Research capacity building integrated into PHIT projects: leveraging research and research funding to build national capacity

The five AHI-funded PHIT projects have been described in detail in other papers: Ghana [20], Mozambique [21], Rwanda [22], Tanzania [23] and Zambia [24]. In summary, PHIT projects were similar in their core principles but different in their specific activities, coverage and implementation partnerships [25, 26]. The primary goal for each project was strengthening the district-level health system [24‐26]. However, focus on clinical and data quality improvement [25, 27], strengthening the health facility management [21, 22] and focus on the community-level of implementation [20, 22, 23] varied across countries. Further, the size of the implementation areas varied from Rwanda’s project covering two districts with 23 facilities to Mozambique’s project covering 13 districts with 146 facilities. Partnerships included Ministries of Health, local and international academic institutions and other non-governmental organizations.

RCB was a common denominator across the five country proejcts, strengthened by DDCF’s emphasis on implementation research and program evaluation [28]. However, each country differed on the amount of research already happening prior to the start of the PHIT project. A short description for each project’s research and RCB activities prior to AHI funding is provided below.

For this manuscript, each PHIT project nominated one to two member(s) of their team who led or coordinated the RCB activities completed using AHI funding. These individuals are all included as co-authors on this paper. For cohesion, we used Cooke’s framework to synthesize details about PHIT project activities specific to RCB [29]. This framework includes six dimensions of RCB: 1) building skills and confidence, 2) developing linkages and partnerships, 3) ensuring the research is “close to practice,” 4) developing appropriate dissemination, 5) investing in infrastructure, and 6) building elements of sustainability and continuity. The representatives responded to how much each of the RCB domains was a priority for AHI-funded RCB activities and provided a list of their specific activities that were linked to these domains via a semi-structured questionnaire (see Additional file 1). Co-authors Bethany Hedt-Gauthier, Jackline Odhiambo and Ayaga Bawah compiled responses, with follow-up skype interviews and emails to clarify questions arising from the completed survey. Responses from PHIT projects’ RCB leaders were collected and refined between September and December 2015.

In addition, respondents provided data on RCB activity inputs, outputs and outcomes indicators synthesized from other articles describing RCB monitoring and evaluation [29‐31]. There were no standardized metrics for RCB activities across the five projects and projects only reported on RCB indicators when the data had been routinely collected during the implementation of the PHIT project. For this reason, projects at times had no data available for some specific indicators. When available, data could correspond to the overall PHIT project or to a specific RCB-activity implemented as part of the project. For each indicator, we report how many countries are represented and whether the countries are reporting for the combined project or single activity. The indicators were reported in aggregate for the 5 years of PHIT implementation as totals and averages.

Finally, each project key informant provided information on the degree to which common RCB challenges, as listed in a systematic review of non-academic RCB programs [30], persisted for their project and described any innovative activities made possible by AHI funding to alleviate these challenges. For each challenge, RCB leaders reported whether the challenge was a major challenge (score = 2), minor challenge (score = 1) or not a challenge (score = 0) and we reported the average response across the five PHIT projects.

This paper includes program descriptions from co-authors and data that is routinely collected through PHIT project monitoring and evaluation systems. As such, this paper falls under non-human subjects research.

In each of the Cooke’s domains, about half of the PHIT projects had some RCB activities happening prior to AHI funding and these activities were either expanded or catalyzed using the AHI funding (Fig. 1). Further, each program prioritized all five domains for RCB activities, and most country projects reported that each domain was a high priority (Fig. 1).

Specific RCB activities varied country-by-country. However, there were common themes to the types of activities supported using AHI funds (Table 1). Four country projects reported using AHI funds to provide scholarships for Masters and PhD degrees either at local or international institutions. All five projects used AHI funding to increase research administrative support and infrastructure. Four countries hired either research or administrative staff and four country projects used funding to construct or renovate spaces used for data collection, research and/or research training.

All five country projects reported using funding to support mentorship activities (Table 1). Four PHIT projects used funding to hire mentors or facilitate mentorship from national researchers, from international researchers locally based or from international researchers using a distance-mentorship model. For the distance-mentorship model, approaches ranged from exclusively relying on email exchanges with periodic calls to a blended model that included these distant exchanges complemented with occasional face-to-face meetings, most often when the international researcher was visiting the project to support other aspects of the work. Four PHIT projects offered writing workshops to consolidate mentorship activities and accelerate dissemination. Finally, two PHIT projects used AHI funding to provide support to local universities by teaching courses within their academic programs.

