Priority setting for health system strengthening in low income countries. A qualitative case study illustrating the complexities

ABSTRACT

Health systems are critical to the realisation of Universal Health Coverage. There has been insufficient attention to the evaluation of priority setting for health system strengthening within low income countries, including evaluation of the local capacity to implement priorities. This study evaluated the extent to which health system strengthening was prioritized in Uganda. The Kapiriri & Martin framework was used to evaluate health system priority setting from 2005–2015. A document analysis was triangulated with interview data (n = 67) from global, national and subnational stakeholders and analysed using content analysis. Health system strengthening was perceived to be circumvented by a lack of resources as well as influential actors with disease focused, rather than system-oriented, interests. There were defined processes with explicit criteria for identifying priorities and evidence was highly valued. But sub-optimal transparency and weak accountability often compromised the integrity of priority setting and contributed to stalling progress on health system strengthening and achieving health system outcomes. The strengths in the current planning processes should be harnessed. In addition, a systematic approach to priority setting, potentially through the establishment of an independent body, and stronger oversight mechanisms, would strengthen health system planning in this setting.

1. Introduction

Strong and effective health systems are critical to the realisation of Universal Health Coverage (UHC) as well as to the achievement of the Sustainable Development Goals (SDGs) (Chisholm et al., 2010; SDSN, 2014). Health system strengthening has been an important focus for most low income countries (LICs), including Uganda (Box 1), as they progress their paths towards identifying and implementing reforms to build responsive, resilient, and sustainable health systems (WHO, 2007). Health system strengthening, as defined by the WHO, is the process of identifying and implementing the changes in policy and practice in a country’s health system, so that the country can respond better to its health and health system challenges (WHO, 2010). Given the scarcity of resources and overwhelming unmet need in most LICs, hard decisions must be made about where resources should be spent to support and develop health systems as a strategic priority. Effective priority setting is thus critical for advancing UHC and health system strengthening (Baltussen et al., 2016; Jamison et al., 2018; Teerawattananon & Luz et al., 2016). Priority setting can be defined as the ranking of different programmes for the purpose of resource allocation and implementation. It occurs at different levels of the health system, including at the national level (macro); at the sub-national (meso-) level and at the bedside (micro-) level.

Box 1.

Snapshot of the health sector in Uganda

As part of the health sector reforms, the Ugandan health system is decentralized. While the Ministry of Health is responsible for policy direction, training, monitoring and evaluation; districts are responsible for implementing the national policies and priorities.

The health sector reforms also involved the re-orienting of services to primary health care, with the use of the basic care package as a vehicle for ensuring universal access to health care. Ideally, health funding should focus on supporting the delivery of the basic health care package. However, to date, the country has been unable to provide adequate funding to ensure universal access to the package.

Part of the reforms required that funding from development assistance partners (which is up to 60% of total health expenditure) is pooled to ensure that the government takes the lead in allocating the budget across the health sector; as opposed to project funding which was prevalent. However, some donors opted out and continue to support their preferred projects as opposed to providing budget support as proposed by the Uganda government in their Health Sector Strategic Plan.

The literature on priority setting in LICs has tended to be programme specific (Bird et al., 2011; Essue & Kapiriri, 2018; Kapiriri, Tomlinson et al., 2007; Wallace & Kapiriri, 2017), focused on the processes within specific institutions (Barasa et al., 2016, 2015, 2017; Kapiriri, Norheim et al., 2007) or limited to particular levels of the health system (Byskov et al., 2014; Kapiriri et al., 2003). Other efforts have focused on the applications of Health Technology Assessment and cost-effectiveness research (Jamison et al., 2018; Li et al., 2016; Teerawattananon & Luz et al., 2016). While this research aimed to provide guidance on how priority setting for health should occur within LICs, there has been a lack of focus on systematically examining the links between priority setting and health system strengthening, particularly in the context of achieving the UHC and SDG agendas. Nuanced evaluations are critical for exploring the complex and interacting factors that affect whether priorities are implemented to support the strengthening of health systems and the achievement of population health outcomes (Borghi & Chalabi, 2017).

There are three main evaluation frameworks have been used to assess priority setting in low income countries. Most widely used is Daniels’s (2000) Accountability for Reasonableness framework (Daniels, 2000). It specifies four conditions for fair priority setting (relevance, publicity, appeals, and enforcement) to operationalise the concept of fairness, but has limited focus on distilling the contextual factors and outcome measures that are associated with priority setting. When used in Tanzania, Zambia, and Kenya, the studies indicated that priority setting did not meet the fairness conditions (Byskov et al., 2009; Maluka et al., 2011). Barasa et al.’s framework (Barasa et al., 2016) focuses on evaluation of the macro and meso levels and has mostly been applied to evaluate priority setting in the context of the Kenyan hospital system (Barasa et al., 2017). Finally, Kapiriri and Martin’s framework (Kapiriri & Martin, 2010) for evaluating priority setting in low income countries has been validated by global and Ugandan policymakers as a useful framework to guide a comprehensive and holistic approach for evaluating priority setting for health in low income countries (Kapiriri, 2017). The framework was developed based on the literature on health systems, evaluation, and priority setting, as well as interviews with experts in the field (Kapiriri & Martin, 2010). It has been implemented to evaluate several cases of priority setting for health in LICs (Essue & Kapiriri, 2018; Kapiriri & Be LaRose, 2019; Wallace & Kapiriri, 2017, 2017).

The objective of this paper was to evaluate the extent to which health system strengthening has been prioritised and achieved in Uganda.

2. Methods

2.1. The analytical framework

The health system was the topic of analysis and was defined according to the WHO’s core pillars of health systems performance (WHO, 2007). The Kapiriri and Martin evaluation framework (the evaluation framework hereafter) was used as the analytical framework in this evaluation. It is a multidimensional framework that entails a proceduralist approach, emphasising the importance of key parameters as determinants of fair and successful priority setting. The development of this framework was informed by Daniel’s Accountability for Reasonableness framework (Daniels & Sabin, 1997). The evaluation framework expands on this work by accounting for both the technical and political aspects of priority setting, by recognising the unique priority setting context of low income countries and by deliberately explicating and considering the contextual factors that affect successful priority setting in LICs (Kapiriri, 2017). The value of this framework is that it harnesses the principles of evaluating priority setting in the literature and proposes a standardised approach for the systematic evaluation of priority setting for health by defining key parameters for evaluating successful priority setting, the evidence that is required for its evaluation (i.e., the means of verification (MOVs)), and the objectively verifiable indicator(s) (OVI) that should be assessed in the evaluation (Table 1).

