This study aimed at exploring the acceptability of the A4R framework from the perspectives of CHMT members, local government officials, the health workforce, and members of user boards and committees. Increasing calls for decision-makers to be explicit about priorities (and the rationales behind them), coupled with the growing acknowledgment that priority-setting in health care is partly subjective and value-based in nature, has led to greater expectations that the A4R framework could help to improve priority-setting and resource allocation in health care institutions. Understanding the perceptions of stakeholders is crucial for the proper implementation of the A4R framework and could, in turn, help to assess the feasibility and sustainability of the innovation in priority-setting and decision making processes in the district. It is thought that this is the first study to document the actual experience of implementing the A4R framework in the planning and priority-setting process in low-income countries.
The picture of the A4R framework that emerged from the respondents was, overall, a positive one. The approach was seen as an important tool that could be used for improving priority-setting and health service delivery: first, all respondents shared the opinion that involving multiple stakeholders would ensure that a wide range of relevant values and principles would be taken into account, and thus this would improve fairness, transparency and legitimacy of the priorities identified. Second, all categories of respondents recognised that transparency had the potential to enhance the democratic process by helping the members of the community to learn how to allocate health care resources thoughtfully and fairly. Further, respondents widely shared the view that a formal appeals mechanism would provide opportunities for people to express their dissatisfaction with decisions and revisions, based on evidence. This finding resonates with the previous study of the district health planners in Tanzania by Mshana et al. [19
]. Mshana et al
. presented the framework to district health planners in a series of capacity-building workshops. Participants liked the framework, especially the extensive participation it called for, the strong expectation of transparency, and the potential for including a wider range of values.
However, in our study, a few aspects of the A4R framework were perceived as problematic by a majority of respondents. First, a majority of the CHMT members felt that adding other values and criteria was complex and easier said than done. They argued that the MoHSW had already highlighted criteria to be followed by the districts when preparing health plans. It was evident from the study that the high level of conditionality associated with the use of planning guidelines gave the CHMT, at least theoretically, little room to add other values and criteria. However, a recent study in the district documented that the planning and priority-setting was seldom evidence-based [17
]. According to these authors, priority-setting usually occurred in the context of budget cycles and the process was driven by historical allocation; the use of epidemiological or cost-effectiveness evidence tended to be only a small component of the decisions. These authors also found that district health priorities were rarely implemented as planned, and lots of unstructured priority resetting happened throughout the year.
Closely related to this, concerns were also expressed by the CHMT regarding the involvement of multiple stakeholders in the planning process. The findings showed that there was fear among CHMT members of including additional stakeholders in planning and priority-setting meetings. One of the reasons provided was that many stakeholders did not have the knowledge, skills and experience to effectively contribute to priority-setting decisions. However, when CHMT members took initiatives to visit villages, to solicit community priorities, they were astonished how lay people provided useful information which was important in the preparation of the district health plan. Other studies carried out in Uganda and Tanzania have also shown that when lay people are provided with evidence they are able to engage in simulated priority-setting decisions [25
]. Similarly, a study of priority-setting in Canada showed that service users can make a strong contribution to the process, provided that they are given time to build trust with other members of the decision making group [27
]. It can, therefore, be argued that well-engaged community members could significantly contribute to non-expert values and criteria such as acceptability, feasibility, community support options and willingness, NGO partnership, age weighting etc.
Lack of funds and planning guidelines imposed by the national government were also frequently mentioned by the CHMT members as barriers to stakeholder involvement in the planning process. A review of the planning guidelines, however, shows that the CHMT enjoys a reasonable level of autonomy in terms of decision making and priority setting. The central government has delegated substantial decision making authority to the CHMT over a number of domains, including the opportunity to decide the number and type of lay stakeholders invited to the planning meetings, should the need arise. While lack of funds may continue to impinge on the CHMT's desire to broaden stakeholder participation, much depends on the willingness of the district authorities to set aside funds to this effect. Besides basket funds, which are often earmarked for specific activities, the district authority receives block grants over which local authorities can make relatively unrestricted choices. In addition, the district authorities often assume larger responsibilities for funding their services by assigning local revenues from taxes and other sources. Similarly, lack of funds might have been falsely used as an excuse against involving multiple stakeholders in planning meetings; part of this conclusion comes from the inconsistencies of the CHMT's actions during the preparation of the annual district health plans. For instance, while the CHMT claimed it had insufficient funds to invite more stakeholders to the planning meetings, including the public, the number of involved and paid for medical professionals has been increasing every year during the annual planning process.
One of the possible explanations of the CHMT's fear of broadening stakeholder participation could be due to resistance to change and the perceptions of CHMT members about the compatibility of the intervention with existing values and past experiences. Indeed, studies in other contexts have demonstrated that innovations that are perceived to be compatible with organisational norms, values and ways of working are more readily adopted [28
]. Overall, the study results suggest that these kinds of complex innovations, which involve changes in behaviour by challenging socio-political conditions and traditions, need more time than can typically be allocated for a research project.
The last concern has to do with the appeals/revision
condition. While stakeholders in Mbarali district widely appreciated the importance of the revision and appeals condition, they also expressed their concern about the applicability and feasibility of it. They argued that the appeals mechanism was difficult to put into action and embed into the daily routine. The challenges of achieving the appeals and revisions condition pointed out by stakeholders in the district should not be surprising, given the socio-political conditions and traditions in which the A4R framework is implemented. The previous study by Maluka et al. found that the district had no culture of appealing against decisions made by authorities [17
]. According to these authors, lack of transparency of the government decision-making bodies in Tanzania, coupled with poor public awareness, seemed to be the major explanatory factors behind the lack of appeals mechanisms.
Experiences from other contexts suggest that if carried out correctly, the revision and appeals condition can close the gap between decision-makers and those affected by the policies, and engage a broader range of stakeholders in the process of deliberation [30
]. However, in this study, little evidence was found to support Norman Daniel's view that even if stakeholders "do not participate in the original decision making about limits, the revision/appeals condition empowers them to play a more effective role in the larger societal deliberation about the issues and to provide wider societal oversight of the limit-setting process" [7
]. Given the lack of transparency of government decision-making bodies in Tanzania, coupled with the low public awareness revealed in this study, opportunities must be provided for service users to participate collaboratively with health organisations and providers in planning, delivery, monitoring and evaluation at all levels, in a dynamic and responsive way. It is not enough for the public to merely be able to follow the prioritisation process from a distance and to appeal against decisions believed to be unfair. People have to be properly informed about priority-setting decisions in order to appeal against them, otherwise they would not even know what exactly they would be appealing against [31
]. This underlines the importance of the relevance
condition aiming for initial inclusiveness of stakeholders in the mechanism for achieving compromise. In this respect, the A4R conditions may be mutually supportive, but the strongest possible initial focus on involvement across formal and informal power differences is likely to accelerate the desired change. In a review of priority-setting in hospital operational planning in Toronto, Gibson et al. proposed a fifth condition, the 'empowerment condition', which requires that there should be efforts to minimise power differences in the decision-making context and to optimise effective stakeholder participation [32
]. Significant efforts need to be made to empower the public, particularly user committees and boards as well as local civil society organisations. The effectiveness of decentralised health care planning and priority-setting is strongly influenced by the ability of the grassroots to hold service providers and local authorities accountable.