management, organisation, and budgeting skills developed through the GLP project to managing their personal lives and households (Quotes: IND4).

Effects on communities surrounding the institution were rooted in often locally sourced solutions to challenges and, in particular, procurement and infrastructure development. By being locally based and finding local solutions, communities around the test facility saw investment in local businesses for consumables, construction materials and construction teams. Also reported was an increase in local employment as new studies were attracted, creating roles such as mosquito collection for experimental hut studies, and improvements in shared infrastructure such as roads. Test facility staff who recognised these effects in the community both took pride in these effects and valued them highly. (Quotes: NAT4).

Despite a focus on the institutional level, the GLP laboratory capacity strengthening project had effects at each level of the research system -individual, institutional and national/ international. These effects are summarised in Figure 1 . These findings align with factors previously identified for evaluation of research capacity strengthening initiatives 24 . The findings from this study emphasise that, particularly at the individual level but also at the institutional level, the "research team" included in evaluations of research capacity strengthening should include auxiliary, administrative and technical staff. These roles are often neglected in RCS evaluations but are vital for implementation of quality research. It is also imperative that quality training is extended to these roles, as happened in several test facilities within the GLP project. Recognition of research leadership and esteem should not be limited to evaluation of outputs of research scientists in middle and senior management roles but should also encompass recognition of excellence in administrative and technical roles.

The programme was institutionally focused, with the end goal of achieving GLP certification. This, however, required inputs and investment at the individual level (especially external training of key individuals, who then went on to implement training in-house or across the network), at the national/international level (for example, by bringing test facilities together to facilitate international networks and collaboration), as well as at the institutional level. A direct effect at these levels was experienced because of this investment, but it also triggered effects across the boundaries between these levels, demonstrating that the three levels within research systems are interconnected (Figure 2) , and reflecting findings from previous evaluations of individual level initiatives that showed positive secondary effects on national and international collaboration 11 .

This finding supports calls for research capacity strengthening efforts to be explicitly aware of what is happening at all levels and to optimise this effect, even if the described goal is at a single level, in order to plan to optimise these secondary and ripple effects 24, 26, 27 . This may be particularly true for research capacity strengthening initiatives that are targeted at the institutional level, as there is scope for triggering effects across the boundaries with both individual and national/international level, and towards the institution. This also has implications for evaluations of research capacity strengthening initiatives that describe a goal at a single level. In this case, the effects triggered across the boundaries away from the institutional level and jumping directly from the individual to the national/institutional level are effects that contribute to a more broadly strengthened research system without being related to the single-level goal. Nevertheless, these effects are important to capture, both to accurately describe the total effect of a programme, but also because strengthening at the national/international and individual levels then has an effect of further strengthening at the institutional level.

Ripple effects were identified beyond the research system, with rich descriptions of how the GLP project was making a wider difference to the lives of the people and communities that surround the test facility ( Figure 1 ). Unexpected effects arising from research capacity strengthening initiatives have been previously identified, particularly in the development of transferable skills 11, 28 . The findings presented here highlight beneficial effects for communities close to the testing sites which were meaningful to those engaged in the GLP project. Explaining these benefits to those involved in research capacity strengthening projects may help to engage and motivate them during difficult times on the project. Future research could further explore these effects, to better understand how they arise, to what extent they are attributable to the research capacity strengthening efforts, and the impact of these effects on both individuals and communities.

Together, these findings show that the GLP project acted at and had primary and secondary effects at all three levels of the research system, that the relationship between these levels is complex and interrelated, and that there are ripple effects beyond the research system itself. These findings should, therefore, inform the design and evaluation of similar programmes to: 1. Use the three levels -institutional, individual and national/ international -as the foundation for programme development, to promote a holistic approach to programme design, and inform evaluation of effect at each level 24,26 ;

2. Explicitly plan for and capture information from each level about the interactions with other levels, and capture ripple effects 24 .

