CODES project description
CODES is a district focused health systems management and community empowerment project that seeks to improve effective coverage and quality of child survival interventions. CODES combines UNICEF tools designed to systematize priority setting, allocation of resources and problem solving with Community dialogues based on Citizen Report Cards and U-Reports used to engage and empower communities in monitoring health service provision and to demand for quality services. The first 2 years of the project were a proof of concept, during which time the tools were adapted to the local situation and their adoption piloted in five districts. The Project has two implementing partners, Child Fund International (CFI), