Attachment J.3-PMI MRA BA Phase II Final Report - 202010230.pdf

October 30, 2020

This Baseline Assessment report was produced at the request of the United States Agency for International Development. It was prepared independently by Wolfgang Stuppert, Beth Anne Pratt, and Esther Sempiira under contract with QED Group, LLC. and on behalf of the USAID/Uganda Learning Activity (ULA) Contract No.

72061719C00003.

FINAL REPORT

Baseline Assessment for an Anticipated PMI Uganda Malaria Reduction Activity

FINAL REPORT

Baseline Assessment for an Anticipated PMI Uganda Malaria Reduction Activity

October 30, 2020

USAID/Uganda Learning Activity Contract No. 72061719C00003

DISCLAIMER

The authors’ views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government.

i

TABLE OF CONTENTS

Acronyms ............................................................................................................................... ii

Executive Summary ............................................................................................................... iii

Introduction

Assessment Purpose

Assessment Questions

Utilization of Assessment Findings

Project Background

The PMI Uganda Malaria Reduction Activity

Assessment Methodology & Limitations

Data collection

Data analysis

Limitations

Findings

Situation Analysis: Summary of Findings from Phase I

Systems for Malaria Response and Coordination

COVID-19 and Its Impact on Malaria Service Delivery and Utilization

Conclusions and Recommendations

References

Annex I: Malaria Reduction Activity Baseline Assessment Scope of Work ................................. I

Annex II: Analysis of Public Health Data ................................................................................. IX

Annex III: Selection of Study Districts ..................................................................................... XI

Annex IV: Key Informant Interviews ...................................................................................... XV

Annex V: IVR Survey ............................................................................................................ XIX

Annex VI: Interview Guides and Survey Tool ........................................................................ XXI

Annex VII: OCA Tool .............................................................................................................. XLI

Annex VIII: Ethical Considerations and Quality Control ....................................................... XLV

Annex IX: Additional Data Tables ...................................................................................... XLVII

Annex X: Sense-Making Workshop ..................................................................................... XLIX ii

ACRONYMS

ACT Artemisinin-based Combination Therapy AHSPR Annual Health Sector Performance Report ANC Antenatal Care APSPR Annual Pharmaceutical Sector Performance Report CDCS 2.0 Country Development Cooperation Strategy DfID

DHO

UK Department for International Development District Health Officer

DHS Demographic and Health Survey DHIS2 District Health Information System 2 DHT District Health Team DMFP District Malaria Focal Person HC Health Center HMIS Health Management Information Strategy iCCM Integrated Community Case Management IDI International Diseases Institute IPTp

ITN

Intermittent Preventive Treatment in Pregnancy Insecticide-Treated bed Net

IVR Interactive Voice Response KII Key Informant Interview

LC1

MAAM

Local Council 1 Mass Action Against Malaria

MAPD

MiP

Malaria Action Program for Districts Malaria in Pregnancy

MIS Malaria Indicator Survey MOH Ministry of Health MOP Malaria Operational Plan NDP National Development Plan NHSDP Government of Uganda National Health Sector Development Plan NMCD National Malaria Control Division OCA Organizational Capacity Assessment OPD Outpatient Department PFP Private-for-Profit PNFP Private-not-for-Profit

PMI

RPD

President’s Malaria Initiative Rapid Diagnostic Test

RHITES

SBCC

Regional Health Integration to Enhance Services Social and Behavior Change Communication

SI Strategic Information SITES Strategic Information Technical Support

SOW

SP

Scope of Work Sulfadoxine-Pyrimethamine

SURMa Strengthening Uganda’s Response to Malaria

UBOS

UCC

Uganda Bureau of Statistics Universal Coverage Campaign

ULA Uganda Learning Activity UMRESP Uganda Malaria Reduction and Elimination Strategy UNCST Uganda National Council for Science and Technology USAID United States Agency for International Development VHT Village Health Team WEP Weekly Epidemiological Reporting iii

EXECUTIVE SUMMARY

The United States Agency for International Development (USAID) through the President's Malaria Initiative (PMI) is designing a new flagship malaria activity called the Malaria Reduction Activity. The purpose of this Baseline Assessment is to establish the status and current trends of malaria prevention and treatment and to document malaria management practices, processes, and capacity for response in communities, health facilities, and districts prior to the implementation of the Activity.

The Baseline Assessment focused on the following questions:

Q1: Across the anticipated Malaria Reduction Activity target regions, what is the current status of health systems, quality of care, and outcome indicators?

Q2: What is the status of key malaria program coverage in targeted districts?

Q3: Based upon available performance indicator data, what is the current perception of malaria program interventions?

Q4: What are the strengths and weaknesses of the NMCD and District Health Teams (DHTs) regarding: (i) national and district-level leadership; (ii) joint planning; (iii) budgeting; and (iv) data reporting capacity, as well as providing oversight and technical assistance of community interventions? What are the strengths and weaknesses of Local Council I’s (LC 1s) regarding oversight and technical assistance of community interventions?

Q5: Who/which are the key public and private stakeholders in malaria prevention and case management at the community and district level? How do they currently collaborate?