Many indicators for evaluating the outcomes and effect of research capacity strengthening initiatives at all three levels already exist, and these may form the basis of evaluations of similar projects 7 . Box 1 summarises some suggested areas for consideration when developing evaluations of institutional capacity strengthening projects. For ripple effects in particular a mixed methods or qualitative approach may be beneficial 29,30 . 

The strengths of this study are in the diversity of participants involved, capturing the views of staff filling a wide range of roles in five test facilities across three African countries. This approach ensured that effects meaningful to staff in diverse roles were reflected in the findings and offered a voice to staff less often heard within research teams, such as those of technicians and administrators. Furthermore, by using a qualitative approach, this study was able to richly describe the perceived effects of the GLP project and reveal and explain interactions between these effects.

This study is, however, limited by several factors. As no quantitative data is included in this study, numerical measures of change resulting from the GLP project are not possible. Instead, the study relies on the subjective experiences and opinions of individuals involved in the GLP project. With a grounding in a specific laboratory capacity strengthening project, caution should be exercised on generalising these findings to all research capacity strengthening projects. Test facilities were at different stages towards GLP certification and this study is unlikely, therefore, to have captured all of the effects of the GLP project. Further effects will likely be identified by staff as the test facilities progress through certification and begin to attract GLP studies. In addition, given the relatively small amount of time specifically dedicated to this question within interviews, it is likely that additional effects may have been identified given more interview time. Finally, changes had to be made to data collection methods due to the COVID-19 pandemic: the responses at the two test facilities that participated via email and video-call are likely to be more superficial due to reduced opportunities to ask follow-up questions on observations.

Building research capacity in public health and related fields is essential to the generation of high quality, reliable scientific data. This study, focussing on a project supporting seven test facilities in Africa towards GLP certification, shows that research capacity strengthening interventions for laboratories with a focus on institutional level goals also require actions at individual and national/international levels. Furthermore, there are interactions that happen in both directions across the boundaries between the individual, institutional, and national/ international levels, with effects at one level triggering a further effect at another level. These interactions can amplify the effects of an intervention, including research capacity strengthening effects which are the primary objective of such projects. Finally, there are additional "ripple effects" that extend beyond the research system, but that are meaningful to individuals engaged in these projects. The significance of these findings are twofold: firstly, it confirms the interactions between the levels of the research system and, therefore, adds to the evidence that research capacity strengthening projects should plan both to address and to evaluate their effects at all three levels; and secondly, it shows that it is possible to capture secondary and ripple effects of investment in research capacity strengthening and that capturing these effects should be planned for explicitly at the instigation of the project to support further engagement of stakeholders in research capacity strengthening.

Underlying data Transcriptions of interviews with facility staff are available from the research group on request (please email ccr@lstmed.ac.uk to request access), on a case by case basis for the purpose of informing further research and on the condition that it will not be published in part or in entirety. They have not been made available as a dataset because they cannot be de-identified without compromising anonymity and the ethical approval conditions for the project stated that only the research team would have access to the data. 

Swiss Tropical and Public Health Institute, Basel, Switzerland

After the revisions taken, the study has gained clarity and has become more concise in its outline and appearance, the figures more sound, and the manuscript should be made available to the research community, in particular the one involved in (global) RCS.

Still, the write-up of the manuscript would benefit from some continued minor revisions to gain clearness and improved readability, as described in the following:

Although the authors' answer mentions that repetitions (i.e. on the ripple effect) are being removed, there are still too many of them. I assess it is possible to still shorten the text by making it more concise and better readable. Still, it is not surprising that such a project has additional positive side effects so that the "ripple effects" should not be overemphasized which I feel still are.

