Advocates Coalition for Development and Environment

67TH STATE OF THE NATION PLATFORM

“Confronting the Challenge of Child Mortality in : Scale-up and Consolidation of Gains from the CODES project”

ACODE Policy Dialogue Report Series, No. 32 Adelaine Aryaija Karemani, Barbara Ntambirweki & Moses Mukundane

Advocates Coalition for Development and Environment

67TH STATE OF THE NATION PLATFORM

“Confronting the Challenge of Child Mortality in Uganda: Scale-up and Consolidation of Gains from the CODES project”

ACODE Policy Dialogue Report Series, No. 32 Adelaine Aryaija Karemani, Barbara Ntambirweki & Moses Mukundane 67TH STATE OF THE NATION PLATFORM

ABSTRACT and lessons learnt from five years of CODES implementation.The dialogue Over the years, Uganda has made brought together a cross-section of key stakeholders in child survival in Uganda significant strides in reducing child including; Women MPs, District Health mortality. Child mortality in Uganda Officers, Chief Administrative officers of declined from 156 per 1000 live births in the 13 CODES implementation districts, 1990s, to 137 per 1000 live in 2005, 90 National Medical Stores, Health Monitoring per 1000 live births in 2011 to the current Unit, UNICEF,ACODE, CFI, MoH and 69 per 1000 live births1. This was partly Makerere University. The objectives of the due to result of increased knowledge and 67th STON were as follows: service in the health sector. The five-year Community and District Empowerment 1. To share with participants the for Scale-up (CODES) project, developed achievements and lessons learnt from 5 jointly by Uganda’s Ministry of Health, years of CODES implementation; UNICEF and Karolinska Institutet in 2. To provide a national-level platform for partnership with Advocates Coalition for participants to discuss the achievements Development and Environment (ACODE), and lessons learnt from 5 years of Child Fund International (CFI), Liverpool CODES implementation; School of Tropical Medicine (LSTM) and Makerere University School of Public 3. To provide a national-level platform Health (MUSPH) put in place a myriad for participants to discuss policy of interventions to reduce child deaths implications for CODES exit and associated with diarrhea, pneumonia and sustainability strategy. malaria. It is against this background that Advocates Coalition for Development B. Background and Environment (ACODE) together with CODES partners held the 67th State of the Child mortality remains a global health Nation Platform (STON) for a discussion concern and this explains its particular locus on “Confronting the Challenge of Child in the global health/child survival agenda. Mortality in Uganda: Scale-up and The recent Millennium Development Goals Consolidation of Gains from the CODES (MDGs) and the present Sustainable Development Goals (SDGs) put emphasis project.” A number of recommendations on reducing child mortality2. Consistent emerged key among which was the call with the global child health agenda, the to adopt the community dialogue model Government of Uganda (GOU) through to increase demand for health services the Ministry of Health (MOH), and in particularly for children and the need to partnership with Development Partners strengthen the capacity of Health Unit and Civil Society Organizations (CSOs) Management Committees (HUMCs) has undertaken several child survival through training and financial facilitation. interventions under the national framework of National Minimum Health Care Package A. Introduction (UNMHCP), Health Sector Strategic Plans (HSSP) I, II and III (2000/1-2004/5; 2005/06- This report focuses on the achievements 2 UNDP: Http://www.undp.org/content/undp/en/home/ sdgoverview/mdg_goals.html. Accessed on October 10, 1 (MFPED, 2015; MOH, 2015a; UBOS and ICF, 2011). 2016