Table 1.

Parameters for evaluating priority setting with corresponding indicators and means of verification

	Parameters of successful priority setting	Objectively verifiable indicators (OVI)	Means of verification (MOV)
Contextual factors	Conducive political, economic, social and cultural context	Relevant contextual factors that may impact priority setting	Follow up intermittent interviews with local stakeholders, systematic longitudinal observations, relevant reports, media reports.
Pre-requisites	Political will	Degree to which the politicians support the set priorities	Follow up intermittent interviews with local stakeholders, systematic longitudinal observations, relevant reports, media reports.
	Resources	Budgetary and human resource allocation to the health sector	National budget documents
	Legitimate and credible institutions	Degree to which the priority setting institutions can set priorities, public confidence in the institutions	Stakeholder and public interviews
	Incentives	Material and financial incentives	National budget documents
The priority setting process	Stakeholder participation	Number of stakeholders participating, number of opportunities each stakeholder gets to express opinions	Observation at/minutes of meetings, media reports, special reports
	Use of clear priority setting process/tool/methods	Documented priority setting process and/or use of priority setting framework	Observation at/minutes of meetings, media reports, special reports
	Use of explicit relevant priority setting criteria	Documented/articulated criteria	Observation at/minutes of meetings, media reports, special reports
	Use of evidence	Number of times available data is resourced/number of studies commissioned/existing strategies to collect relevant data	Observation at/minutes of meetings, media reports, special reports
	Reflection of public values	Number and type of members from the general public represented, how they are selected, number of times they get to express their opinion, proportion of decisions reflecting public values, documented strategy to enlist public values, number of studies commissioned to elicit public values	Observation at/minutes of meetings, public values study reports,
	Publicity of priorities and criteria	Number of times decisions and rationales appear in public documents	Media reports
	Functional mechanisms for appealing the decision	Number of decisions appealed; number of decisions revised	Observations/minutes at meetings, media reports, special reports
	Functional mechanisms for enforcement	Number of cases of failure to adhere to priority-setting process reported	Observation at/minutes of meetings, media reports, special reports
	Efficiency of the priority setting process	Proportion of meeting time spent on priority setting, number of decisions made on time	Observation at/minutes of meetings, annual budget documents, health system reports
	Decreased dissentions	Number of complaints from stakeholders	Meeting minutes, media reports
	Allocation of resources according to priorities	Degree of alignment of resource allocation and agreed upon priorities, times budget is re-allocated from less prioritised to high prioritised areas, stakeholder satisfaction with the decisions	Annual budget reports, evaluation documents
	Decreased resource wastage/misallocation	Proportion of budget unused	Budget documents, evaluation reports
	Improved internal accountability/reduced corruption	Number of publicised resource allocation decisions	Evaluation reports, stakeholder interviews, media reports
	Increased stakeholder understanding, satisfaction and compliance with the priority setting process	Number of stakeholders attending meetings, number of complaints from stakeholders, percentage of stakeholders that can articulate the concepts used in priority setting and appreciate the need for priority setting	Observation at/minutes of meetings, special reports, stakeholder satisfaction survey, media reports, stakeholder interviews, evaluation reports
Implementation of the set priorities	Improved internal accountability/reduced corruption	Number of publicised resource allocation decisions	Evaluation reports, stakeholder interviews, media reports
	Strengthening of the priority setting institution	Indicators relating to increased efficiency, use of data, quality of decisions and appropriate resource allocation, percentage of stakeholders with the capacity to set priorities	Training reports, evaluation reports, budget documents
	Impact on institutional goals and objectives	Percentage of institutional objectives met that are attributed to the priority setting process	Evaluation reports, special studies
Priority setting outcomes	Impact on health policy and practice	Changes in health policy to reflect identified priorities	Policy documents
	Achievement of health system goals -improved population health -reduction in health inequalities -fair financial contribution -responsive health care system	Percent reduction in DALYs, percent reduction of the gap between the lower and upper quintiles, percentage of poor populations spending more than 50% of their income on health care, percentage of users who report satisfaction with the healthcare system	Ministry of Health documents, Demographic and Health Surveys, commissioned studies
	Improved financial and political accountability	Number of publicised financial resource allocation decisions, number of corruption instances reported, percentage of the public reporting satisfaction with the process	Reports, media reports, interviews with stakeholders
	Increased investment in the health sector and strengthening of the health care system	Proportion increase in the health budget, proportion increase in the retention of health workers, percentage of the public reporting satisfaction with the health-care system	National budget allocation documents

Non-italics = immediate parameters (i.e., can be assessed in a budget cycle); Italics = delayed parameters (i.e., assessed over a longer term, such as the planning cycle)

2.2. Study design

This evaluation was part of a large qualitative study that aimed to describe and evaluate priority setting for health in Uganda, using six casestudies at the national and district levels, including emergencies (Kapiriri & Be LaRose, 2019), non-communicable diseases (NCDs) (Essue & Kapiriri, 2018), HIV/AIDS (Kapiriri et al., 2019), maternal, neonatal and child health (MNCH), new technologies (Wallace & Kapiriri, 2017) and health systems.

A mixed methods approach, involving document review and interviews, was used. We critically reviewed policy documents that outlined the strategic focus and implementation of health initiatives in Uganda. The documents were accessed through the Uganda Ministry of Health webpage. These documents covered the period 2005–2015. Two National Health Policy statements and three Health Sector Strategic Plans (HSSPs) and annual health sector review reports, which covered the period of the evaluation, were reviewed.

Key informant interviews were conducted between 2013–2015 with global, national, and sub-national stakeholders who were involved in health system planning and implementation as well as in influencing priority setting for the health system. The study team, in collaboration with local partners at the Ministry of Health, compiled a preliminary list of key policymakers, development assistance partners (DAP), and other stakeholders. A snow-balling strategy was used to recruit additional stakeholders. The interviews were conducted by telephone or face-to-face by trained Ugandan research assistants, using a pilot-tested interview guide. Interviews lasted approximately 45 minutes and were audio-recorded with permission from the respondents. The interviews used open-ended questions that aligned with each domain of the evaluation framework. The study received ethics approval from the investigators’ institutions.