Which kind of support could be provided to improve their collaboration and strengthen local malaria prevention and case management?

Q6: How is COVID-19 currently impacting the delivery of malaria services at the national, district, health facility, and community levels? What changes to current interventions, approaches, or systems need to be considered to adapt to the COVID-19 pandemic in the future?

Q7: In the fight against malaria, where can we see major health disparities (e.g. according to gender, age, at-risk groups, refugees)?

The Baseline Assessment was carried out in two phases. This report integrates findings from both Phase I and Phase II. A full description of findings from Phase I can be found in a separate report (Wandera et al., 2020).

ASSESSMENT DESIGN AND METHODS

The Baseline Assessment was focused on high burden areas targeted by the Malaria Reduction Activity: Acholi, Busoga, Karamoja, Lango, and West Nile. The assessment used a mixed-method approach. In Phase I, the assessment team carried out statistical analyses of public health data and a desk review of key documents to validate and explain empirical findings. Phase II involved iv primary data collection. The assessment team collected qualitative data through phone interviews with key informants on national, regional, district, and community levels (n = 49). Quantitative data was collected through an interactive voice response (IVR) survey of public and private sector health facility workers and village health teams (VHTs) (n = 818). Results were triangulated across methods to ensure robust conclusions.

FINDINGS

Situation Analysis

The Phase I situation analysis covered malaria outcomes, prevention, treatment, and surveillance.

Regarding outcomes, the analysis found that based on DHIS2 data and UBOS population projections, regional incidence rates for 2019 exceeded or came close to 1,000 per 1,000 population in Acholi and West Nile regions. Both regions host large numbers of refugees. The high incidence rates therefore point to the importance of including refugees in an effective malaria response. Female incidence rates were significantly higher than males in all regions except West Nile. With regard to prevention and treatment during pregnancy, data from the DHIS2 2019 and MIS 2019 suggest different regions (and districts within those regions) may have different drivers influencing both IPTp3+ coverage and, more broadly, IPTp and ANC access and uptake.

Behavioral data from the MIS shows that ITN coverage and usage are still below national targets in many districts in the proposed Malaria Reduction Activity regions. Regarding malaria treatment, it was found that there were high rates of RDT stock out in 2019 in all target regions, yet many districts – with the exception of Karamoja - nevertheless had low rates of treating without testing, while weekly timeliness of reporting by facilities and quarterly reporting by VHTs remained weak.

Using MIS survey data, the Baseline Assessment described how a variety of reasons were given by survey respondents for treatment-seeking behavior, but that these reasons seemed to differ depending on the extent to which 1) respondents also reported private sector utilization and 2) DHIS2 reported key commodity stockouts. Finally, regarding surveillance, the assessment found that weekly reporting rates and monthly reporting rates for facilities are generally high among Malaria Reduction Activity regions, yet weekly timeliness of reporting by the facilities and quarterly reporting by village health teams (VHTs) remain sub-optimal.

Systems for Malaria Response and Coordination

The assessment looked at the capacity of district and community systems to mount the sort of malaria response envisioned by both the Malaria Reduction Activity and the UMRESP.

Capacity for Planning, Budgeting, and Target Setting

First, the assessment looked at capacity for planning, budgeting, and target setting at district level.

The assessment found that districts valued such platforms as the Mass Action for Malaria Task Force, extended DHT, and quarterly performance review and that, in the context of these platforms, malaria was a prioritized agenda item. However, key informants also suggested that such platforms were highly dependent on partners’ input. Planning meetings that were not dependent on partners (e.g. weekly and monthly DHT meetings) were less likely to meet and the extent to which these meetings prioritized malaria varied from DHT to DHT and depended a lot on the initiative of team members. Key informants also stated that districts had no specific budget for malaria (i.e. budgets tended to be organized around general activity categories and not specific v programs) and that all malaria activities were paid for by partners. Annual planning documents also often did not prioritize malaria and, if there were special malaria plans (e.g. sustainability plans, epidemic response plans), these too were largely supported by partners and were neither costed nor tied to a budget.

Additionally, no district in the sample had district-specific targets for malaria but rather used national program targets or, alternately, confused targets with the malaria Normal Channel.

These findings are problematic in that both the Malaria Reduction Activity and the UMRESP are meant to encourage a more targeted, tailored, and stratified response to malaria prevention, control, and treatment in high burden regions and, yet, most districts do not have the planning capacity, data, or domestic financing to support such a response in the absence of partners.

Sustainability and ownership, therefore, will be an issue. The sense-making workshop held with key NMCD and implementing partner stakeholders proposed several means by which to improve ownership, for example supporting NMCD and Ministry of Health to become primary conveners of quarterly review meetings rather than partners or utilizing central Primary Health Care (PHC) funds for data review activities.