All PHIT projects used AHI funding to implement research trainings to build skills. The target competencies, length and style of training varied by program (Table 1). Often PHIT projects implemented multiple training activities of increasing complexity to allow an emerging researcher’s skills to grow over the training series. Finally, all five country projects used AHI funding to support research dissemination, including scholarships for conferences (four projects) and writing workshops with the goal of producing scientific publications (four projects).

Table 2 synthesizes indicators across the five PHIT projects, indicating how many projects had data available and the scope of that data. A country project-level summary is available in an online supplementary table (see Additional file 2). Over the 5 years, there were 201 participants in RCB-related trainings (Table 2). Though rarely documented, 14 of these were known to be practitioners or clinical staff and 27 were programmatic staff. For the four PHIT projects with data available, 22 PhD, Masters or research scholarships were awarded using AHI funds. Three country projects had recorded data relevant to research mentorships, documenting 56 research mentor-mentee pairings. Forty-four research protocols were documented, and 116 manuscripts were attributed to the PHIT projects. Over the 5 years, for the four country projects that tracked this data, there were 59 peer-reviewed publications, of which 29 (49.2%) were first authored by a national researcher and 18 (30.5%) had a national senior author. Four PHIT projects reported that trainees presented at 99 conferences, workshops, or public lectures. Although very few PHIT projects systematically recorded this information, 43 trainees were documented to collaborate on new research projects, 4 trainees were documented to be leading new research projects and 19 trainees went on to become facilitators for a training course or mentors.

At the organizational level, 54 research trainings were held using AHI funding, ranging from 2 to 14 in-person training days with 2–150 days of additional contact (practicums with additional mentorship). Twenty-five PhD- or Masters-level trainers were available for RCB trainings, of whom at least 11 were national trainers. Research promotions were rarely tracked, and only nine promotions were documented and attributed to the PHIT-related trainings. Only two PHIT projects reported eight different guidelines developed or used as part of the AHI funding.

In terms of using the AHI funding to build research networks, for the two PHIT projects with data available, 9 new networks and collaborations were established. Four PHIT projects reported holding 37 forums between policy makers and researchers. Only one country had routinely collected data on the number of research projects that impacted policy and no PHIT projects had documentation on the frequency that research findings impacted quality of care or outcomes or reduced costs of programs/service. For the three PHIT projects reporting, eight additional external donors had expressed interest to fund research and/or RCB activities.

Common challenges for RCB activities were grouped into five categories – trainee background; mentorship, teaching and trainee support; infrastructure and logistics; institutional support and buy-in; and sustainability and funding (Table 3). For trainee background, the most noted challenge was ability to publish in international journals, which respondents linked to the fact that trainees were best suited to produce research close to practice, which is often difficult to publish. In addition, for several PHIT projects, English was not the first language for trainees, hence a language barrier in research writing, and few had any previous experience in scientific writing. The PHIT projects addressed these challenges by identifying journals that published operational and implementation research, providing language and writing support to trainees, and “co-producing” research, with writing teams including implementers and researchers. While PHIT projects reported having mismatches between trainees’ skills and initial training goals, they also noted the importance of tailoring training goals to the abilities of the trainees, offering a range of activities to enable different levels of skills building and using advanced trainees to provide support to less experienced trainees.

All PHIT projects reported that the demand for mentorship exceeded the mentorship resources available (Table 3). The pool of available mentors for each PHIT project was small. Leveraging long-distance mentors expanded the number of mentors available, but introduced additional challenges of providing adequate support to new researchers when language and communication infrastructure barriers existed. Trainee drop-out due to inadequate mentorship was an ongoing challenge. Respondents noted that provision of mentorship was a high priority for AHI-funding and most PHIT projects used formal group skills-building activities, such as writing workshops, to efficiently spread the mentorship resources.

Many of the infrastructure and logistics challenges reported in RCB activities in other settings were not challenges to the AHI-funded projects. All respondents attributed the absence of these barriers to the ability to flexibly target funds from AHI to strengthen or build research infrastructure and procure necessary research materials where gaps were anticipated or emerged during the grant period. The most common challenge in this domain was poor internet. While many of the PHIT projects found work-arounds such as cell-phone modems, all PHIT projects noted that internet availability limited the effectiveness of RCB activities. Similarly, many of the institutional buy-in challenges for RCB activities seen in other settings were not reported by PHIT projects. Respondents attributed the ease of buy-in to the fact that PHIT projects were collaborative across all institutions, including local research institutions, and that research was closely linked to the trainee’s work and identified priorities. However, competing work demands was a noted challenge for trainees and an area where no novel solutions were provided by respondents.