The reader still does not spot easily that only 5 centres are being investigated on if not reading the revision comments before ("Clarified throughout that effects described in this study are in the 5 test facilities involved in the study (described in Table 1 ) and highlighted where national/international level effects arose from the wider group of 7 test facilities") True that mention is at the beginning of the results but should be mentioned at least towards the end of the introduction and more visibly in the methods section. ○ And yet Table 3 is still difficult to digest, in particular since the mention "Selected two illustrative quotes for each theme in Table 2 and provided all other illustrative quotes in supplementary material." raises different expectations: Table 2 I understand is now the framework without quotes. Table 3 (which I assume is meant) still has way too many quotes ○ for them to be selective so that there is no substantial difference to the supplemental material. An exemplified selection of quotes would be capturable by the reader. In this context, quotes shall be transcribed under best practices (i.e. false starts and fillers should be deleted when overwhelming, i.e. "they have -how they have gone, how far have they gone") -they may be left in the original version in the suppl. material but adapted in the illustrative quotes.

I still believe that a road created is a very rare and not in all cases directly linked side effect of an educative project and hence inappropriate (in particular for the figure) , also there are many more examples of infrastructure.

The discussion encompasses some sentences which remain opaque to the reader even after several attempts (i.e. page 13 left column, upper part starting with "and to optimese this effect..." down to "...related to the single-level goal."; or dto middle part "testing sites which were meaningful to those engaged in the GLP project") ○ Delete: page 13 left column, middle part "Future research..." until end of sentence ○ Seems to be a repetition: page 13 left column middle part "Together, these findings..." until "...the research system itself." and is taken up again in the last sentence of this column.

The two numbers points (1. Use the three... and 2. Explicitly plan...) should be the end since they point to Box 1 (?) and could be part of the conclusion.

The conclusion is yet another repetition. I trust the text could gain poignancy by revising both the discussion and conclusion.

The language seems to become increasingly nonchalant towards the end of the textplease keep up high quality working throughout the text, i.e. the "GLP project" (see next point), increased redundancies, use of RCS in the discussion as acronym although never used nor spelled out before.

○ Inconsistent use of the project name: whereas in the beginning there is no real mention of the project name, it seems to be referred to as "the GLP certification project" but then towards the end turns into slangish "'the GLP project'". This report is another very nice example on assessing collaborative investment into research capacity, with a major resource allocation into people, and descriptively identifies details on those effects. Nicely, these effects are looked at on various levels, with a view on the interconnectivity between these layers. While they can be applied to similar research capacity building activities in the context of vector control, the transfer to other life science areas is being discussed. Surely, the community will learn from, and build upon, these experiences. Also, the study contextualizes the results into existing frameworks of RCS analysis not least through baseline consideration which renders the results more comparable.

While it is not surprising that the study identifies positive side or unintended effects which are being named "ripple effects", it is interesting to see how these additional effects are being carved out of the study participants' views.

This work should definitely be made available to the research community, in particular the one involved in (global) RCS, through publication in the proposed journal. However, the manuscript would benefit from revisions to gain clearness and improved readability, as described in the following:

While 5 facilities were included into the survey, 7 facilities are involved and are being discussed; it is not always clear which number is referred to throughout the results and discussion.

Since an interesting aspect of the work is the contextualization of the results to the existing frameworks, mention of it in the methods section would strengthen the methodology part.

○ Some more information on the maximum-variation purposive/purposeful sampling should ○ be added in addition to Ref. 19 .

Obviously the centres were investigated in their nature of facilities and equipment. Since it is assumed that these information do not stem from the survey, the description of the centres could be outlined in the methods. 

The logic of the levels starts from institutional, presumably because this was the focus of the RCS activities under investigation. It then goes over individual to national/international level. This order is reflected in many of the passages and Table 1 . In other sections, however, the order seems to follow the more natural one, starting with individual to continue with the institutional in the second place. This is seen in Table 2 and in Fig. 1-3 and in Box 1 and in some sections throughout the text (discussion page 14, second column).

Harmonization could assist the reader in grasping the discussed points.

The ripple effects: even when nicely carved out of the data, such effects are not so much of a surprise since they are identified in all RCS studies. The related content in the results is rather short and may explain the catchy word in the title but the discussion point seems to be a bit overstretched, i.e. through the description of transferrable skills. While unintended positive effects are expected in such RCS projects and were being described earlier, i.e. shared infrastructure such as roads may be caused by a mixture of reasons beyond RCS actions.