Secondly, some districts were better than others at being able to understand malaria as an integrated issue and, thus, conceive of a role for every member of the DHT. Other districts, by contrast, saw malaria as an issue tasked to the Biostatistician (for surveillance) and the District Malaria Focal Person (DMFP). In the case of the DMFP, too, the district’s approach to malaria sometimes depended on the DMFP’s specialization (i.e. environmental health or clinical care). If the DMFP was an environmental health officer, the approach centered on vector control. If the DMFP was a medical officer, the approach centered on case management. This often left out other key personnel such as the District Health Educator or other DHT members with specialization in maternal and child health. In the absence of an integrated response, malaria often ended up sitting as a specialized silo within the DHT.

Capacity for Engaging the Private Sector

Key informants suggested that quality and supply chain bottlenecks, as well as distance to health facilities and convenience of drug shops, are the main drivers of utilization of the private for-profit sector (PFP). While districts are increasingly bringing PFP facilities into the DHIS2, their capacity to keep track of, engage, and regulate PFPs was hampered by a lack of human and financial resources. Many DHTs were also frustrated by inadequate regulatory mechanisms at their disposal. Additionally, key informants reiterated that PFPs were not a unified category and that engaging PFP facilities was slightly easier than attempting to understand and manage the role of PFP drug shops.

Capacity for Responding in the Context of Refugee Populations

Key informants reinforced the Baseline Assessment hypothesis from Phase I: that refugee and cross-border populations are not included in district population estimates but, nevertheless, receive treatment for malaria in public and PNFP facilities, thus driving up incidence rates. Key informants believed that this had major consequences for supply chain management in refugee-hosting districts as stock orders are based on population estimates and, thus, may not take into account true commodity utilization by facilities treating refugees. Additionally, it is sometimes difficult for DHTs and their health partners to coordinate and align with humanitarian partners serving refugee communities and this means that these communities are often not served with vi the same key malaria program components – notably VHT inputs and iCCM packages – as their non-refugee neighbors.

Capacity for Community Response and Coordination

Key informants pointed to the fact that the key stakeholder group for community-level malaria response was VHTs, a cadre with tenuous links to local government structures that was primarily overseen by health facility staff. LC 1 and LC II chairmen - who did play a role in mobilizing communities for health-related activities, including key malaria interventions - had a limited role in coordinating VHTs and were often left out of data dissemination. They also had limited capacity to plan for malaria in the absence of partner support. Additionally, VHTs ability to respond - as well as report data - was limited by the well-documented motivation gap that exists, since VHTs are volunteers and remain unremunerated and dependent upon partner compensation and facilitation. This greatly impacted iCCM in the community, a program that was highly valued but also recognized as largely unsustainable in its current form.

COVID-19 and Its Impact on Malaria Service Delivery and Utilization

COVID-19 had a number of impacts on malaria prevention, treatment, and surveillance. IVR survey results and DHIS2 data suggest that community outreach work was strongly impacted across the health system. More than 75 percent of VHTs and health facilities reported that they reduced home visits and community outreaches. Additionally, since the beginning of the pandemic, significantly fewer pregnant women sought ANC care and IPTp coverage seems to have dropped accordingly.

Regarding malaria treatment, the data indicate that while there was no uniform reduction in district-level malaria incidence rates, either fear of COVID-19 or lockdown-related mobility restrictions seemed to be driving away malaria patients from public and PNFP health facilities. A vast majority of VHTs reported that they engaged with fewer fever patients since the beginning of the pandemic, while only a minority of health facilities reported the same. At the same time, respondents reported that ACT and RDT stock out increased in 40 percent and 43 percent of health facilities, respectively. Additionally, the treatment of severe malaria cases reportedly became more difficult for a majority of health facilities. A few key informants suggest that this may be in part due to the reduced availability of blood donations for the treatment of anemia which may result from severe malaria.

With regard to the impacts of the pandemic on malaria surveillance, the majority of VHTs reported that they received less support from supervising health centers and reporting became more difficult for them. In contrast, health workers at facilities reported that the analysis and reporting of malaria data was not significantly affected by the pandemic.

CONCLUSIONS

Secondary data analysis from Phase I found that:

• The context-driven approach prioritized by both the UMRESP and the Malaria Reduction Activity was justified by the evidence. Inter- and intra-regional diversity warranted improved tailoring, targeting, and stratification of interventions on a district-by-district basis.

vii

• The PFP sector clearly plays an important role in service delivery for malaria but the magnitude is still not well documented and the dynamics that drive its use may differ from district to district.

• The drivers that inform utilization of services or programs, including treatment seeking, uptake of ANC/IPTp, and use of ITNs, may also differ from district to district, which justifies a shift toward a demand-driven approach designed around the underlying factors that influence and inform patient and caregiver choices.

• The capacity of VHTs and other community systems varies widely and the dynamics of community health systems and the performance of iCCM at the district level need to be understood and targeted better.

• In the target regions, an understanding of, and response to, both the malaria burden and program coverage bottlenecks depends on robust data systems, which urgently require strengthening.

Primary data collection in Phase II validated many of these findings and led to the following conclusions:

• Building district capacity for malaria response means focusing on whole-team building:

Districts and their partners need to understand and invest in malaria as the responsibility of every member of the DHT, not just those with specialization in a technical malaria response.