2 Confronting the Challenge of Child Mortality in Uganda 67TH STATE OF THE NATION PLATFORM

2009/10; 2010/11-2014/15), Health and interventions that lead to an array Sector Strategic Investment Plan (HSSIP) of improvements in health outcomes, (2010/11-2014/15), and the Health Sector especially concerning the control of Development Plans (HSDP) I and II (2010/11 diarrhea, pneumonia, and malaria. – 2014/15; 2015/16 - 2019/20), inter alia (MOH, 2015a3; 2015b4). The totality of the The CODES project had 3 pillars; interventions resulted into considerable strides in reducing child mortality in the 1. Ensuring that the best interventions were country, especially the type of mortality used to reduce child death and illness in resulting from the top killer childhood each of the participating districts. diseases of diarrhea, pneumonia, and malaria, as well as diseases resulting from 2. Helping District Health Managers lack of or inadequate administering of timely to use simple tools to improve their vaccines. Consequently, contemporary performance on health service delivery. reports indicate a trend of drastic decline in child mortality rate in Uganda over the last 3. Empowering communities to demand two and a half decades, an era that overlaps and receive better quality and gain more with the United Nations MDGs epoch. access to all health services. The Uganda MDG report 2015 indicates that Uganda “missed narrowly” to meet The three pillars epitomized both MDG 4 target of reducing child mortality to C. the “Supply-Side” and “Demand-Side” the rate of 56 per 1000 live births by 2015 components of the CODES project. The (MFPED, 2015)5. “Supply Side Component” interventions were intended to improve district level The CODES project put in place a myriad health systems and health facility based- of interventions to reduce child deaths continuous quality improvement (CQI) in associated with diarrhea, pneumonia and health service delivery while the “Demand- malaria. The interventions involved activities Side Component” was intended to mobilize that were supported by the Government and galvanize citizens to demand for of Uganda, CSOs, and other non-state effective and quality health service delivery. actors all of which focused on reducing ACODE has shouldered the mandate of child mortality. Designed as a learning implementing the third pillar/Demand-Side project, CODES set out to demonstrate component of the CODES project. how strengthening of district health systems management and community empowerment could be scaled-up to D. Presentations: attain intended objectives. CODES was anticipated to help the government to Dr. Flavia Mpanga Kaggwa, boost its capacity to implement policies UNICEF, Principal Investigator, 3 Ministry of Health (2015a). Annual Health Sector CODES project: Performance Report for Financial Year 2014/15.MOH. Kampala, Uganda. 4 Ministry of Health (2015b). Health Sector Development Dr. Mpanga presented a brief overview Plan (2015/16- 2019/20) .MOH. Kampala, Uganda. of the 5-year CODES project, the policy 5 Ministry of Finance, Planning and Economic Development (2015). Millennium Development Goals Report for prospects and exit/sustainability strategy. Uganda 2015. Results, Reflections and the Way Forward. Dr. Mpanga extended appreciation to the MFPED with Support from United Nations Development Programme (UNDP). Kampala, Uganda. CODES project implementers (ACODE +

Confronting the Challenge of Child Mortality in Uganda 3 67TH STATE OF THE NATION PLATFORM

CFI) for their contribution towards reducing as to supervise and to track performance child mortality in Uganda, which further led of health units. Health officials were further to achieving the MDG 4. Uganda’s child empowered to put in place improvement mortality rate was noted to stand at 55/1000 plans and continuously implement in a live births which was acknowledged to SMART way for the next cycle. This has be a huge achievement. However, this resulted in strengthening all six pillars of observation was put in context and noted the health system; leading to increased to be inadequate because so many children coverage of key health interventions. Dr. continued to die needlessly. Mpanga emphasized that all this was done with very little resources—district health Dr. Mpanga shared the CODES project teams used existing resources and where hypothesis which consisted of several resources were not available or sufficient, inputs with an overall aim of reducing districts mobilized resources from other child mortality in the 13 implementation donors because the districts were now districts. The CODES hypothesis adapted able—while using data, to tell partners an approach that harnesses data that the where their need was and therefore districts already had from the HMIS/DHIS2. influenced where resources could be They also obtained data from dialoguing allocated. With this, the hypothesis was with communities to understand their that; all these inputs together would reduce experiences and reasons for not seeking child morbidity and mortality and help the health care services. With this data, districts nation get to Vision 2040 and eventually health teams were approached to perform also look into the sustainable development evidence based planning that targeted goals.

Districts that participated in the CODES project included; Wave 0 (Proof of concept) namely: Buikwe, Bukomansimbi, Masaka, Mukono, Wakiso; Wave 1 (Intervention districts) of Apac, Arua, Bugiri, Buhweju, Buvuma, Luuka, Masindi and Maracha and Control districts of Alebtong, Kole, Kiryandongo, Kamuli, , Kasese, Mitoma and Sheema. The project started with a proof of concept phase in 5 districts. There were 8 intervention and 8 control districts. The project design deliberately intervened in 8 districts. In control districts, baseline and Figure 1 Dr. Flavia Mpanga, UNICEF making end-line data collection ascertain status in her remarks at the 67th STON session terms of child indices was the main activity. sub-counties that were performing sub- Looking at key processes that were optimally as well as made sure that the little employed by the CODES project; Child resources available were efficiently utilized. Fund International (CFI) collected data using Lot Quality Assurance Sampling CODES project empowered district health (LQAS) and District Health Management officials to be able to take decisions as well Information Software System version 2

4 Confronting the Challenge of Child Mortality in Uganda 67TH STATE OF THE NATION PLATFORM

– Bottleneck Analysis (DHIS2-BNA) and health workers were put to task to develop accordingly brought evidence to the work solutions. This process enabled the analysis plan processes. ACODE was in-charge of of system gaps by subjecting them to a the ‘demand side’ component of CODES root cause analysis to identify solutions and through the use of community dialogues. strategies for districts. Dr Mpanga explained that a community dialogue all about talking and holding a Some of the achievements on the supply discussion-- and that this had power in itself. side were; CODES showed that talking to household members reduced child mortality. Issues • Evidence based work-plans in 13 like talking to households about family care districts increased funding for child practices, sleeping under mosquito nets, health from 4% to 6% in 2 years immunizing children and washing hands • Improved reporting through DHIS from with soap and clean water were small things 23.5% in 2013 to over 80% in 2016 that many people take for granted but that across the 487 facilities in the 13 districts these had really reduced mortality. • Testing/assessment before treatment A Tanahashi model was employed as a for malaria and pneumonia improved backbone for the CODES project to assess from 23% to 69% at facility level in 2 the health system across the six determinants years of coverage on the supply and demand • Decreased health workers absenteeism side including (i) availability of commodities, from 44% to 29% in 2 years (ii) HR availability—health workers and VHTs, (iii) physical access to services—are • Increased male involvement: ANC with these health centers in reasonable reach?, spouses. Giving men a service has are the VHTs close enough and do they increased their involvement. have drugs in their medicine boxes? The demand side which could many times be Dr Mpanga concluded by sharing policy complex was simplified and looked at; (iv) prospects that would sustain the gains from initial utilization—if you have made available Codes and these included; all supply side services, will the services be utilized? Do they come for their 1st • Data: suggested that there should be a immunization when a child is born and what shift from expensive surveys to the use is their experience when they do come that of routine data: from LQAS to DHIS2 for will make them come again?, (v) continuous planning at district level. utilization and (vi) quality coverage. Dr Mpanga further illustrated that when a child • The national planning guidelines recently is suffering from pneumonia, the desire is to revised to include the use of Reproductive see whether there are antibiotics, whether Maternal New-born Child Adolescent there are health workers in the facility to Health (RMNCAH) score cards which prescribe the antibiotics and if they are they have digitalized into DHIS2—it there, whether they are accessible enough. has 24 indicators and at a glance the For this particular case, the CODES project different colors direct interrogation of saw that there were gaps among health performance, digitalized the bottleneck workers with antibiotics and hence showed analysis, and causal analysis and that children did not get their antibiotics community aspects to inform evidence within 24 hours of onset of illness. The based planning at district level and that