PHIT projects reported sustainability and funding of RCB activities as the largest ongoing challenge (Table 3). Trainees dropped out, during or after the training, due to changing employment especially when the AHI funding was coming to an end and new funding for research or RCB had not been secured. PHIT projects did not identify donors not wanting to fund RCB activities as a large issue; however, dependence on external funding for RCB activities was a major challenge. PHIT projects are currently trying to expand their funding sources for research and RCB activities to continue after the end of the AHI grants and to continue to link RCB activities to existing institutions in-country and at the US partners’ home institutions. All respondents commended DDCF for the level of funding and the flexibility of funding for RCB activities and recommended that other granting organizations follow suit.

Prior to DDCF funding, all five country projects had difficulties obtaining sufficient resources for specific RCB activities. However, the availability and flexibility of AHI resources enabled each PHIT project to implement the suite of activities that matched its needs. Funding with small budgets or restricted timelines encourage one-off training programs without the needed support, particularly mentorship, to build national research capacity [18]. Funding agencies should make sufficient and flexible resources available for each program to develop the most appropriate plan for their current research landscape. Further, funding timelines should span years to reflect that RCB requires a long term commitment to be most effective.

Funding restrictions and limited number of trainers also encourage isolated RCB training programs that target only one skill [30, 32, 33]. This model fails to facilitate the translation of research skills into research projects after the training [30, 34]. Flexible DDCF funding through AHI allowed the PHIT projects to implement a spectrum of RCB activities, in parallel and sequentially, which reflected local context and range of baseline capacities. Individuals started with trainings appropriate for their skill levels and continued to develop skills until research independence. This also created an environment where individuals more advanced in the training spectrum could support and provide mentorship to the more junior trainees, increasing the quantity of training and mentorship possible with limited resources.

For four PHIT projects, existing national research institutions were at the core of their projects’ research. These projects noted that building on existing infrastructure accelerated the initiation of projects and enabled the continuation of research implementation and development of research skills outside of the scope of PHIT-specific projects. Building on existing institutions has been a noted RCB challenge [35]; however, it promotes the sustainability of research capacity [8, 36] and can ensure the continuation of activities even after the initial funding has ended.

Linking PHIT-related research to implementation, program evaluation and quality improvement enhanced the commitment of trainees to the training process until the completion of the training deliverables (most often, a research paper). The connection between research and challenges and priorities encountered during work also increases the likelihood of the research influencing policy and practice [1, 29, 37, 38]. For the PHIT projects, it also reduced the perception that the research training was in competition with the individual’s work, making the trainees and their host institutions more willing participants. While trainees faced publication and time constraint challenges, solutions such as identifying implementation research focused journals and having protected research and writing time have been proposed [29].

Mentorship is critical for developing capacity [39, 40]; however, it is often lacking in RCB activities in LMICs [38, 41, 42]. While each PHIT project noted the essential role of mentorship for their RCB activities, mentorship resources were limited requiring creativity in supporting trainees. All PHIT projects used some form of “learning-by-doing”, including “deliverable-driven” training models that provided trainees opportunities to learn and immediately apply concepts while receiving mentorship, resulting in tangible products such as a protocol or paper. This hands-on mentorship fosters confidence with the key research concepts and skills [39, 40]. PHIT projects also created mentorship groups, pairing teams of trainees with one dedicated mentor to complete a deliverable, expanding mentorship resources and facilitating peer-to-peer learning. Finally, many PHIT projects used e-mentoring, linking in-country trainees with international researchers within their network.

Ongoing measuring and monitoring of RCB activities is vital to ensure effective implementation of RCB activities [29, 34, 36]. All PHIT projects monitored some aspects of their RCB work and many noted that the ability to report on these helped advocate for additional resources. However, a weakness of the RCB implementation was that metrics were not standardized and what was collected varied from site-to-site. To date, no internationally agreed upon metrics exist and we recommend that a standardized tool of RCB indicators be developed and required by funders for future grants with an emphasis on RCB. This standardized tool should include clear indicator definitions and modes for data collection, particularly for more abstract concepts, such as “number of research projects that impacted policy.” PHIT projects also suggested that more of these indicators of RCB activities assess long-term impact and sustainability.