• Investing in planning tools, structures, and platforms should include investment in building capacity for domestic finance mobilization: Partner-supported planning platforms (MAAM Task Force, quarterly performance reviews, District Malaria Sustainability Plans, etc.) are highly valued, but unsustainable in the absence of partner support, and so sustainability must be a key focus of the Malaria Reduction Activity.

• A targeted response requires improving target setting: No district had district-specific incremental and five-year targets tailored to the local context, nor did any of the LC 1s.

Data utilization across all levels - from village to parish to sub-county to district - needs to include a proactive planning focus, as well as investment in dissemination, understanding, and utilization of data at the community level.

• There is no formal mechanism to link village local government to VHTs: There are no formal channels of data dissemination, nor platforms to support LC I data utilization, and LC Is have little-to-no training in planning, target setting, costing, or budgeting (and, indeed, LC I government lacks a formal budget). For the UMRESP and the PMI Malaria Reduction Activity, bridging the political and technical divide at the community level must be prioritized.

• iCCM requires better ownership: iCCM is highly valued but also entirely dependent on partners and lacks sustainability. iCCM is a key strategy to achieve the goals set forth in the new UMRESP so addressing local ownership and domestic financing is important going forward.

viii

• PFPs urgently need integration into data systems, including creative integration of drug shops: While PFP facilities are increasingly being registered and incorporated into the DHIS2, most districts still struggle with engaging PFPs in their communities. Improving the information system’s ability to capture how many malaria cases are being seen by PFP providers and how they are being managed is critical.

• There is an important demand-side component to the PFP sector which has not yet been addressed: PFP interventions are predominantly supply-side (e.g. trainings, guideline briefings, data reporting). Understanding and addressing consumer demand for PFP services, including those provided by drug shops, is an important area for focus.

• Mainstreaming PFPs and their clients into all program interventions might improve engagement: Engagement of the PFP sector is often conceived programmatically as a stand-alone strategic objective. Understanding PFPs, including drug shops, as part of an overall system of malaria service delivery and utilization and, therefore, targeting PFPs with all activities and interventions intended for the public sector may be a more useful way of engaging them.

• Malaria can and should be part of an integrated, multilevel, multisectoral approach, crisis or no crisis: While COVID-19 and its mitigation measures seriously disrupted malaria service delivery and utilization in the first few months of the pandemic, eventually district and community stakeholders adapted. The lesson that COVID-19 can teach malaria stakeholders is that a multisectoral approach involving the integration of health issues and the involvement of lower level political structures and communities is key.

The assessment concludes that district malaria response takes place in a complex system and activities targeting the prevention, treatment, and surveillance of malaria intersect across different health thematic areas (nutrition, child health, maternal health), levels of health systems (DHT, HC, VHT), and levels of governance (district, sub-county, parish, village). Capacity assessment -and capacity building - should be whole-district systems focused, and target each district’s unique network of relationships. This perspective best aligns to the goals and pillars set forth in the new UMRESP with its focus on targeting and stratification, decentralization and ownership, multisectoralism and broad engagement with stakeholders beyond the public sector, and demand-side considerations.

RECOMMENDATIONS

District Capacity for Planning, Budgeting and Target Setting

• Build capacity and improve advocacy to identify other sources of domestic financing (e.g.

leverage MAAM or, alternately, PHC funds by ensuring malaria is understood as a multisectoral issue integrated across most health areas).

• Increase sustainability and ownership of existing platforms by local government by supporting central government to improve its mandate over them.

• Improve data utilization especially with respect to, first, closing the gap between when data is reported and when it is utilized and, second, creating feedback mechanisms.

ix

• Help districts set appropriate district-specific, proactive, annual and five-year targets.

• Reach the entire DHT with capacity-building initiatives to promote a whole-DHT, integrated response to malaria.

• Work with districts and central government to include estimates of refugees’ impact on local health systems in planning processes.

District Capacity for Community Response and Coordination

• Ensure VHT coverage and iCCM are extended to refugee populations and harmonized to existing DHT approaches by working more effectively with humanitarian partners.

• Address the VHT motivation gap in the absence of salaries via creative ways (support with supplies; access to loans and savings program; mobilizing financing at village level).

• Support linkages of VHTs not just to facilities but to LC 1 government structures (e.g.

health committees), including building accountability with respect to data dissemination and local ownership; also support linkages of LC chairpersons to VHT quarterly review meetings.

• Invest in supply chain systems strengthening for VHTs, VHT coordinators at LC II level, and health facility supervisors to better assess, quantify, and redistribute supplies.

• Engage key stakeholders such as cultural and religious leaders since they play a key role in the mobilization of both community members and domestic financing which, in turn, may help address sustainability issues.

Capacity for Engaging Private-For-Profits

• Strengthen district capacity to bring PFPs into the DHIS2 and to respond to data (e.g.

potentially bring PFPs or their representatives into DHT quarterly review meetings).

• Creatively engage drug shops.

• Remember demand-side approaches that target PFP customers.

• Mainstream PFPs into all activities targeting health facilities and health workers.

Strengthening Malaria Response as a System

• The COVID-19 pandemic has shown that malaria can and should be part of an integrated, multilevel, multisectoral approach.