A total of 67 interviews were conducted and analysed to evaluate health system priority setting from the perspectives of global, national, and sub-national stakeholders. This included (a) 26 interviews with stakeholders working at the national and subnational levels of the health system in either government departments (e.g., Ministry of Health and the Department of Planning) or NGOs and development assistance partner (DAP) organisations (identified as “HS” in participant ID); (b) nine interviews conducted with stakeholders who worked at the global level in bilateral or multilateral organisations (e.g., USAID and UNICEF; identified as “G” in participant ID); and (c) 32 interviews from priority setting case studies from the authors’ larger study (identified as “oHS” in participant ID; national or sub-national stakeholder (n = 22); global stakeholder (n = 10)).

2.3. Data analysis

The policy documents were analysed with a focus on understanding what policy commitments were made and what evidence was used to substantiate priority setting decisions. This information was evaluated using the lens of the framework.

The interviews were transcribed verbatim and analysed using NVivo 11. The initial analysis involved describing and evaluating the various aspects of priority setting in the interview transcripts against the parameters for priority setting as described in the framework. Analysis included coding interview data into relevant themes (termed “nodes”) as per the interview questions, and subsequently composing discrete narratives based on each node. Each discrete narrative was then synthesised, refined, and collated into a report. A breath of perspectives were included to generate a rich qualitative dataset. These data were explored to discern how decisions were made (or not) and gain insights on the factors that were influential in setting agendas in this context. As a standard interview guide was used in all interviews, concepts that arose were explored from multiple perspectives, including at the level of respondent (i.e., how was this issue discussed by other global respondents) and where relevant, between levels of respondents (i.e., how was this issue discussed by global respondents versus national and sub-national respondents). The range of perspectives were reflective of the experiences at the global, national, and sub-national levels so that it was possible to explore, for example, how decisions made at the global level or national levels impacted those working at the subnational level where implementation occurred.

In addition, the raw transcripts from disease-specific cases studies (i.e., HIV/AIDs, MNCH, new technologies, NCDs, emergencies) from the authors’ larger study were also analysed with a lens of health system strengthening, and these data were incorporated in the main analysis.

Three researchers were involved in the analysis. They met regularly to review and discuss the findings and clarify assumptions. The initial report of the findings was presented to Ugandan health system experts, policymakers and collaborators at a dissemination meeting held in Kampala in May 2018; their feedback on the interpretation of the results was incorporated in developing the final report as well as this paper.

All framework parameters were evaluated, with the exception of public values and resource wastage as they could not be assessed using the available data. The results are described according to the five domains and corresponding parameters of the framework.

3. Results

3.1. Priority setting context

The political, social, cultural, and economic contexts influenced the prioritisation process and the implementation of the priorities (Table 2). The main issues identified were the politically driven changes within the Ministry of Health leadership, which, as discussed below, impacted the implementation of some programs. In addition, a conservative political ideology held by many in government as well as by the public was thought to impact the socio-cultural context for priority setting. The example discussed was the passing of the 2014 Anti-Homosexuality Bill which resulted in several donors withdrawing funding and exemplified how reliant the government was on conditional foreign aid.

Table 2.

Evaluation summary of priority setting for health system strengthening

	Parameters of successful priority setting	Key findings
Contextual factors	Political, economic, social and cultural context	Politically driven changes within the Ministry of Health leadership impacted the implementation of some programs. A conservative political ideology was thought to have jeopardised donor assistance.
Pre-requisites	Political will	Weak institutional capacity and willingness to defend and pursue priorities. Scarce resources (financial, human resources, infrastructure). Legitimate priority setting institutions that were deemed credible and inclusive of a diverse range of stakeholders, with the public perceived to be under-represented. Consultation inbuilt in the processes perceived to incentivise buy-in for priorities and implementation. A recognition that greater incentives were required to attract and retain staff.
	Resources
	Legitimate and Credibleinstitutions
	Incentives
The priority setting process	Fairness	The process lacked fairness.
	Stakeholder participation	Participatory process, with under-representation by some stakeholders. Evidence that power imbalances between stakeholders were not well managed in the process, resulting in some stakeholders dominating the process as well as parallel priority setting outside of the sector.
	Use of clear prioritysetting tools/methods	Transition from a technically driven process towards a process of collaboration and consensus-building. Iterative and hierarchical processes; no explicit discussion of formal tools or methods used systematically.
	Use of explicit relevant priority setting criteria	Explicit criteria discussed: epidemiological need; equity; cost-effectiveness; burden of disease; appropriateness of solutions; DAP interests or core policies; domestic political interests; consistency with global agenda Additional criteria: consistency with national health strategies and/or other prevailing policy documents; donor duplication
	Use of evidence	Evidence highly valued Types of evidence available to be used in the process: expert advice health and health system indicators local, regional and global data Problems identified with the evidence base: not effectively or regularly collected not consistently used during planning processes relevant evidence was missing from existing systems (i.e., HMIS) contemporary data was lacking
	Reflection of public values	-
	Publicity of priorities and criteria	Policies, strategies, and decisions made by the Ugandan Government were supposed to be disseminated and/or publicised across the health sector but in reality, this policy dissemination was often poorly executed, failed to include all relevant stakeholders, or did not occur at all. The media had not been engaged extensively to publicise policies and strategies.
	Functional mechanisms for appeal and enforcement of the decision	The mechanism responsible for ensuring accountability and district-level adherence to the national strategy were ineffective in practice. Reallocation of resources occurred regularly. No formal appeals discussed
	Efficiency of the priority setting process	Inefficient priority setting processes
	Dissentions	Several dissentions discussed
	Public understanding of and confidence in the process	There was a perception among interviewees that the public, by not being informed about priorities, were disempowered and limited in their ability to demand accountability from planners and DAPs.
	Allocation of resourcesaccording to priorities	Main challenge discussed was that resources were often pre-allocated, making Ministry-level priority setting seem irrelevant and not feasible. Funds were often reallocated at the point of implementation and there was a lack of transparency about why and where the funds were reallocated
	Decreased resource wastage/misallocation	-
	Increased stakeholder understanding, satisfaction and compliance with the priority setting process	Broad consultation processes fostered stakeholder understanding and buy in for the decisions. Dissatisfaction that the implementers (districts) and the beneficiaries (public) were under-represented in the planning processes DAPs were thought to shape 70% of the health sector agenda due to inflexible funding interests and rigid specifications
Implementation of the set priorities	Impact on internal, financial and political accountability and corruption	Weak financial accountability, linked to a history of corruption that negatively impacted how DAPs invested in the health system
	Strengthening of the priority setting institution	The quality of decision making for health sector priorities and resource allocation was thought to be sub-optimal
	Impact on institutional goals and objectives	Implementation of priorities to support health system strengthening was thought to rarely be achieved or successful
Priority setting outcomes	Impact on health policy and practice	-
	Achievement of health system goals	Several key health indicators were thought to be achieved
	Increased investment in the health sector and strengthening of the healthcare system	Evidence of increased investment in the health sector but not in the areas of defined need or aligning with the priorities of the sector

