Governance Interventions to Improve Public Health Delivery in Uganda
- Rural population
- Urban population
- Citizen satisfaction
- Empowerment
- Health outcomes
- Community participation
- Targeting
While numerous proven and affordable health interventions exist, many low- and middle-income country governments fail to provide them to the public. The researcher conducted a pilot randomized evaluation to test the feasibility of a regular community accountability meeting program and a politician health leadership program, as well as the feasibility of associated data collection activities. Both interventions realized considerable stakeholder participation and engagement, providing a foundation on which to conduct a larger randomized evaluation in the future around the impact of these interventions on health service delivery and health outcomes.
Problema de política pública
Even though many inexpensive, life-saving health interventions have existed for decades, many governments fail to deliver such interventions to the public. While a range of global and local resource misallocations may contribute to these delivery failures, many such health care delivery shortfalls arise in the “last mile” of service provision, where widespread accountability gaps hamper public health facilities in poor countries. These accountability shortfalls include health worker absenteeism, low health worker performance, diversion of supply and medication inventories, and bribery in the provision of services. For example, in Ugandan public health centers, absenteeism hovers around 50 percent. Without addressing the institutional structures of health care systems that underlie these deficiencies in service provision, attempts to improve care at local facilities may be insufficient.
Top-down monitoring, specifically through audits, is one potential way to better align incentives for health care delivery. However, monitoring requires considerable government capacity and resources, in addition to a willingness by the government to establish audit programs. An alternative and potentially more promising approach to improving performance incentives in health care delivery is to enable citizens to exert pressure on local political leaders to better fulfill their governing roles. However, citizens may not have insights into how their leaders are performing and whether politicians are trying to address accountability issues like health care quality.
There is little evidence on effective ways to address political economy inefficiencies, and few interventions focus on citizen-centered approaches to influence government service delivery. Can accountability meetings between citizens and local politicians and health leadership-building for these politicians improve health service delivery?
Contexto de la evaluación
In Uganda, local, democratically elected political leaders have the potential to mitigate management problems by monitoring service provision at government health centers. Uganda’s Local Governments Act tasks these leaders with overseeing government performance within their jurisdictions. However, few local leaders engage in regular monitoring, and their constituents exert little pressure on them to do so. Some of the researcher’s previous work in Uganda suggests that citizens face substantial barriers—such as community organization and access to local institutions—to holding leaders accountable for poor health services. Furthermore, leaders themselves may not have the appropriate training to monitor health services.
Since 2009, the Office of the Prime Minister (OPM), one of the partners for this evaluation, has been implementing a local citizen advocacy initiative called the Baraza Program. The interventions tested in this pilot evaluation will be used to inform potential new accountability strategies to build on the Baraza Program.
Detalles de la intervención
The researcher partnered with OPM and the non-profit Progressive Health Partnership (PHP) to conduct a pilot randomized evaluation to test the feasibility of (1) quarterly community accountability meetings on the quality of health service provision and (2) a health leadership program for local politicians, along with associated data collection activities, to prepare for a longer-term future study.
The researcher divided twenty localities into five strata based on margin of victory of council chairpersons in the 2016 election as a proxy for political competitiveness and quality of democratic functioning. Within each of the five strata, the researcher randomly selected four localities and randomized them into four groups:
- Citizen-Politician Health Meeting: Households in localities in this group attended quarterly meetings between the Council Chairperson and citizens organized by PHP to discuss local health care quality.
- Politician Health Leadership Training: Chairpersons from the localities in this group received training on their rights and responsibilities as well as on how to monitor local health facilities and take action based on their findings.
- Citizen-Politician Health Meeting and Politician Health Leadership Training combined: Households in localities in this group attended quarterly meetings, and their chairpersons received the health leadership program.
- Comparison: Localities in this group received no intervention.
PHP initially implemented both the Citizen-Politician Health Meetings and the Politician Health Leadership Training between January and March 2020 before pausing the activities due to Covid-19. PHP then resumed and completed implementation of the activities from January to August 2022.
PHP’s field interview team collected baseline data from October to November 2019 via household surveys, politician surveys, and health facility quality assessments. The team also collected data via a meeting survey. Baseline survey data collection included a collection of measures for the ultimate outcomes of interest (health care utilization; health care quality; health outcomes; and citizen satisfaction, agency, and empowerment) to inform a future fully scaled study. Due to the disruptions caused by Covid-19, the researchers forewent follow-up data collection. Given that this was a pilot study, the key outcomes assessed included measures of politician and citizen participation in meetings, such as attendance, gender inclusiveness, problems reported, and elite versus non-elite participation. Feedback from citizens on their participation in meetings and feedback from political leaders on the training program’s quality was also collected.
Resultados y lecciones de la política pública
While the small sample size of the pilot study did not allow for the impacts of the intervention to be calculated, the pilot study was a success from the standpoint of implementation. In addition, an associated qualitative focus group study suggested impacts of the meetings intervention on health service delivery. The OPM is committed to collaborating on a larger randomized evaluation in the future to evaluate the impact of community accountability meetings and politician health leadership building on health services outcomes, individual health outcomes, and citizen satisfaction, agency, and empowerment.
Research Implementation: The pilot study generated crucial research materials and protocols that the researcher can use in a future expanded study on the same topic. Research materials created include surveys, in-field data auditing tools, and reporting templates. Data collection was also conducted with little difficulty. The household survey had a refusal rate of only 1.3 percent, while all health centers participated in the health facility surveys. The politician survey demonstrated similar feasibility.
Program Implementation: The implementation team also developed a similar breadth of program materials for a future expanded evaluation, including handbooks for meeting and training facilitators and trainees as well as activity-based reporting templates. The activity-based reporting templates minimized errors in activity implementation during the pilot and helped the implementation team devote additional time to discretionary tasks that contributed to program quality.
Politician Training: The Politician Training intervention group achieved a minimum 80-percent participation rate by political leaders across all of the workshops and fostered a rich discussion among the chairpersons. The participants provided positive feedback after the workshop.
Community Accountability Meetings: Sixteen meetings across eight localities were conducted from January to March 2020 before the intervention was paused due to Covid-19. Following the acute phase of the pandemic, the implementation team resumed the meetings and completed two more quarters of implementation in ten localities. Based on the pre-pandemic data, an average of 19.7 percent of households in each local catchment area sent at least one person to the meetings. Additionally, on average, 110 people attended the meetings. Community members comprised, on average, 96 percent of the attendees and 99 percent of them were from the non-elite portion of the population. Over 60 percent of community member attendees were women, who also made over 60 percent of the comments during the meetings. From a qualitative perspective, the discussions focused on health worker absenteeism and shortages, health centers’ limited operating hours, lack of supplies, charges and bribes for services, and disrespect from health workers.
The results of the pilot were presented to Uganda’s National Monitoring and Evaluation Technical Working Group, a group chaired by the OPM and comprised of representatives from all government ministries, departments, and agencies in the country. The OPM also hosted a panel discussion and dialogue meeting with key governmental and non-governmental stakeholders to discuss the results of the pilot and to generate additional ideas for strategies to enhance local government performance in relation to public service delivery. Many local stakeholders have expressed interest in the pilot’s findings, and the OPM now hopes to work jointly with the researcher on a quantitative evaluation of the interventions.