○ Table 2 is hard to digest. It's name "Target level for RCS" is unmentioned before, the abbreviation never used, "societal" used as synonym but never in the text. It appears that some 74 comments from the 66 survey participants were collected and grouped. While the entire table provides a wealth of opinions the minority of readers will be able to go through this table as such, so it could go into a supplementary table. The manuscript would benefit from the authors selecting categories and exemplifying representative citations to create a concise table with language-corrected statements by some survey representatives. Throughout the text, there seem to be quite some redundancies, i.e. parts of the results (which could be also shortened) seem to be repeated in the discussion and parts of the ○ discussion in the conclusions, and in particular when it comes to the ripple effects. Less repetition will not diminish the emphasis of this point.

Limitations: as a limitation the lack of a quantitative analysis part to support the results should be mentioned. The community assumes that Consent Forms are also administered in French for the West African countries and in addition in local language, or reliably translated on site into the latter, as the version given is written in English -can this be reconfirmed by the authors? ○ manuscript. Simplified figures into 2 figures; effects at the three levels and additional "ripple" effects, and illustration of the interrelated nature of the three levels in RCS efforts.

Highlighted that unexpected/ripple effects have been previously observed in studies of individual level RCS efforts, and included additional references to support this.

Simplified the text and removed repetition, particularly summaries of results within the discussion.

Added acknowledgement of the absence of quantitate data in this study.

Amended all of these, including additional confirmation on the language of consent documentation (available in French in West Africa, Swahili in Tanzania).

No competing interests were disclosed.

Measuring the outcome and impact of research capacity strengthening initiatives: A review of indicators used or described in the published and grey literature

PubMed Abstract | Publisher Full Text | Free Full Text

A systematic approach to capacity strengthening of laboratory systems for control of neglected tropical diseases in Ghana

Laboratory support of global health research

PubMed Abstract | Publisher Full Text | Free Full Text

Strengthening capacity for natural sciences research: A qualitative assessment to identify good practices, capacity gaps and investment priorities in African research institutions

Research Capacity Strengthening in Low and Middle Income Countries -An Evaluation of the WHO/TDR Career Development Fellowship Programme

Innovative Vector Control Consortium: African Trials Facilities. IVCC. 2020. Reference Source

SumiLarv 2 MR, Chlorfenapyr 240 SC. World Health Organisation: Geneva. 2017. Reference Source 15. World Health Organisation: Prequalification Vector Control. 2020. Reference Source 16. World Health Organisation: Prequalification Vector Control | Prequalified Lists. World Health Organization. 2020. Reference Source 17. World Health Organisation: Overview of the WHO Prequalification of Vector Control Product Assessment

Impact of Health Research Capacity Strengthening in Low-and Middle-Income Countries: The Case of WHO/ TDR Programmes

Reference Source 20. Centre for Capacity Research: Interview Guide and Information Sheets for: Developing laboratory capacity for Good Laboratory Practice certification: lessons from a Tanzanian insecticide testing facility

Purposeful Sampling for Qualitative Data Collection and Analysis in Mixed Method Implementation Research

Developing laboratory capacity for Good Laboratory Practice certification: lessons from a Tanzanian insecticide testing facility

Research Capacity Strengthening in Low and Middle Income Countries -An Evaluation of the WHO/TDR Career Development Fellowship Programme

): e0006112 PubMed Abstract | Publisher Full Text Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and is the work technically sound? Yes Are sufficient details of methods and analysis

We thank the Bill & Melinda Gates Foundation, USAID and UK Aid for their support of the test facilities. We also thank Jameel Bharmal, for his support in data collection and translation, and Russell Dacombe, for his technical input.