• Capacity assessment – and capacity building – should be whole-district systems focused, targeting each district’s unique network of relationships that feed into an effective malaria response.

INTRODUCTION

The United States Agency for International Development (USAID), through the President's Malaria Initiative (PMI), is designing a new, flagship malaria activity called the PMI Malaria Reduction Activity. The purpose of this anticipated USD $38 million activity is to support the National Malaria Control Division’s (NMCD) implementation of the Uganda Malaria Reduction and Elimination Strategy (UMRESP) in order to reach the goal of a “malaria-free Uganda.” The UMRESP marks a strategic shift for the NMCD toward a response that involves a) improved utilization of data through targeting and stratification; b) building capacity for multisectoral action;

c) prioritization of decentralized, community-level approaches and ownership; d) redirection of focus toward demand-side interventions; and e) better engagement of the private-for-profit (PFP) sector.

The PMI Malaria Reduction Activity’s primary objectives, therefore, are in line with UMRESP’s goal to “strengthen the capacity for malaria prevention and treatment, as well as ownership of health, at the community and household levels, while maintaining gains made at national, district and the health facility level in focus districts” (Ministry of Health, 2020).

PMI Uganda is seeking the assistance of the Uganda Learning Activity (ULA) to carry out a Baseline Assessment to establish the status and current trends of national malaria prevention and treatment in Uganda, as well as current malaria management practices and processes.

The Baseline Assessment was carried out in two phases:

Phase I: Secondary Data Analysis of malaria status and program performance indicators.

Phase II: Assessment of Processes, Relationships, and Policies within the Malaria Sector.

Phase I of the Baseline Assessment was completed on July 24, 2020. Data collection for Phase II was carried out in September 2020. A full description of findings for Phase I can be found in the Phase I Final Report (Wandera et al., 2020). This integrated report briefly summarizes Phase I while also addressing at length a set of assessment questions that were answered primarily in Phase II. For the combined Phase I and II Scope of Work (SOW), see Annex I.

ASSESSMENT PURPOSE

The purpose of the Baseline Assessment was to establish the status and current trends of malaria prevention and treatment, as well as malaria management practices and processes in Ugandan communities, health facilities, districts, and national-level spaces, prior to the implementation of the PMI Malaria Reduction Activity.

ASSESSMENT QUESTIONS

The Baseline Assessment addressed seven questions:

Q1: Across the anticipated Malaria Reduction Activity target regions, what is the current status of health systems, quality of care, and outcome indicators?

Q2: What is the status of key malaria program coverage in targeted districts?

Q3: Based upon available performance indicator data, what is the current perception of malaria program interventions?

Q4: What are the strengths and weaknesses of the NMCD, District Health Teams (DHTs) regarding (i) national and district-level leadership; (ii) joint planning; (iii) budgeting; and (iv) data reporting capacity, as well as providing oversight and technical assistance of community interventions? What are the strengths and weaknesses of Local Council I’s (LC 1s) regarding oversight and technical assistance of community interventions?

Q5: Who/which are the key public and private stakeholders in malaria prevention and case management at the community and district level? How do they currently collaborate?

Which kind of support could be provided to improve their collaboration and strengthen local malaria prevention and case management?

Q6: How is COVID-19 currently impacting the delivery of malaria services at the national, district, health facility, and community levels? What changes to current interventions, approaches, or systems need to be considered to adapt to the COVID-19 pandemic in the future?

Q7: In the fight against malaria, where can we see major health disparities (e.g. according to gender, age, at-risk groups, refugees)?

Phase I addressed Q1, Q2, Q3, and Q7. The Findings- Situation Analysis section of this report summarizes the outcomes of Phase I research.

Phase II addressed Q4, Q5, and Q6. The Findings- Systems for Malaria Response and Coordination and Findings- COVID-19 and Its Impact on Malaria Service Delivery and Utilization sections of this report address these questions in detail.

UTILIZATION OF ASSESSMENT FINDINGS

The Baseline Assessment will be used, first, to assess the capacity of public institutions involved in planning and coordinating Uganda’s malaria response. It will provide an understanding of the community- and district-level partnerships and the types of collaboration that are critical to malaria prevention and treatment. It will evaluate the impact of the COVID-19 pandemic specifically on malaria service provision. Finally, it will analyze the current status of key programmatic indicators. This information, taken together, provides critical input to the planning and design process of the new malaria follow-on activity.

The assessment team believes that this information is essential for the development of effective and efficient interventions in Uganda’s health system, to improve health system governance, and to create an enabling environment for malaria reduction and elimination.

PROJECT BACKGROUND

Uganda has made significant progress on malaria prevention and treatment, with estimated incidence dropping from 272 cases per 1,000 population in 2017 to 191 cases per 1,000 population in 2018 (MOH, 2018a). Meanwhile, malaria prevalence among children <5 years has decreased over the last decade from 42 percent in 2009, to 19 percent in 2014, to 9 percent in

2018 (UBOS, 2019).