Non-italics = immediate parameters; Italics = delayed parameters; - = not assessed; SH: Stakeholder.

3.2. Pre-requisites

3.2.1. Political will

There was a perception among interviewees that the health system was a lower government priority, relative to other sectors. This was explained by some as a “crowding out effect” caused by large foreign investment in global health priority initiatives. With ongoing and substantial development assistance resources invested directly in health and health-care services, the government focused domestic resources in other sectors such as infrastructure and agriculture. The challenge, as noted by several interviewees, was that this meant that the government had never met its commitment to allocate 15% of GDP to health. Describing this lower prioritisation of the health system, both a national and global interviewee commented:

“The Health sector at national level since the late 2000s has not been a priority … both at a political level, but also at the technical level … because they know that there’s a lot of global health initiative money”. (HS04)

“the larger the amount of money that donors like ours put into a particular problem, the more it crowds out partner countries’ allocations to that same problem. They go spend their money on something else”. (G01)

3.2.2. Resources

There was consensus that inadequate financial and non-financial resources were the most pressing barriers to effective priority setting and implementation of priorities, often rendering priority setting a “senseless activity” (HS01). These issues were also thought to pose barriers to developing sustainable financing. This was discussed in terms of financial, human and infrastructure resources.

3.2.2.1. Financial resources

Participants were concerned that low levels of domestic health system investment had placed the Government in a disempowered position. The Ministry of Health was thought to be disempowered as funds tended to be allocated off-budget and often directly to the districts, resulting in weak governance and a lack of authority in the Ministry of Health to influence, in a meaningful way, how funds were used. One interviewee commented:

“It’s good that funding is decentralized but, bad that the Ministry of Health is the only organization in a position to take oversight of what’s happening across the country [but] it’s not pulling the purse strings”. (HS03)

Global interviewees also stressed that off-budget investments were not the preferred way of working in LICs, for example,

“because of the poor governance and the corruption usually money is tied to certain conditions and to a certain schedule which is a catch 22 situation … we are all routing [resources] through different NGOs and as a fallback choice not the default choice.” (oHS07).

While in general most stakeholders commented that vertical, disease-focused funding models were less preferable, there was a recognition of the potential for shared benefits by way of strengthening human resource capacity, infrastructure and access to medicines in the system (HS05) Other issues that were identified as challenges to developing sustainable health financing included a lack of consensus on health financing and as a consequence, divergent opinions impeded the advancement of health financing reforms. There was also a perception that the Ministry of Health did not view health financing as an urgent priority. Finally, practical and logistical issues were also identified, for example, the challenge of tax collection in a setting where population-wide data were still incomplete.

3.2.2.2. Human resources

Human resources for health were perceived by most interviewees as continually neglected, yet critical for strengthening the health system:

“the key priority … human resources. That one is appreciated but not tackled” (HS11)

At the national level, interviewees discussed an ongoing shortage of technical expertise to support priority setting. A shortage of skills in quality assurance, finance, and economics was thought to have negatively impacted the system’s ability to plan for the future. At the district level, the lack of human resources was thought to impede planning, with few individuals able to champion effectively for local issues identified in their districts.

Interviewees also discussed the challenge of unmotivated health system workers who were unable to progress health system agendas. This was thought to manifest in high rates of absenteeism, presenteeism (i.e., attending work but not performing at one’s full capacity) and attrition.

A lack of research capacity was also identified as a resource gap, with one interviewee commenting:

“If one wants to strengthen health systems … the issue of domestic research capacity is quite important and it is unfortunately a stepchild … That’s not happening a lot. It’s mainly driven by outsiders”. (G04)

3.2.2.3. Infrastructure

Both national and global interviewees stressed inadequate infrastructure posed a critical challenge for implementation. Inaccessible health facility locations, non-functional buildings (i.e., power outages, water shortages) and inadequate supplies of necessary medicines, materials, and technologies were cited as examples that compromised efforts to strengthen the health system.

3.2.3. Legitimate and credible institutions

Weak institutional capacity, which was in part a consequence of what was described as a “culture of aid dependency”, was thought to have impacted the political leadership and the commitment to drive national efforts to strengthen the health system. Several interviewees discussed “health” as a global priority rather than a national priority and this was seen to perpetuate the dependency on foreign aid. Furthermore, aid dependency was thought to reinforce conditions that allowed the DAPs to maintain a leadership role in priority setting.

The Ministry of Health and its health development partners should have been guiding the prioritisation process, which was governed by the National Development Plan and National Health Policy (Box 2). The processes were seen as legitimate, credible, and consultative by the interviewees. However, several interviewees discussed the ways in which the legitimate processes were often compromised by multiple, uncoordinated and at times competing strategic plans and vertical programs:

“A bigger part is internal. But that does not mean that some of these activities by the donors don’t fragment here [the health system] – they can impact the country, especially when countries have weak institutional capacity. They [can] create parallel arrangements which really make it hard for the country”. (HS01)

In addition to this, without meaningful engagement of the Districts and end users with the Ministry in its role as the main institution for setting priorities, there was some concern that this impacted the extent to which priorities were viable and implemented; some priorities were seen to conclude in the Ministry without weighting given to alternative priorities.