Confronting the Challenge of Child Mortality in Uganda 5 67TH STATE OF THE NATION PLATFORM

scale up of developing action trackers check-up every year. This data is analyzed for all districts and national planners was using a bottleneck assessment tool on an expected to start in November 2016. annual basis. Step 2: A causal analysis using the Dr. Agnes Kobusingye, Child management analysis and 5 WHYs model Fund International/ Liverpool was also conducted. Here a qualitative School of Tropical Medicine, Team survey of community and health providers was conducted using a management check leader CODES project list as well as results from the bottleneck analysis. Managerial short comings, Dr. Kobusingye presented about the decision space, resources, capacity and supply-side interventions. Lessons learned; motivation are identified through a forum challenges and prospects. She reiterated with district health teams and if a response the processes of the CODES supply side I was No, it was subjected to the 5 whys intervention by listing the steps undertaken model. as schematically presented below. Step 3: This process was done in fora with Step 1: In identifying bottlenecks, a district health teams. For each identified quantitative household survey in conducted cause (factor responsible for the observed by employing an LQAS. This was conducted gap in coverage), the district teams every 2 years. A health facility assessment suggested possible solutions. Potential which was conducted annually, a village solutions were prioritized using a rank- health team survey conducted every 2 years scoring approach that took into account and the Health Management Information effectiveness of the suggested intervention, System (HMIS) was used for mid-term feasibility, affordability and acceptability.

STEP 1 STEP 2 STEP 3 STEP 4 dentify Analyse root causes Identify Consult bottlenecks (casuality analysis) solutions Stakeholders & strategies

HMIS Supply side Analysis

Select & Stakeholders Equity and Quality-of- Management prioritise consultation solutions and to validate LQAS surveys Bottleneck Care Analysis Analysis Assessment develop findings and strategies solutions

Demand Facilities Barriers surveys Analysis

Equity and Community Management Bottleneck dialogues analysis, BNA, Assessment tool U-report, Citizen CQI

6 Confronting the Challenge of Child Mortality in Uganda 67TH STATE OF THE NATION PLATFORM

13 districts had color coded canvas charts. The red meant that they did not meet the standard and the green meant that they met the standard. The aim was to turn all the reds into green. Dr. Kobusingye went on to share some findings from the supply-side component of the CODES project such as;

• Health facility score cards: Figure 2: Dr. Agnes Kobusingye, CFI, making her presentation - She said that health facility staffs had been trained to score themselves with an aim to turn Sample findings from were all the reds into green. presented. For all the indicators under antibiotics for pneumonia for instance, an • District score cards: - These were improvement was realized and in several simple tools that leaders could use to other districts, they hoped to see a lot understand the status of events in health of these improvements from the endline facilities. findings as virtually all the determinants in • Aligning CODES project with already the Tanahashi model at the health facility existing cycles. If there was data for use and district have shown an upward trend in a bottleneck analysis shortly before but commodities. This was because the national budget framework, then one factors influencing the availability of drugs would see a lot of the output of analysis and commodities are usually beyond the of problems being transferred into the districts control. budget framework. Another example was at the health facility in Masaka which was working on the client • Leadership and Governance from MoH waiting time in the laboratory. By the time and the district managers were very CFI intervened, the average client waiting crucial: Support supervision reports: - time was 3 hours to pick results from the there was a book to guide health workers laboratory and they worked it out using on how to improve performance at the the continuously quality assessment health facility. However, some reports approach— the In-charge warned the lab do not indicate the gaps and actions to person about late coming and absenteeism, be taken by the health facility following introduced an attendance register and a the supervision. Some supervisors group approach to testing. As a result all, only made comments like “good work” these approaches put together reduced “struggle continues”… etc and this still on the client waiting time to less than 30 needed to improve. More still, health minutes. facilities were conducting staff and CFI has further worked with health facilities departmental meetings as required. and trained personnel on how to score However, the frequency of meetings was themselves as well as ably work with not regular, minutes were not available those scores. All the health facilities in the

Confronting the Challenge of Child Mortality in Uganda 7 67TH STATE OF THE NATION PLATFORM

in some facilities and they didn’t show a review of progress on action points from previous meetings. One of the reasons wass that, health In-charges were not trained to be administrators and performance would improve if they were trained.

