Through our analysis, we identified a number of common approaches to engaging with policy and practice and lessons learnt about effective approaches for enhancing the impact of operational research on health equity. These findings are presented below against the key areas identified in the ODI RAPID framework.
External influences and policy context
We use this part of the framework to analyse the enablers/drivers behind the different operational research projects. In all cases this involved impetus from both within the research country and internationally so we have blended policy contexts and external influences.
In
Kenya (CS1) in 2000 the Government of Kenya stated its aims to scale up quality VCT services to five sites in each of the 70 districts in Kenya [
14] and incorporated this into the Kenya National AIDs Strategic Plan. At the same time new technologies in the form of rapid tests were being promoted by the WHO and donors were adding their voice and funds to the proposed scale up of services. With the government committing to support salaries of VCT sites co-located in health facilities this sustainable model became popular and up to 80% of VCT facilities to date remain in primary health care settings. It was unclear both how this model would be best rolled out, what the training and supervision requirements were and what the patterns of uptake among youth, men and women would be and whether there would be any prevention benefits[
15]. Further operational research was required to elucidate some of these questions.
In
Malawi (CS2), a partnership exists between the National TB Control Programme and an independent research organisation - Research for Equity and Community Health (REACH Trust). For a decade, the World Health Organisation (WHO) has reported that the NTP is detecting fewer than 50% of the estimated 48,000 (cases of TB arising in the country each year)[
16]. Research conducted by REACH Trust (formerly EQUI-TB Knowledge Programme) highlighted the multiple challenges poor women and men were facing in accessing a TB diagnosis. These included the time and expenditure involved in multiple visits to different care providers, including multiple visits to local grocery stores[
17]. These findings together with a need to try to meet the WHO target for case finding of 70%, meant that the NTP had an enthusiasm for piloting new approaches to increase tuberculosis case finding among poor and vulnerable groups. REACH Trust, together with LSTM and the NTP secured funding from the Norwegian Heart and Patient Lung Association to test whether empowering grocery storekeepers to refer of TB suspects for formal diagnosis could increase TB case finding. Grocery storekeepers in two poor, peri-urban sub-districts in the capital Lilongwe were trained. The TB case detection rate from the two intervention sub-districts more than doubled while the TB case detection rate remained static in a third, comparable sub-district (control) where there was no interaction with grocery storekeepers[
18].
In Nigeria (CS3), the Federal and State Health Ministries recognised that in order to improve their health indicators such as maternal and child mortality rates they needed to provide better access to diagnostic services in the community. In this case there was a clear policy requirement for action, which also meant finding ways to provide better quality laboratory systems that could be incorporated into national control programmes. The DFID funded Partnership for Transforming Health Systems (PATH) which focused on 5 states responded to requests from state level to improve access to quality tests for hard-to-reach communities. By empowering and building the capacity of teams of laboratory supervisors the project set up simple diagnostic tests in 92 primary health facilities serving a population of >1,000,000. Aspects of this project are now being incorporated into national disease control programmes.
In the three case studies there was a policy momentum for action; new knowledge and new approaches were needed to fulfil policy requirements. All cases also included partnership between researchers/research organisations and policy makers and service providers from the inception of the operational research.
Links
All three case studies included partnerships with key policy makers throughout the cycle of research beginning with problem formation (CS1, MoH, National AIDS and STD Control Programme; CS2 National TB Control Programme and CS 3 the federal and state ministries. But the links do not end here - developing partnerships at multiple levels and with multiple players in the health system was key in all three case studies. For example, the process of producing guidelines for scaling up HIV testing and counselling in Kenya (CS1) involved: 1) establishing a national taskforce; 2) involving counsellors from the districts in iterations and testing of the guidelines; and 3) incorporating clients' concerns into guidelines. CS2 on TB services in Malawi involved engaging with grocery store keepers, community leaders, urban assemblies, district health officers and district TB programme officers thereby forming a bridge between informal health providers and the formal health system. CS3 required effective communications between state authorities and those responsible for local government activities, as well as ensuring engagement with national policy makers and decision-makers. Harmonisation of the project activities with those of various NGOs and vertical programmes was essential to avoid duplication and for sharing of resources.
Also common across the 3 case studies and summarised in additional file
1 were
capacity building activities to consolidate links and partnership. These were part and parcel of the operational research approach which required capacity building in the provision of new services (e.g. diagnostic tests at community facilities in CS3, and VCT within different sites and modalities in CS1) or new players in service provision (e.g. grocery store owners in CS2). In Kenya, CS1, district staff were selected as VCT support supervisors and eventually trained as trainers, able to establish sites from scratch in neighbouring districts. They were also able to offer support supervision and regular refresher courses to counsellors. The Malawi case study (CS2) involved ensuring grocery store owners could provide a screening service for TB diagnosis, training community groups in TB awareness and working collaboratively with the health workers and the NTP to develop case finding activities aimed at poor and marginalised groups. Establishing diagnostic systems in Nigeria (CS 3) involved equipping health workers with the skills to perform accurate tests as well as establishing sustainable systems between state referral and community facilities to check and improve quality. This strengthened teaching and supervision systems between secondary and primary tiers and facilitated constructive engagement with Federal policy and programme planners.