3.2.4. Incentives

Wide stakeholder consultation was formally inbuilt in the processes through which priorities were developed and this was thought to foster buy-in and collaboration between diverse stakeholders. It was also believed to incentivise stakeholders to not only engage in the process but also to support implementation of the priorities.

At the implementation level, several interviewees commented that better incentives would attract and retain staff which were being lost to competing agencies and countries that offered better employment conditions.

3.3. The priority setting process

3.3.1. Stakeholder participation

Several interviewees commented that priority setting and policy formulation had transformed from a mostly technical process to one that emphasised consultation and collaboration with diverse stakeholders as well as technical assessments of evidence (Box 2).

Box 2.

Overview of institutions and consultations that guide priority setting in Uganda

	Stakeholders	Contribution to priority setting
	Ministry of Health Departments and Planning Directorate	Together with the Planning Directorate, each of the 6 Departments of the MOH consult relevant evidence and literature, and submit their respective goals and priorities for the next five years to the Senior Management Committee Priorities are determined based on the department’s data collection, consultations with various stakeholders (DAPs, communities, district health offices, experts, health providers, etc.), evaluation reports, etc.
	Joint Review Mission	An annual review that includes diverse stakeholders – domestic and foreign, ministry and non-ministry, district- and national-level, DAPs to assess the performance and impact of the past year’s health and development activities, the successes and challenges experienced, and the appropriateness of current allocations. The review’s findings are summarised in the “Aide Memoire”, which delineates the priorities of the health sector for the upcoming financial year
	Ministry of Health Senior Management Committee	Comprised of the heads of all MOH Departments Amalgamates departmental plans into a comprehensive internal technical document and submits to the Top Management Committee
	DAPs and other external stakeholders	Various stakeholders are sent a draft of the departmental plans for review and comments
	Ministry of Health Top Management Committee	Comprised of Directors, Commissioners and the Minister Receives, revises and ultimately approves of the internal technical plan
	Health Policy Advisory Committee	Sits monthly. Reviews the working document and provides final guidance All planning documents from the MOH reviewed by this committee; DAPs have a significant amount of control over the final product
	Ministers of Other Departments	If relevant, input sought from other Ministries affected by the strategic plan (i.e. Education, Agriculture, Water and Sanitation) prior to presenting the document to cabinet in order to gain cross-Ministry consensus and buy-in
Policy Arena	The Sector WorkingGroups	Works with a number of technical working groups. Provides advice on broad strategic areas identified in the planning documents and specific funding amounts to be allocated to each district in accordance to a formula that accounts for criteria such as: population size, infant mortality, burden of disease, distance from city centre, etc. Ultimately produce a budget framework paper, that is passed to Senior Management through to Cabinet. Play an important and continual role in monitoring budget allocations
	Cabinet andParliament	Cabinet provides final review and sign off of all planning and budget policy proposals.

The respondents identified several stakeholders involved in health system priority setting: members of the Ugandan Government (including politicians, parliament and cabinet members, staff from the ministries of Health and Finance); DAPs; technical experts (i.e., academics); civil society organisations; private-for-profit and faith-based private not-for-profit organisations; district and community representatives (including health service providers, managers, and operators of health facilities) and to a much lesser extent, the public.

The legitimacy of DAPs role in influencing priority setting was questioned by many of the national interviewees who suggested that priority setting should be the remit of the local leaders.

“ … it is illegitimate [the role and influence of DAPs]. … We are the people in charge. We have the data, we set the priorities so the moment you start guiding me into a direction, honestly that’s not a legitimate way to do things”. (HS11)

A sub-national respondent commented

“You know those, our partners are opportunists. They get opportunity where we have the gaps. … And the technical working groups always incorporate the partners – just like we are incorporating partners at this level”. HS20

Interviewees also discussed the role that politicians had in derailing the selection and implementation of prioritised interventions because of proclamations and commitments made during election campaigns. They were thought to regularly use their platform to draw attention to their specific interests, regardless of whether these were national priorities identified through the formal planning channels.

There was consistency among interviewees that while the processes were consultative, they did not support meaningful engagement of all stakeholders, especially the district officers who should implement the identified priorities:

“the planning is done very much in silos and with very little buy-in of the implementing people. So much of the planning is centralized. When the district people for example, are called in it’s mainly you know, a lip service thing they don’t really have much say at the priority setting table”. (HS07)

Another sub-national respondent commented on the influence of national health priorities on District health priorities and implementation:

“The national health priorities and district health priorities are similar because most of the priorities we pick must align with the national health policy. For example last year they said part of the money should be put on maintenance. That we should not start new projects that were needed. We had to follow that ....So by and large our local health priorities are also influenced by the national priorities” (HS21)

3.3.2. Use of clear priority setting process, tools, and methods

The interviewees discussed an explicit and hierarchical structure of governance and management that was in place to support the development of the Health Sector Strategic Investment Plans (HSSIPs) (Box 2), for example, “ … if we look in the [HSSIP], governance and management structures are starting right from village to the senior top management … each structure at each level looks at what is planned and then they recommend to the higher so that by the time you get the aggregate that is going to be taken to the senior top management … everybody has had at least a place where they can articulate what they think”. (HS09)

Several global interviewees discussed their longstanding commitment to build local capacity to independently make priority allocation decisions using robust methods such as data on the burden of disease and cost-effectiveness of interventions, for example, one interviewee descibed the training they provide to countries as covering the following:

“look at the problems, find their epidemiological demographic severity and magnitude, look at the burden of that disease and then look for available solutions and find the cost effective ones that apply to your situation and turn them on and scale them up”. (G01)

Despite this training, several national stakeholders highlighted that this approach was not systematically followed in practice.

3.3.3. Use of explicit priority setting criteria

National and global interviewees all discussed the relevance of considering the burden of disease and cost-effectiveness of available interventions as guiding principles for priority setting. There were distinct areas of divergence between interviewees about other relevant criteria and the extent to which they should be valued in the decision-making processes.

Most DAPs discussed the high value given to their organisational priorities, which often aligned with advancing global health priorities. Many also discussed the importance of ensuring that health system priorities addressed inequities in access to services and in health outcomes. Only one global DAP discussed the importance of minimising DAP duplication as a key criterion when working in partner countries.