Nevertheless, the country faces several health challenges related to malaria and a significant proportion of Uganda’s population continues to live in districts classified as having ‘high endemicity.’ Prevalence also varies widely between regions. In the highest-burden regions, prevalence rates range from 34 percent in Karamoja, to 22 percent in West Nile, 21 percent in Busoga, 13 percent in Lango, and 11 percent in Acholi. In refugee communities the estimated prevalence is 13 percent (MIS, 2019).

To respond to regional health disparities, UMRESP 2021-2025 has adopted a stratified approach based on prevalence rates (MOH, 2020). According to this strategy, interventions shall be tailored to each region’s epidemiological profile and deployed to:

• Accelerate the reduction of the malaria burden in high-endemicity areas, and;

• Maintain a low burden -or aim for the elimination of malaria altogether- in areas of low endemicity.

UMRESP also aims to reduce malaria infection and morbidity rates by 50 percent, and malaria-related mortality rates by 75 percent, by 2025 (MOH, 2020).

THE PMI UGANDA MALARIA REDUCTION ACTIVITY

USAID/Uganda, through PMI, is designing a new flagship malaria activity, the Malaria Reduction Activity. The purpose of this anticipated USD $38 million activity is to support the NMCD’s implementation of the UMRESP in order to reach the goal of a “malaria-free Uganda.”

The PMI Malaria Reduction Activity will follow the current five-year, PMI-funded Malaria Action Program for Districts (MAPD) activity (2016-2021). MAPD targeted 52 districts in five regions in Uganda with a range of interventions, including those related to case management and treatment, bed net distribution and use, malaria in pregnancy, social and behavior change communication (SBCC), data collection and utilization, and district and NMCD capacity building. In the past, PMI funding has also supported similar packages of interventions through the Regional Health Integration to Enhance Services (RHITES) projects in another five regions of Uganda.

The PMI Malaria Reduction Activity is presently in the project design stage. Aligned with the UMRESP, the PMI Malaria Reduction Activity redirects PMI’s focus toward communities in districts in the highest-burden regions, including refugee-hosting districts. A number of key interventions are anticipated under this award. These interventions include: 1) strengthening case management and treatment, especially for children aged <5 years, school-aged children, and pregnant women; 2) improving bed net distribution and utilization; 3) building village health team (VHT) capacity; 4) strengthening surveillance and response systems, especially at the community level; 5) engaging private sector facilities and drug shops; and 6) strengthening data systems and referral linkages between communities, health facilities, and DHTs.

Figure 1 provides an overview of potential Malaria Reduction Activity target districts (yellow) compared to former MAPD districts (blue), with overlapping areas indicated by stripes.

Figure 1 - Anticipated Malaria Reduction Activity Target Districts

ASSESSMENT METHODOLOGY & LIMITATIONS

The Baseline Assessment used a mixed-method approach that included statistical analyses of public health data, key informant interviews (KIIs) with stakeholders within the Ugandan health system, and an interactive voice response (IVR) survey of health workers and VHTs. The assessment team triangulated results across methods to ensure robust conclusions. A sense-making workshop was also held to enable key stakeholders to participate in the analysis (Annex X).

The Baseline Assessment was carried out in two phases:

• Phase I focused on statistical analyses of public health data. This was combined with a desk review of key documents to validate, support, explain, and contextualize empirical findings. The analysis focused on case management, surveillance, treatment seeking and its drivers, intermittent preventive treatment in pregnancy (IPTp) 3+ coverage and utilization, as well as bed net coverage and utilization.

• Phase II involved primary data collection. The assessment team collected qualitative data through KIIs with national-, regional-, district-, and community-level stakeholders.

Quantitative data was collected through an IVR survey of public and private sector facility health workers and VHTs. The KII transcripts provided content for structured district-level organizational capacity assessments (OCAs), thematic analyses of the capacity for oversight and technical assistance of community interventions at community, district, and national level, and an analysis of stakeholders at the district level. The IVR survey provided the basis for an analysis of the impacts of the COVID-19 pandemic on malaria prevention and treatment. To comply with government-imposed COVID-19 travel and social distancing restrictions, the team conducted the survey and all interviews by phone or using online conferencing tools.

Detailed descriptions of the methods employed in both phases of the assessment are available in Annexes II-V. Data collection tools can be found in Annex VI. The following sections provide an overview.

DATA COLLECTION

KIIs

KIIs provided the content for OCAs and stakeholder analyses, as well as additional information on the impact of COVID-19 at the community, district, regional, and national levels. Semi-structured interview guides were designed to prompt a guided discussion. In the case of OCA-focused interviews, the guides ensured that organizational capacity topics of interest and questions about stakeholders and their relationships were touched upon during the interviews.

KIIs focused on ten study districts in the five target regions of the anticipated PMI Malaria Reduction Activity. District selection was based on key malaria indicators and ensured an adequate regional representation. The key indicators are: (i) antenatal care coverage, 4+ visits (ANC4); (ii) artemisinin-based combination therapy (ACT) stock out; (iii) VHT reporting on time, and (iv) weekly epidemiological program (WEP) reporting rates. The details of the district selection process are available in Annex III. Table 1 shows the selected districts.