The time and resource costs of developing these multiple links and partnerships were clearly highlighted in discussion with PIs and research partners. Researchers stressed the importance of being flexible and responsive to new opportunities for partnership given the fluid and changing context - locally, nationally and globally. In Nigeria (CS3), negotiations and consensus building for joined up service delivery in different states and across different tiers within states and managing potential conflicts with the objectives of NGOs and vertical programmes, was hugely time consuming. Researchers faced challenges in embedding capacity building within different institutional cultures, especially in contexts of high staff turnover resulting in limited institutional memory. The importance and difficulty of capacity building at multiple levels - for example for research, for delivery and scale-up - was also highlighted. Difficulties were exacerbated because of limited budgets, relatively short research project time scales and the number of players involved. Further difficulties may be faced in the uptake of operational research findings and approaches in policy. For example in Malawi (CS2) the research team were concerned that the NTP had underestimated the resources that would be required to operationalise and scale up the new strategy of engaging with informal providers. The complexity of the partnerships necessary for scaling up proven approaches can also delay uptake.
In Nigeria (CS3) the long term plan is to have a dedicated unit at federal level to take forward the integration of community testing and quality assurance processes into state and national plans, but this is a complex process requiring agreement across several programmes and departments as well as involvement of external funders and NGOs.
Evidence
We use this part of the ODI framework to analyse key issues in the creation of evidence for equity and pro-poor approaches given our explicit aim of learning from operational research that promoted policy uptake for equitable interventions. Of note was the
framing of the research problem: for example CS3 specifically chose to focus on anaemia, malaria and TB, which typically affect poor girls, boys, women and men[
19,
20]and where there had been less investment in diagnostic resources (as compared to HIV for example). The aim of CS2 was to increase case finding (which is a key concern of TB programmes) amongst poor and vulnerable communities and individuals by bringing services closer to poor communities and reducing the costs and opportunity costs of TB care seeking.
The approach to sampling was also guided by equity considerations for example the operational research on diagnostics in Nigeria (CS 3) specifically selected hard-to reach sites on the basis of lack of access to diagnostic services, and the operational research in Malawi specifically selected poor areas in Lilongwe to pilot new approaches to case finding. CS1 included work in underserved rural health facilities and dispensaries, normalising HIV testing and linking quality assurance of services to the local community. All three cases demonstrated that new approaches to service delivery can work in poor, under-resourced areas making advocacy to scale-up successful interventions more compelling.
All the research projects also used multiple methods to gather robust evidence on poverty and equity. Multiple methods used included analysis of routine data from health records, exit surveys, questionnaires and qualitative and participatory methods, such as participatory gender sensitive poverty assessments. These enabled the interpretation of client and/or different community members' experiences (disaggregated for example by age, gender and socio-economic status) against a background of statistical analysis of Health Information Systems Data when investigating uptake of HIV Counselling and Testing and TB case finding (CS 1 and 2).
Developing
indicators and assessing progress against them is a critical component of operational research[
21]. The use of indicators that addressed equity considerations was central to all case studies. For example, the operational research with grocery store owners in Malawi (CS 2) used a poverty scale[
22] to determine who benefited. CS1 included indicators on the number of individuals tested and receiving their results disaggregated by gender, age and HIV status[
23] and also systems to incorporate feedback from counsellors. Operational research also requires ongoing assessment of progress against indicators and being responsive to emerging challenges. In Kenya, counsellor feedback on increasing numbers of clients reporting gender based violence led Liverpool VCT, Treatment and Care to collaborate with the Ministry of Health and other stakeholders to design, pilot and roll out comprehensive post rape care services within the HIV testing and counselling setting (CS 1)[
24]. Limited access to HIV testing in rural areas led to the development of a mobile and outreach programme that brought services closer to potential users and made information available in sign language, through the use of deaf staff [
14,
25].
Researchers faced challenges in data availability on poverty and equity. Even where a poverty scale exists (CS2) it did not capture all axes of poverty and inequity, such as disability. Researchers stressed the importance of developing additional data generating strategies, to enable a more comprehensive equity analysis. Another challenge to ongoing equity analysis lies in funding the recommendations emerging from this, such as the need for mobile VCT programmes, and the development of post rape care services and VCT services for the deaf,, which arose from equity analysis of VCT scale-up data in Kenya (CS 1).