All national interviewees agreed that DAP priorities were an informal criteria that often influenced the selection of priorities, given the health system’s reliance on donor assistance. One national interviewee stated that up to 70% of the health system agenda had been shaped by donor priorities. Another national interviewee stated:

“So [DAP] will come in and fund H disease, because H disease is big and it makes them look good on the global scene. So it is development partners’ priorities” (HS15)

For the national stakeholders, first and foremost, individuals believed the health system priorities should align with the commitments or recommendations made in the national policy documents. They also emphasised the importance of considering the appropriateness of solutions in their local context and the capacity for implementing priorities, including the availability of human resources and technology.

Despite having formal structures in place to support a consultative process for deliberation of relevant criteria, priority setting was thought to proceed in an ad hoc manner, allowing the values and politics of the most powerful stakeholders to dictate which criteria were considered and used in the identification and implementation of priorities. This was thought to result in “a laundry list of priorities … with no sense of prioritization” (HS07). One global interviewee summarised:

“The more I work in developing countries the more I realize this [priority setting] is really a political process. There is no intervention that doesn’t have evidence behind it … and one will say this is cost effective, this other one is also cost effective, I can save this many lives, that can save so many lives … by and large it’s a political process”. (G02)

A sub-national interviewee commented

“So the politicians also now come in … and if you don’t have a good backing of your priorities, then they have their own political priorities”. (HS20)

Participants also noted a more nuanced approach to priority setting, specifically one in which political will is both complemented with and informed by technical knowledge about program efficacy. In this way, evidence and research are used to inform political actors about which health concerns were priorities.

3.3.4. Use of evidence

Evidence was highly valued as an important input for priority setting by all interviewees. They reported that decisions were based on both qualitative and quantitative information. Qualitative information was accessed through the extensive consultative meetings with a range of diverse stakeholders. These meetings generated knowledge on health system performance, action areas and interventions. One interviewee explained:

“But chiefly we do the consultative process … We still hold consultative meetings through the gathering of information that would bring stakeholders from districts, from the donor community, civil society and … pick as many topics as possible to discuss when you have the political will also there in the districts. They tell you exactly what they hear from their people”. (HS11)

Quantitative evidence was obtained from local and international sources. Locally, evidence on health and the health system was obtained from the HMIS databases, global burden of disease estimates, (which were acknowledged to be outdated), sector reviews, research studies and commissioned rapid epidemiological assessments. One interviewee explained:

“they try as much as possible to make it evidence based. The burden of disease study that was done for Uganda in 1993 still influences very much the priorities and the interventions therein that are used for the health sector”. (HS15)

At the regional and international level, evidence and lessons were gleaned from relevant case studies on successes and failures experienced by similar African countries, for example,

“ … we can also look at the East African community in general how are the other member countries performing? Like in Rwanda, how have they improved their health indicators? And we see if we can cut out those kinds of interventions, maybe we can improve”. (HS05)

Interviewees discussed their concern about the lack of contemporary data to guide decision-making. Identified evidence gaps included locally derived cost-effectiveness estimates, recent burden of disease estimates, and good quality data. A national stakeholder commented:

“But you don’t even know [who’s] there, who are you planning for? What we usually have [are] numerators … Slowly people start confabulating, they make up denominators one way or another … if we could deal with health information we definitely [would be] able to act better to implement better policies”. (HS12)

Interviewees also discussed the value of being able to use a “bottom up” process where possible, such that data generated at the program delivery level, where front line staff operate, should be the primary source of the information used to guide future priority setting processes:

“we say this is a bottom up … if you follow me correctly because we don’t generate the data … the data is generated from the service delivery level that is health facility sub-county and those levels also have players. The Health Unit Management Committee, the local politicians, (they all generate issues)”. HS22

3.3.5. Publicity of decisions and criteria

Interviewees elaborated on the ways in which policy decisions were intended to be disseminated across the country, including “regional workshops”, policy documents and policy and implementation guidelines, in addition to the reports detailing the strategic directions (e.g., the HSSPs) and the media to reach the public.

However, effective, accurate and inclusive policy dissemination was identified as sub-optimal, or neglected altogether. Interviewees reported that often the public and the districts did not access this information. One stakeholder linked the inadequate publicity of priorities to failed implementation at the district level, explaining that district facilities often proceeded with “business as usual”, which could be inconsistent with the national strategies:

“it [the HSSP] was very poorly disseminated most of the copies are in the Ministry of Health rather than at the grassroots where the actual implementation is happening. There is no translation of what the priorities mean for the planners at implementation level so in a sense they [implementers] end up doing things business as usual rather than actually being informed by the plan”. (HS07)

Several national interviewees commented that there had been missed opportunities to use the media to publicise policies and strategies.

3.3.6. Functional mechanisms for appeal, enforcement, monitoring, and evaluation of decisions

The mechanisms for appeal and enforcement of decisions were not discussed. In terms of monitoring and evaluation, interviewees indicated that supervision and follow-up at the district level to ensure that they were operating according to set priorities, was very weak. Several discussed the challenge of having inadequate systems in place for effective monitoring and evaluation of implementation, which had the effect of rendering priority setting ineffective. For example,

“ … Again this shows that people who believe this is important don’t go to follow up. You need to be able to say if this is important you follow it ….It should go with the budgeting process and the implementation process to make sure that you follow and ensure that what your plan says gets implemented” (HS01)

3.3.7. Efficiency of the priority setting processes

While the HSSP meetings were thought to be conducted efficiently, the lack of- or fragmented implementation of the priorities made the process inefficient. Interviewees discussed off-budget donor support, which was often poorly accounted for by the Ministries of Finance and Health. DAPs were also thought to introduce parallel priority setting initiatives which resulted in either duplicity and hence resource wastage; or the implementation of programs that did not align with the priorities identified in the HSSP. This was thought to ultimately undermine the efficiency of priority setting.

