Green suggests that grants should seek to develop problem-solving skills and community leadership and confidence rather than to seek to institutionalise programmes that may become ‘sterile bureaucracies’. This is echoed by Doyle and Patel who describe it as the survival of the organisation for survival’s sake. While CRAIDS funding was project based, the funding mechanism recognised the importance of building the capacity of the organisation, through facilitating and then funding a bottom-up, community-articulated needs approach, supported by community IGAs, with CBOs being the main implementers of the projects / interventions. CRAIDS built on existing capacities by funding CBOs that had already been active in the communities.
The general trend in Zambia however, during the period of Global Initiative support to HIV programmes, has been to move towards top-down project support and away from investing in institution and community-building. Finances and materials are provided for project-based activities by donors and cooperating partners but not for overheads to contribute to institutional functioning. The ZANARA project, through the CRAIDS funding mechanism, attempted to build on existing community capacity. With greater demands for CBOs to produce results and show visible impact, which comes particularly from donors, the pendulum has shifted further towards a utilitarian approach. This has not meant that participation was seen as an end in itself, either by the World Bank or by the communities themselves. CBOs were formed to improve lives of people living with HIV and their families. Most frameworks to evaluate sustainability do not consider health services as a component[24, 39], instead paying more attention to organisational/community capacity. This suggests that the literature attaches more importance to the empowerment framework.
Harman, reporting on World Bank MAP supported programmes in Kenya, Tanzania and Uganda, proposed that under this programme, CBOs were merely the implementers of projects, thus following a more utilitarian framework. This Zambia case study suggests that the World Bank (ZANARA-CRAIDS) support succeeded in delivering on many elements of the empowerment framework by placing the community at the centre of health service delivery, with the need to build the capacity of such groups.
For primary health care, the debate of the last three decades focused on selective (or vertical) versus comprehensive (horizontal) health service delivery. Vertical health programmes are focused on a single issue or disease, where health activities occur in parallel and in addition to routine primary care activities. Integrated programmes (also known as horizontal programmes) “tackle overall health problems on a wide front and on a long term basis through the creation of a system of permanent institutions commonly known as general health services.”. Integrated health care provides care for interrelated health problems for entire populations. Recent years have seen a shift towards combining the strengths of both approaches.
Our interest in the vertical versus integrated debate is in how it relates to sustainability. Atun et al. suggest that limited evidence does not allow for clear conclusions about when vertical approaches are desirable. Specifically in relation to sustainability, two studies[53, 54] found that vertical programmes create unfavourable conditions for sustainability once donor funding ceases, and a decrease in community self-reliance. The limited evidence shows that vertical programmes are effective as a temporary time-limited measure (the key term here is temporary). While the dichotomy between vertical and horizontal is not as rigid in practice as in theory[55, 56], horizontal approaches are seen as fostering more holistic approaches to health that are more aligned to local needs.
Most CBOs implementing HIV care and support activities in Mumbwa have been vertically funded, meaning that funding agencies and NGOs at a higher level selected, designed and funded the HIV activities that the CBOs undertook. A number of studies[24, 26, 57] have found that vertical programmes are less likely to be sustained than programmes that are well integrated with existing (community) systems and structures. Vertical programmes focus resources and activities on well-defined goals, but are less likely to attract indigenous sources of funding, making them vulnerable to demise when external funding ends.
Harman found that organisations that had expanded their practices were able to attract funds from a wider range of sources, as donors perceived them to have a more holistic approach in responding to the epidemic. Also, there are less transaction costs for a donor who funds one group that engages in multiple activities, than if funding several organisations to undertake different activities. Funding 1,800 CBO HIV-related projects represented a significant effort on the part of ZANARA-CRAIDS; and by 2011, donors in Zambia were shifting their focus from HIV to broader poverty-alleviation interventions such as social cash-transfer schemes. CBOs in Mumbwa remained within their niche focus of a specific care or support activity for HIV/AIDS, through the decade.
Integrating HIV/AIDS into other development initiatives, for example around Mother and Child Health and other priorities that are once again on the agenda of donors and the Zambian Government, could be a logical strategy for CBOs to continue to attract funding and build sustainability. Whatever the changes in focus at the national level, from Government and donors, CBOs will continue to be particularly well placed to build on existing relationships of trust with communities.
A key challenge for community participation in health lies in how it should be institutionalised with the formal health service. An integrated health system needs to incorporate a community and population dimension, which CBOs are best placed in Zambia and elsewhere in Africa to support. These are different and essential roles. The Brazilian Family Health Programme has institutionalised community health committees, as part of municipal health services to sustain local participation, ensuring that community participation does not become an alternative to, but an integral part of the state’s responsibility for health care delivery to the population.
An ideal world of community participation in health would be side-by-side involvement of community members with health care professionals and a responsible sharing of both power and responsibility. Findings in this study showed the CRAIDS funding had supported the development of supportive links between the formal (health facilities) and informal (communities and volunteers) health sectors, supporting treatment adherence and effective healthcare seeking behaviour. This appears to have been an unintended and important effect, which may not have been captured in the World Bank’s evaluation of its MAP project in Zambia. It appears that lessons from the CRAIDS era are being incorporated into Zambia’s National Community Health Worker Strategy, recognising the importance of the work carried out by community health workers for the effective running of the health system. However, other successful components of the CRAIDS project, such as the delivery of care and support by CBOs to people living with and affected by HIV/AIDS, have been neglected by donors following the cessation of the CRAIDS funding.
Lehman and Sanders wrote that there is little evidence that volunteerism can be sustained for long periods of time. Community volunteers and health workers in lower and middle income countries expect and require an income. Community needs and demands often require full-time health workers, and volunteers need to spend time on other breadwinning activities. Evidence points to higher attrition rates associated with volunteers than with formal health workers.
CBOs rely on volunteerism because Government and donors are reluctant to fund costs outside of core services to a target group, and rarely support payments to community level workers. The ZANARA-CRAIDS 2006 Annual Report stated that trained volunteers were unwilling to participate in outreach programmes without being paid allowances. This was shown not to be the case in Mumbwa, where most volunteers worked for no allowances, except those that the CBOs had generated through IGAs. However, these allowances were themselves not sustainable, due to the unsustainability of the IGAs. Other factors are essential to sustainability, including ensuring that volunteers know that they are appreciated and are being fairly compensated for out of pocket expenses and losses of earnings. In general, volunteer labour and capital costs are locally generated resources that have a better chance of being sustained, if mobilised through community / CBO-driven responses to community problems such as HIV and AIDS.
CBO sustainability depends on the commitment of those volunteers who make the services happen. Findings from this study showed that health facility staff had begun to accept and appreciate the importance and usefulness of the CBO volunteer in providing health services. Other studies have shown that lay counsellors relieve the workload of overstretched health care workers. According to CBO representatives in this study, health workers recognised and appreciated the services provided by lay counsellors. However, while commitment amongst existing volunteers was high, decreases in numbers of volunteers were reported across all CBOs.
Community participation and so-called ‘soft services’, such as those provided by CBOs in Mumbwa, are often perceived as less measurable and therefore more difficult to evaluate. An internal evaluation of the World Bank stated that poor performing projects had no targeting mechanism for reaching the poor and recommended that the World Bank create new incentives for monitoring and evaluation for both the Bank and the borrower. This would include requirements for baseline data, evaluation designs for pilot activities as well as evaluation of main project activities on an ongoing basis. It would be more accurate to state that the Bank evaluation did not succeed in identifying how communities targeted the poor and those in their midst who were most in need of support. NAC Zambia has recognised that the lack of quantified or collated M&E data from civil society means that its contribution to the national response is often underestimated and as a result it has been difficult to argue for appropriate resource allocations to the sector.
Deciding what should be sustained at the end of a project cycle should begin with an assessment of needs and existing community strengths (its social capital) before the start of the project; followed by measurement of achievements during and at the end of the project cycle. Torpey et al. suggest that quality assurance tools should be based on national standards. Measurements of the costs of the activities that contributed to those achievements, and the costs of maintaining core activities needed for sustaining achievements, are also essential. There is no evidence to show that this occurred in Mumbwa, although the field appraisal approach can be considered to be an informal or proxy method of measurement. The CRAIDS implementation manual states that there will be a focus on “impact indicators”, however there is no evidence to show that these were developed. This does not mean that little or nothing was achieved, or that efforts to identify successes and generate useful lessons – qualitatively if not quantitatively – are not of value. The greater inherent risk of bias in qualitative evaluations may have precluded the World Bank from identifying and ‘making more’ of such positive lessons; as may have an inherent preference for quantitative measurement and gold standard evaluation design methods.
Globally, frameworks, assessments and tools exist for carrying out sustainability analyses of programmes. However, there is little evidence to show that these sustainability assessments and tools are an integral part of donors funding mechanisms. There was disagreement between CBOs and district officials in Mumbwa as to what HIV care and support services were needed. Sustainability assessments from the outset might have avoided this and could have been an important first step towards evaluating what should be sustained for CBO services in Mumbwa. Individual CBOs were not well placed to identify IGAs with a good chance of longer term success and lacked the capacity, technical advice and support to translate good ideas into sustainable IGA programmes. While the CRAIDS application procedure required that each target should include an IGA element, in reality only about 45% of all community projects included an IGA.
The difficulty in costing for service sustainability for ‘soft services’ is reflected in the lack of studies addressing the subject. A paper by USAID analyses costs from a CBO programme in Zambia for children affected by HIV/AIDS, providing a comprehensive set of services. Health and nutrition were the most costly programmatic areas and psychosocial support was one of the least costly programmatic areas. There is no evidence from CRAIDS or World Bank documents that a costing of sustainability of those services took place in Zambia.
A particularly important feature of the HIV-support CBOs in Mumbwa that received funding was that they all pre-dated CRAIDS funding; and while their ability to provide specific HIV care related services increased with CRAIDS funding, the project was able to build on existing structures. These CBOs continued to function and deliver HIV support services, though at lesser capacity, following the cessation of CRAIDS funding. Therefore, it seems likely that this level of sustainability was due at least in part to CRAIDS having engaged with and supported pre-existing community structures, as it also worked with existing district, provincial and national level systems.