INGO/academia collaboration in research: challenges

Myriads of challenges, operating individually or in concert, hamper the utility of INGO-derived data in research. These include weak health systems with poor quality of routine data and objective outcome measures and conflict of interest between donors, implementers and researchers.1 5 8 Many INGO projects are implemented via public sector or in public–private partnerships for cost-effectiveness and sustainability with varying degree of success.1 In most cases, however, services are provided either by local staff employed by INGOs or by the existing care providers. In both situations, personnel lack research skills or research intent is not prioritised or the facilities are not strengthened to support research activities.1 8 In many INGO-led and implementation science programmes, although monitoring and evaluation is always integral, the what? (outcomes and impact), the how? and the when? to monitor and evaluate programmes are (too) often not well defined at inception of health interventions.5 A recent trend among donors is simultaneous engagement of third-party monitors (TPMs) or professional evaluators to work along with programme implementers with the view to improving evaluation outcomes. However, motives are mainly on care evaluation and independence of evaluation findings rather than asking pertinent scientific questions.7 8 In addition, TPMs come into projects often too late to collect relevant quality data to determine an effect, and less than 25% of TPMs engage programme implementers in planning, implementing and interpreting of results.7

Another drawback with INGO data is the tendency to focus predominantly on process indicators (inputs and outputs) with less emphasis on health outcomes (most important in both care and research in contemporary time) and impact due to (un)justifiable belief that health impact and outcomes result from multiple complex pathways involving numerous stakeholders operating simultaneously, and therefore are not appropriate to be claimed by one project or organisation.9 In addition, time frame for most programme is too short to report on outcomes which occur years later. Donors’ interests are largely responsible for INGOs’ overemphasis on process indicators.10 Unfortunately, policy-changing and practice-changing research derives its values in proving that interventions result in improved health outcomes. These conflicting expectations further render INGO-derived data inadequate to answer emerging questions in global health research. Furthermore, INGOs do not routinely capture and report the unintended effects of their programme interventions.11

Finally, because INGO interventions are donor driven, there is often a conflict of interests in terms of what donors want and what science expects; donors frequently ask for indicators of accountability, coverage and value for money.1 9 In contrast, science requires high-quality data based on rigorous methodologies that validly explain existing and/or emerging health challenges. Since INGOs are accountable to their donors, they focus preferentially on service delivery coverage indicators as opposed to generating valid scientific evidence. In fact, many INGOs are of the view that proposals with a focus on research outputs would not be funded by donors.9 While there is no direct evidence to support or refute this claim, studies have found that donors’ funding for primary healthcare in low-income countries is predominantly towards service delivery as opposed to system strengthening.12

Promoting NGO/academia collaboration in research: opportunities

Given INGOs’ superior comparative advantages over academic institutions in coverage and access to LMIC settings where research gaps exist, an enabling environment that sets the stage for INGO–academia collaboration in research is in place. Over the past years, the emphasis has been on promoting collaboration between INGOs and public sector health systems in poor-resource countries on one hand1 and ‘openness’ and ‘transparency’ of research findings on the other hand.13 While this is, undoubtedly, yielding progress, there has not been proportionate growth in our ability to answer key research questions on the health of LMIC populations. Thus, promoting and building INGO–academia research collaboration in LMICs could potentially advance our quest to realise this and subsequently reduce research inequity in global health.

One way to achieve seamless INGO–academia collaboration is building a culture of developing joint proposals in response to donors’ calls that serve both academic and programme’s interests. At this stage, professional evaluators can be invited to contribute their expertise in designing rigorous evaluation methodologies. How subsequent project(s) implementation, monitoring and evaluation is conducted can be agreed on by the collaborating parties at inception through improved communication,14 and can follow the initial joint project designs with objectively well-defined interventions and outcome measures that are relevant to both parties and the product of teamwork.

Governing issues between collaborating parties can be addressed through an early initiation of partnership with a mutually developed project protocol (based on trust, transparency and respect),15 including development of detailed validated data-capturing tools. For instance, project implementation and routine evaluation can be carried out by partner NGOs and professional evaluators, while academic collaborators are given unlimited access to the collected data for subsequent relevant analyses and interpretation. Alternatively, academic institutions can be given direct access to project sites for direct data collection as appropriate. This approach should lead to the collection of pertinent scientifically relevant information that minimises measurement biases. In essence, a key to harmonious and productive collaboration is transparency in communication throughout the process.15 In either approach, strengthening the local research capacity within INGOs and of local academic collaborators (in deficient cases) should be included in the design for sustainability.

An unintended but foreseen unhealthy consequence of joint NGO–academia response to a proposal is the potential to generate competing interest in the distribution of financial resources and ownership of results.15 This can be avoided with open and fair agreement at the start. In one model, academic institution must not necessarily be grants recipient as partner on health programmes. Grants could go to NGOs but postgraduate students from academia should be given access to project sites and collect scientifically relevant information as originally jointly conceptualised. This should create a win-win situation between the parties in the sense that while NGOs receive funding and implement in line with donors’ expectations, academia use their platforms to collect sound scientifically relevant data at no costs (this assumes adequate existing funding and solid research capacity for academic partners to ensure sustainability). This approach could also minimise potential conflict with donors’ needs. Information that is relevant for donor’s needs is summarised and reported by INGOs and professional evaluators while academic researchers retrieve information that is relevant for scientific consumption.