3.3.8. Dissentions

Interviewees discussed dissatisfaction about the lack of clarity on how criteria were valued by stakeholders and the lack of a systematic approach for assessing the various criteria. This often resulted in a long list of unfunded priorities and the adoption of priority areas that were strongly advocated for by the most powerful and vocal players. Speaking about their dissatisfaction with DAP influence on priority setting, one national interviewee explained:

“the global initiatives that come with specific areas of interventions, there our government doesn’t have much say on what interventions should be implemented … [they say] ‘our money is for disease X, Y and Z. Put it there or leave it’. So that are areas where … they go a little bit off the spirit of the partnership”. (HS04)

There were also challenges identified with reaching consensus among stakeholders and several interviewees discussed their dissatisfaction that some stakeholders were perceived to make decisions and operate relatively independently:

“With every planning process you have multiple voices. Some are more dominant than others and they don’t always agree … issues around the content of what should go in the strategic plan that has caused it to stall because people haven’t agreed … That is not happening, the give and take”. (oHS10)

There was also dissatisfaction with the overall preference in the planning stages for treatment over prevention; reflecting a lack of foresight and proactive planning for the health system.

3.4. Implementation

3.4.1. Allocation of resources according to priorities

The allocation of resources according to priorities often did not happen due to a general scarcity of resources in the system and the ear-marking of resources by Parliament or DAPs. It was not uncommon to find that investment in the health system was not allocated to priority areas. For example,

“So it could be a relatively good budget … but our allocation is rigidly determined by the budget department or Minister of Finance. So that curtails prioritization as we would want. So numerous cases you find a list of unfunded priorities”. (HS11)

“ … the development partners also have their own agenda … for example, we’ve not seen much resources being pooled into non-communicable diseases or reproductive health … yet those are time-bombs for the future”. (HS06)

Working groups had been established to track and monitor the allocation of funding to ensure appropriate use of funds before additional resources were availed. However, several national interviewees felt that these bodies were not working effectively as it was quite common for “slippage” of resources from the defined priorities in the national strategic plans to other unexplained areas.

3.4.2. Internal, financial, and political accountability

Accountability was perceived to be critical to credible priority setting and resource allocation. Some national interviewees discussed the establishment of “sector working groups” as an element of priority setting that were intended to improve budgeting processes, resource allocation management, and internal accountability. However, the lack of information to facilitate effective oversight and monitoring was thought to contribute to reduced accountability. Other respondents alluded to the absence of functional mechanisms to ensure accountability which resulted in an inappropriate or unintended use of funds.

The dysfunctional accountability mechanisms were thought to have, in part, contributed to the previous cases of corruption within the health system that had a consequence of weakening donor trust in contributing towards sector-wide budgets. This was thought to be a reversion from the progress that had been made in funding and partnerships with DAPs. One national interviewee explained:

“ ...the corruption scandals in Uganda have undermined donor confidence in the sector-wide approach especially when it comes to pooled funding … at the moment donors are looking for ways of channeling aid within the principals of the Paris Declaration but in a manner that does not expose their money to risk of misappropriation … ” (HS07)

A lack of effective oversight measures, indicators, and technical expertise to ensure monitoring of how and to what extent priorities were implemented and were also discussed as factors that compromised accountability.

In addition, several national interviewees discussed the impact of what was described as “the disempowered public”, who, due to a lack of awareness of their rights to health and the contents of the national strategies, were unable to demand accountability to ensure that population health needs were being met.

3.4.3. Stakeholder understanding, satisfaction, and compliance with the decisions

Interviewees thought that the broad consultation processes fostered stakeholder understanding and buy-in for the decisions. However, key stakeholders such as the implementers (districts) and the beneficiaries (public) were under-represented in the planning processes. Limited involvement of districts impacted their compliance:

“ … when the document comes they [the districts] have no interest … the document just sits there in the office gathering dust while they go ahead with their annual planning as usual”. (HS07)

3.4.4. Strengthening of the priority setting institution

As the main priority setting institution, the Ministry of Health had established processes, defined relevant criteria and recognised the value of data and evidence, including in the form of stakeholder advice. However, interviewees felt the quality of decision-making for priorities and resource allocation was too heavily influenced by actors external to the Ministry of Health. This led to the fragmentation of the health system, and undermined the legitimacy and credibility of the Ministry of Health as a priority setting institution.

The priority setting institution was further compromised by poor longevity of national strategies and frequent priority changes due to discontinuity in political leadership and high turnover of staff and politicians. For example,

“If you have one Minister who has been thoroughly prepared into one concept and if that Minister is reshuffled before the thing becomes institutionalized … you have to go back and start afresh. And many of them have a feeling that the other one went because of some of these bad policies. So … they fear to touch them”. (HS09)

3.4.5. Impact on the priority setting institutional goals and objectives

Most interviewees discussed the impact of ineffective priority setting as directly affecting the extent to which the system was able to achieve the goals and objectives as set out in the HSSIP. For example,

“You have the vertical programs … having their own strategic plans. These strategic plans for the most part are not aligned to the time horizon of the HSSIP … [and] they’re all planning for the same health system without taking into account the effect on that health system. So they come up with these very ambitious plans which the health system cannot deliver”. (HS07)

3.5. Outcome and impact

3.5.1. Achievement of health system goals

There were noted achievements in many key health system performance indicators, for example, achieving the target neonatal mortality rate, nearly achieving the under-five mortality rate, broader HIV control, successful implementation of immunisation programs and the establishment of an NCD office to coordinate NCD control measures. Interviewees acknowledged that while these achievements collectively pointed to a stronger health system, they actually resulted from achievements in individual program areas, as opposed to a systematic approach to strengthen the health system.

While it was not possible to formally assess fairness in financial contributions and meeting of public expectations, some national interviewees discussed the ongoing impact of out-of-pocket costs for individuals, despite the removal of user fees at first level government health facilities in 2001.

3.5.2. Increased investment in the health system and strengthening of the health-care system

While there had been increased funding in the health system, the majority of interviewees acknowledged that this investment was rarely spent on health system strengthening and was not often spent on the areas the government had identified to be most urgent to achieve this goal. Hence, while there was increased investment, it may not have contributed to meaningful health system strengthening. For example,

“ [DAPs] say ‘We want to start up a National Public Health Institute’ and yet that is not necessarily what was agreed in the strategical plan. And so there is lots of money coming into the health system but it’s not necessarily going to the priorities that we have chosen as a sector … . A donor to remain relevant comes up with some idea, runs with it, brings in loads of money but it’s not necessarily addressing the problems on the ground”. (HS07)

4. Discussion

The Kapiriri and Martin framework was used to describe and evaluate priority setting for health system strengthening in Uganda. To the authors’ knowledge, this is the first study to systematically describe and evaluate how health system strengthening has been prioritised and implemented across the health system of an LIC. The mixed methods design used multi-perspective stakeholder interviews and allowed us to explore the impact of priority setting decisions across the health system, including at the level where the decisions were ultimately implemented (or not).