Table 1 - Study Districts

District Region Regional IP ACT Stock Out <5%

VHT

Reporting On Time >43.0%

WEP

Reporting

Rate >74.7%

ANC4

Coverage >47.8%

Apac Lango RHITES-Lango SURMa Yes Yes Yes Yes

Gulu Acholi RHITES-Acholi SURMa Yes Yes No Yes

Lamwo Acholi RHITES-Acholi SURMa Yes No Yes Yes

Koboko West Nile MAPD No Yes Yes Yes

Abim Karamoja RHITES-E SURMa No Yes No Yes

Lira Lango RHITES-Lango SURMA No No Yes Yes

Agago Acholi RHITES-Acholi SURMa No No No Yes

Alebtong Lango RHITES-Lango SURMa Yes Yes Yes No

Madi-Okollo West Nile MAPD No Yes Yes No

Kamuli Busoga RHITES-EC No No No No

Notes: For details regarding the indicators and applied thresholds, please refer to Annex III.

For each district, the assessment team selected key informants from four broad categories of stakeholders: 1) members of DHTs, 2) members of the LC1 Committee, 3) implementing partners at regional/national level and community based organizations (CBOs), and 4) PFP malaria service providers. In addition, the assessment team interviewed NMCD staff.

Due to travel restrictions, the assessment team conducted all KIIs via phone or online communication platforms. A total of 49 KIIs were conducted. Table 7 provides a breakdown

Table 2 - Selection of Key Informants

Category Target Actual

a. NMCD 3 3

b. Members of DHTs 3 persons x 10 districts 23

c. Members of LC I Committee 1 person x 10 districts 8

d. Implementing Partners and CBOs 7 12 (9 IPs, 3 CBOs)

e. PFP Representatives 1 person x 5 regions 3 Total 49

IVR survey

IVR survey questions focused on the impact of the COVID-19 pandemic on malaria prevention, treatment, and surveillance in the target regions, especially on treatment-seeking and fever case management in children <5 years and on malaria in pregnancy. Health care professionals at public and private health facilities were asked about the potential impacts of COVID-19 on malaria-related services provided by their health facilities. VHTs were asked about impacts to the malaria-related services that they provide.

The questions and the answer options were pre-recorded. Invitees were called on their mobile phones and the pre-recorded messages were played to them. Respondents were able select answer options by using the number pad on their phone.

The assessment team collected the contact data of 7,407 public and private sector health care professionals across all 48 PMI Malaria Reduction Activity target districts. All contacts were invited to participate in the survey. A total of 818 health workers provided full answers.

Respondents who used a different number when calling back could not be matched, and so we do not know which region they are from. Table 3 shows the distribution of respondents according to region and type of stakeholder.

Table 3 - IVR Survey Sample

Region Government Health Centers

Private For-Profit

Private Non-Profit

Village Health Team Total

Acholi 42 10 7 3 62 Busoga 44 12 17 89 162 Karamoja 98 9 24 14 145 Lango 53 11 12 153 229 West Nile 49 22 20 70 161 NA (call backs) 15 2 1 41 59 Total 301 66 81 370 818

DATA ANALYSIS

Analysis of public health data

The analysis focused on the high burden regions targeted by the Malaria Reduction Activity and extracted - as much as possible - data on children <5 years, pregnant women, and refugee-hosting districts. Annex II lists the data sets that were used. Data analysis involved the following:

• Correlation Analysis: The team computed correlations of key outcome indicators and key input indicators to determine the degree of association and its implications for programming.

• Trend Analysis: The team plotted key program level indicators over time to establish direction and, therefore, implications for programming.

• Tabulation: The team tabulated secondary malaria data to permit quick access and utilization by PMI and the Strategic Information (SI) teams when needed.

• Heat Maps: Heat maps were created to identify service-level disparities among districts.

• Choropleth Maps: For key indicators, the team generated maps to see the shift over time (2015-2020) in major areas of service delivery.

This report provides a summary of the findings. A complete documentation can be found in the Phase I Final Report (Wandera et al., 2020) and supplementary annexes.

KIIs

KIIs were transcribed and thematically analyzed. A set of codes was developed a) based on the assessment questions and OCA categories, and b) derived from emerging themes developed collectively during data analysis via grounded theory. Only themes that were found to be shared by more than two key informants were treated as “findings,” except in instances where the theme was highly context-specific.

Information identified in the DHT transcripts that is relevant to OCA was analyzed in a separate matrix tool (see Annex VII). The OCA tool provided themes to support the analysis of the transcripts. A tool such as this is normally intended to be embedded within participatory activities tied to action planning in order to ensure organizational ownership of both the findings and action plans. Due to COVID-19 mobility restrictions, however, DHTs could not be convened. Thus, the assessment team could not utilize the tool in the way originally intended. Instead of using the tool to rank or assess capacity, the team focused on OCA as a means by which to code transcripts thematically.

IVR Survey

The survey data was analyzed with the help of statistical software (R). The analysis included descriptive statistics and a systematic comparison of different stakeholder groups. Correlational analysis was employed to understand the role of regional and institutional contexts in mitigating or exacerbating the impact of the COVID-19 pandemic on malaria prevention and treatment.