Further small typos (i.e. Table 2 effect to effects), please check for typos and don't leave this to the reviewers ○ Box 1: the header mentions "consideration" and three bullet points list them as considerations whereas most of the bullet points start with "consider" -reformulate to avoid considering considerations. ○ Competing Interests: Although the reviewer himself does not have any co-authorships with any of the authors, the institution he works in has collaborations with two of the authors' institutions. This fact, however, does not influence any of the reviewer's comments.I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Reviewer Report 08 February 2021 https://doi.org/10.21956/gatesopenres.14395.r30103 © 2021 Hunsperger E. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Centers for Disease Control and Prevention -Kenya Country Office, Nairobi, KenyaThe authors do not provide any quantitative data to support their findings. Although the authors fully describe the GLP process used to certify 5 laboratories in Africa continent spanning multiple countries, actual quantitative data to measure impact is lacking. The article is full of jargon and defining acronyms that is distracting from the main point. The article would benefit from editing and reducing the content to provide a concise report of the "ripple effect" of research in building lab capacity in these 5 countries. For example, table 1 is completely not helpful to the main point of the article. Similarly Table 2 is very long and overall not very informative. It seems to repeat most of the text. The most important point of this article is how the research activities improved on infrastructure of laboratories, this point gets lost in poor representation of these data or the lack of data.Overall this article would benefit significantly by trimming, removing tables and figures that are not informative and the addition of data to support the findings.

Are all the source data underlying the results available to ensure full reproducibility? No

Competing Interests: No competing interests were disclosed.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.Author Response 14 Apr 2021 Sara Begg, Liverpool School of Tropical Medicine, Liverpool, UK Thank you for taking the time to review this study. The study was designed to be qualitative as opposed to quantitative, and this is why there is no quantitative data presented here. The absence of quantitative data has been added to the limitations section. Comments on further amendments are available in our responses to the report provided by Michael Käser. The objectives of this study were clearly spelt out. This is a model that evaluates the contribution of capacity strengthening project on important stakeholders involved in the generation of quality and reliable scientific data that can influence good policy and decisions in vector control. The study provides useful information on the effects and outcomes of the capacity strengthening on individuals and other relevant institutions. The study also contributes to existing knowledge by providing definite parameters that researchers can use to assess or monitor the impact of capacity building projects.

Are all the source data underlying the results available to ensure full reproducibility? No source data required

Competing Interests: No competing interests were disclosed.Reviewer Expertise: Laboratory and field oriented vector control research. I am also into use of scientific data to make sound public health decisions. I am also involved in capacity strengthening of institutions involved in public health entomology research in Nigeria.

Author Response 14 Apr 2021 Sara Begg, Liverpool School of Tropical Medicine, Liverpool, UK Thank you for taking the time to review this study, and we're pleased that you have found this to be useful and relevant.

No competing interests were disclosed.

Are all the source data underlying the results available to ensure full reproducibility? Yes

Competing Interests: Although the reviewer himself does not have any co-authorships with any of the authors, the institution he works in has collaborations with two of the authors' institutions. This fact, however, does not influence any of the reviewer's comments.Reviewer Expertise: Basic research, cell biology, immunology, molecular genetics Technology development and application, with focus on neglected tropical diseases (NTDs) Capacity building with focus on NTDs and in LMICs Evaluation of research capacity strengthening programmes I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.Author Response 14 Apr 2021 Sara Begg, Liverpool School of Tropical Medicine, Liverpool, UK Thank you for taking the time to review this study, we appreciate your input and experience in this area. We have made the following amendments to the manuscript based on your recommendations:

Clarified throughout that effects described in this study are in the 5 test facilities involved in the study (described in Table 1 ) and highlighted where national/international level effects arose from the wider group of 7 test facilities.○ Added detail on the three-levels of research capacity strengthening used as a framework, in the methods section ○ Added detail on the purpose and approach to sampling in this study (maximumvariation purposive/purposeful sampling) ○

Reordered methods, results and discussion so content always follows the order of individual, institutional and national/international level.○ Clarified that the extent to which ripple effects can be directly attributed to the RCS efforts is inevitably difficult since within a 'real life research context there will inevitably be other factors that influence these effects. Stated that further exploration could be an area for future work.

Selected two illustrative quotes for each theme in Table 2 and provided all other illustrative quotes in supplementary material. Two quotes were retained to ensure views of individuals in a diversity of different roles were still reflected within the main ○