This evaluation revealed that there were key strengths of the planning processes, including a shared recognition among stakeholders that the Ministry of Health had a legitimate mandate for setting priorities for the health system; a defined, evidence-based and consultative process to govern the development of the health system’s strategic and priority areas; some consistency among stakeholders in the relevant criteria for evaluating competing priorities and some existing capacity for conducting technical assessments of competing priorities. But the findings also identified a need to streamline and better coordinate the multiple, parallel planning processes that are undermining the Ministry of Health’s role in leading priority setting.

One potential solution to strengthen priority setting in Uganda could be through the establishment of an independent priority setting institution that sits at an arms-length and independent from government, similar to models that have long been used in high-income countries such as Norway, the United Kingdom, Canada, and Australia. Such independent institutions have been shown to support more rational and fair decision-making (Ham & Robert, 2003), and can contribute to strengthening and championing of fairer priority setting.

This evaluation revealed that Uganda has many of the requirements that demonstrate its readiness to establish an independent body to institutionalise a more systematic approach to priority setting for health (Li et al., 2016). The independence of such an institution could help to rebuild trust with DAPs, ensure accountability in decision-making and potentially, strengthen oversight for implementation. However, its success will hinge on ensuring that the assessment process remains valued by all stakeholders and that the recommendations that come from this body are well aligned with the government’s policy and planning cycles (Glassman et al., 2012; Tantivess et al., 2017; Yothasamut et al., 2009). It will be important for evaluation to be a core element of the priority setting processes within such institutions. Evaluations could be used as an opportunity to track and compare against agreed upon benchmarks. To be most meaningful, the evaluations should be participatory and inclusive of stakeholders at all levels of the system, including the public.

The implementation challenges discussed in this evaluation cannot be solely attributed to inadequate resources. Foreign investment in Uganda’s health system and in health system strengthening has actually been steadily increasing over the last decade (Stierman et al., 2013), which is consistent with global trends (Borghi & Chalabi, 2017; Dieleman et al., 2016). The disease burden has also been changing, including the health system demands associated with managing a double disease burden of infectious and noncommunicable disease and this has placed additional demands on the available resources (Essue & Kapiriri, 2018). But this evaluation has reinforced the challenge of lacking full control over resources, and the impact that this has on efforts to effectively strengthen the health system. Ear-marked funding, parallel priority setting processes that did not engage with the Ministry of Health-led processes coupled with limited national capacity to increase domestic investment in health meant there was limited governance or capacity for oversight. It also compromised the Ministry of Health’s capacity to work across the whole system to progress a health system strengthening agenda as defined in the HSSPs.

This evaluation revealed a clear focus on health system support rather than health system strengthening. Priority setting in Uganda was focused on the improvement of existing services (i.e., improving outputs by increasing inputs) as opposed to prioritising comprehensive changes to performance drivers (e.g., policies and regulations), organisational structures, and relationships across the health system to encourage changes in behaviour and foster more effective use of resources (Chee et al., 2013); the sort of changes required to progress the path to UHC and to achieve SDG 3. The impact of this emphasis on supporting existing services was seen in the health system goals that were achieved. Most achievements remained linked to siloed areas, for example, HIV/AIDs and child and maternal health (The Republic of Uganda and Ministry of Health, 2014, 2015), stemming from long-standing, ongoing, and program-specific donor assistance. Siloed investments in each of these areas work to strengthen elements of the health system; however, the evidence is mixed on whether this siloed approach has a collective impact on strengthening the health system (Grépin, 2012; Stierman et al., 2013). Nonetheless, this is a fragmented and passive approach that has long been criticised as hampering the development of sustainable health systems (Luboga et al., 2016). There remain missed opportunities in Uganda to better coordinate resources across silos to improve efficiencies within the health system, for example, through use of shared human resources, medicine procurement practices and diagnostic services (Luboga et al., 2016; Stierman et al., 2013) This cannot be achieved without greater flexibility and harmonisation in the conditions stipulated by DAPs, who are also facing pressure to better coordinate efforts to address the population health disparities that contribute to global inequities as defined in the SDG agenda.

This study has limitations. We relied on interview data and policy documents. However, we interviewed the majority of stakeholders involved in influencing priority setting for the health system who worked at a global, national and sub-national level. While their reported experiences may be subjective, we expect that any bias in the data should be minimal given the number, range in perspectives and depth of interviews completed. We were not able to objectively assess two parameters using the available data: public understanding and confidence in the process and resource wastage and misallocation. Many stakeholders commented indirectly on these parameters but future studies should include methods to incorporate public perceptions and values of priority setting. We completed a retrospective analysis of policy documents and a contemporaneous analysis of current priority setting using interview data. These data provided a snapshot of the strengths and weaknesses of the priority setting processes in place during the evaluation reference period. They contribute to learning in the system and are an important input for guiding improvements in priority setting. This work provides an important reference point that should be used in future comparisons of health system priority setting in order to monitor progress and changes over time.

5. Conclusion

This study makes an important contribution to the health system strengthening and the priority setting literature, by focusing on understanding how a LIC is grappling with prioritising health system strengthening in their context. In incorporating parameters that are relevant to the five key domains of priority setting (context, pre-requisites, processes, implementation, and outcomes), the evaluation moved beyond an assessment of how the priorities were identified to critically examine the nuances and local conditions in place that affected implementation and outcomes associated priority setting. In so doing, the results provide a deeper understanding of why LICs such as Uganda may be struggling to strengthen their health system and where support is required. The lessons gleaned from this evaluation are relevant to other LICs and global actors in their efforts to support the development of sustainable and responsive health systems, in line with SDG 3.

Funding Statement

This work was supported by the Canadian Institutes of Health Research [10558616].

Disclosure statement

No potential conflict of interest was reported by the authors.