LIMITATIONS

• The team did not have the time or the means to investigate the data generation process of existing data sets. Secondary data utilization is always constrained by the purpose for which primary data was originally collected and intended to be used. Additionally, the timing of data collection and frequency of data points differ from source to source. This means that reliable comparisons and correlation analyses were not always possible. Finally, indicators that are collected by multiple agencies are likely to differ in the data they provide on the same indicators. This creates the challenge of determining which is the most authentic source among the sources available.

• Given the limited time and resources available for this assessment, a representative sampling of key informants was not possible. While a range of key informants was targeted, interviews were necessarily opportunistic based on which key informants agreed to participate. KIIs were affected by participation/non-response bias as those who agree to participate may have represented stakeholders who are already more proactive, engaged, or possess higher capacity than those who do not agree to participate.

• The IVR survey relied on a sampling frame with an uneven representation of types of health workers, regions, and districts. It was built by contacting implementing partners and district officials. No random sampling procedure was employed. Instead, all contacts in the sampling frame were invited to the survey. Finally, with a response rate of 14.5 percent, self-selection played an important role in the realization of the final sample. This means that simple distributions may be subject to various biases. The assessment team employed tests of association to ensure that only significant correlations would be interpreted substantially.

• Additionally, there may be a degree of observer bias as some key informants -such as NMCD staff- are simultaneously engaged in data analysis and will be sought out for input into the findings.

The purpose of this assessment, however, is to gain an understanding of stakeholder perspectives on aspects of the malaria response in Uganda. The methods and activities have therefore been selected to facilitate (virtual) on-the-spot joint debriefing and learning and are adequate for this purpose.

FINDINGS

SITUATION ANALYSIS: SUMMARY OF FINDINGS FROM PHASE I

Phase I of the Baseline Assessment consisted of statistical analyses of public health data on malaria program indicators, coupled with a desk review of key documents in order to validate, support, explain, and contextualize empirical findings. Phase I concentrated on the following subset of questions:

Q1: What is the current status of health systems, quality of care, and outcome indicators?

Q2: What is the current status of key malaria program coverage in targeted districts?

Q3: What is the current perception of malaria program interventions based upon available performance indicator data?

Q7: What are the areas of major health disparities in the fight against malaria (especially according to children <5 years, pregnant women, and refugees)?

The assessment encountered numerous issues with the existing secondary data. A major finding of Phase I was the extent of data quality challenges and incompatibilities between data sets. The assessment team carried out correlation and trend analysis, heat mapping, and choropleth maps.

The Phase I findings in full can be found in the Phase I Final Report (Wandera et al., 2020).

Table 4 provides an overview of findings of patterns in regional performance, as they emerged from the analysis of public health data. The signs indicate a region’s performance relative to other regions, in particular those targeted by the Malaria Reduction Activity. Individual dimensions of the malaria response (outcomes, prevention, etc.) are rated based on the indicators listed in the table. Only substantial differences in regional averages were taken into account. Plus (“+”) and minus (“-“) signs indicate above- or below-average performance or indicator values, while a “o” indicates an average performance or value. Double plus (“++”) or double minus (“--“) signs were introduced to further differentiate the relative ratings of individual dimensions of the malaria response, with “++” indicating a performance strongly above average and “--” indicating a performance strongly below average.

Table 4 - Overview of Patterns in Regional Performance

Indicator Acholi Busoga Karamoja Lango West Nile

Reliance on Public Health System o - + - o Outcomes* - - - o -- Malaria Prevalence + o o o + Children <5 All-Cause Mortality Rate o + + o +

Prevention* - - -- + + IPTp3+ Coverage - o - o + SP stock out o + + - o Treatment* o -- -- - --

ACT stock out o + o + + RDT stock out o + + o + Treatment without testing o o + o o Surveillance* o -- + + ++ HC Reporting o - o + + VHT Reporting o - + o +

Notes: *Ratings express the average degree of performance. High performance may imply low indicator ratings and vice-versa.

The patterns are discussed in detail in the following paragraphs. They highlight a number of key issues for malaria program development going forward.

Outcomes

Findings from Phase I concluded that it will be critical for the Malaria Reduction Activity to:

• Conduct a deep dive into drivers of female vulnerability on a district-by-district basis;

• Establish a better understanding of the role that treatment seeking by refugee populations and clients of the PFP sector play with respect to district and regional malaria status outcomes, as well as improved engagement with humanitarian and other partners in regions where these partners might be providing parallel malaria services to communities.

Figure 2 – Annual Malaria Incidence Rates across Subregions (2019)

Notes: Based on DHIS2 data and UBOS population projections. Information on refugee populations taken from UNHCR (UNHCR, 2020).

An important finding from Phase I was that, according to District Health Information System 2 (DHIS2) data and UBOS population projections for 2019, annual malaria incidence rates per 1,000 population exceeded or were close to 1,000 cases for Acholi and West Nile regions. In the Phase I report, the team explained these high incidence rates partly by the presence of either large refugee populations in certain districts and/or cross-border facility use by patients from neighboring countries.