Mr. Moses Mukundane, ACODE CODES project Team leader

Mr. Mukundane presented about the demand-side of the CODES intervention, lessons learnt, challenges and prospects. Figure 3: Moses Mukundane ACODE, In the demand-side component, ACODE presenting demand side of CODES Intervention th set out to mobilize and galvanize citizens at the 67 STON to demand and receive better quality of health services. A multitude of activities were conducted to achieve this goal. These included; (i) a baseline study, (ii) developing citizen report cards (CRC) and, (iii) conducting two day robust community Highlights of Achievement of the dialogues. Community dialogues involved: demand side (a) Day one where there was a discussion of CRC findings in separate breakout Highlights of achievements that the demand session of caretakers of children U5 side registered were: years, VHTs, community leaders and • The robust awareness creation in health workers and; (b) Day two where communities on basic health issues and there was an interface session combining collective action for child health survival breakout groups and sub-county/district through for example; leaders to dialogue which climaxed with the creation of joint action plans (community Establishment of community- contracts). Other activities included; (iv) initiated health outposts - i.e., Post-dialogue follow up/ monitoring visits Buwenda, Satellite of Butenga HC- 3-6 months after the date of dialogue, (v) IV, Bukomansimbi district and Kimi media campaigns/advocacy which involved island, Satellite of Koome HC-III, airing of public health messages on radio Mukono district and Increased latrine stations in 13 CODES districts, U-report coverage and other sanitary facilities SMS platform for both survey questions in most homesteads in communities and health promotion/education messages that hosted the dialogues- Role of and mTrac 8200 to capture community LCs, VHTs, and community-based compliments and complaints about health ‘CODES’ committees. services delivery issues. The last activity was of conducting a he demand-side end- Enhanced participatory planning for line study. community health/ child survival; Creation of Joint Action Plans to address challenges of healthcare

8 Confronting the Challenge of Child Mortality in Uganda 67TH STATE OF THE NATION PLATFORM

A snapshot of accomplished activities:

ACTIVITY RESULTS Baseline study and development 2 baseline study reports (Wave Zero and Wave of citizen report card One).

Community dialogues 25,800 Citizen Report Cards.

Media and advocacy 151 Community dialogues in 13 districts

Capacity building of CODES 13,355 participants (Caretakers of children district-based NGOs/ CBOS under five years, VHTs, Parish-level community leaders, sub-county and district leaders); 60% Capacity building of DHTs caretakers of children under five years

Post-dialogue monitoring 151 community contracts (“Joint Action Plans); Endline studies 60% implemented. 151 Community-initiated ‘CODES’ Committees

6,807 U-report enrolees; 30% are caretakers of children

8 Radio adverts / public health messages on most popular radio stations in 13 CODES districts

122 Key staff from 43 NGOs / CBOs

13 DHTs (DHEs)

161 Visits – health facility and community

2 Endline study reports (Wave zero and Wave one)

service delivery and utilization. There were a number of implementation Prompt response from sub-county challenges experienced by the demand and district leadership towards side of CODES project; critical issues emerging from community dialogues • Discrepancy between created Creation of new health facility knowledge through community outreach posts for immunization and dialogues and actual practice due to ANC both demand and supply barriers.

Increased knowledge among the • Unbalanced demand and supply- caretakers of children under five side interventions where demand was about prevention and treatment of increased for health services but could Malaria, Pneumonia, and Diarrhea. not be met with health/medical supplies

Confronting the Challenge of Child Mortality in Uganda 9 67TH STATE OF THE NATION PLATFORM

stock outs. He acknowledged the gap in health systems • Constrained civic competencies to financing. Money noted that is often spent enforce meaningful social accountability centrally (Kampala) and not distributed to due to vulnerability syndrome. For districts. Resources are earmarked and not example, community members fear to reallocated according to need/gaps and this send complaints using mTrac 8200. was because of the limited decision space that planners and managers experience. Associate Prof. Peter Waiswa Accordingly, discussions needed to be Makerere University School around the expansion of fiscal and decision of Public Health/ Karolinska space within the district. Institute, Team Leader CODES He re-echoed Dr. Mpanga and Dr. project Kobusingye’s recommendation to move away from conducting expensive surveys Assoc. Prof. Waiswa presented about but utilize the readily available data from what worked and perspectives from HMIS. This data is available and can be project monitoring and evaluation (Quality used for dashboard planning. Community Assurance). He said that the CODES dialogues are also effective for community hypothesis assumed that areas receiving engagement as they help to explain survey CODES intervention will perform “better” findings and guide planning/prioritization. and show accelerated “improvements” on However to ensure their sustainability, the key protective, preventive and curative dialogues should be aligned to already quality coverage indicators for pneumonia, existing community structures. diarrhea and malaria compared to those that have not received CODES intervention. District planners need to plan exit/ sustainability strategies learned from the The role of Makerere University was to CODES project while a consideration of monitor, learn and evaluate effective expanding the CODES project nationwide coverage and cost-effectiveness of should be considered. He noted that the the CODES project. It recognized the there was need to build the capacities of importance of using data within limited district health managers in order to improve resource districts to plan and manage the prioritization process towards targeted health services better based on identified child survival interventions. Community gaps. He said that when you have data dialogues should be integrated into existing and you use the data to plan better and community structures and lead by CSOs, understand which areas are doing well cultural groups etc. and which areas have gaps. You then use the gaps to allocate resources for better The combined cost of CODES per district implementation. Prof. Waiswa continued per annum amount to UGX 149,247,391 to say, projects brought to district need (or US$ 44,065). Scale up of CODES to align with the planning cycle, when the interventions to all the current 112 districts government is planning so that they can would require about UGX 16.7 billion (or leverage the resources of government and US$ 4.94 million) per annum [Adjusted for other partners. He urged planners and inflation UGX 18.2 Billion (or US$ 5.4 million)]. stakeholders to endeavor to align with the These estimates reflected the annual costs planning process. to be incurred after the initial year when set

10 Confronting the Challenge of Child Mortality in Uganda 67TH STATE OF THE NATION PLATFORM

Figure 4: A cross section of participants at the 67th STON session up activities have been undertaken. by men in communities and that they needed to be involved more. In Buvuma Prof Waiswa concluded by saying that this district, health facilities were inaccessible was a complex intervention that was able due to islands and poor and/or expensive to bring tools together at scale in many transport means. Health facilities continued districts and communities. It introduced to be understaffed with limited resources systematic evidence based planning for allocated to districts and limited decision child health to districts. However, despite space for managers to reallocate resources. the capacity built, limited district resources and decision space remain bottlenecks for sustainability. E. Policy Recommendations: Conclusion: Districts /DHTs and MoH The plenary provided more evidence on the achievements/successes of the • Need to adopt community dialogue CODES project. Participants recognized model in hard to reach and peculiar the shortcoming within the health systems communities to increase demand for financing as well as capacitating district health services particularly for children. health managers and health facility In- Ownership and accountability was an charges to plan and implement better important avenue for demand creation – for child health in districts. Participants but getting an effective model for scale also highlighted the limited roles played up could be a challenge.

Confronting the Challenge of Child Mortality in Uganda 11 67TH STATE OF THE NATION PLATFORM

• Need to prioritize quick wins amidst financial constraints; for instance, regular supervision of health facilities to crackdown of errant behavior of some health workers, especially those that charge illegal fees, suffer from habitual absenteeism, and/or use of abusive language, etc.).

• Need to strengthen the capacity of Health Unit Management Committees (HUMCs) through trainings and adequate financial facilitation.

• Need to sustain CODES gains through supporting local NGOs and CBOs involved in child health engagements at district and community levels. It is further important for implementing partners to develop exit strategies early on to allow for district ownership of intervention

• Build capacity of district health managers’ to plan and prioritize better, as well as, ensure that interventions are more targeted and not determined haphazardly

• Change resource allocation strategies: Despite the capacity built, limited district resources and decision space remain a bottleneck for sustainability.

• Data utilisation to foster targeted interventions: District planners should use Bottleneck, causal and management analysis tools in service areas beyond Child data to guide resource allocations and interventions.

12 Confronting the Challenge of Child Mortality in Uganda

Advocates Coalition for Development and Environment Plot 96, Kanjokya Street, Kamwokya, P.O.Box 29836, Kampala, Uganda Tel: +256 312 812 150 Email: [email protected] Website: www.acode-u.org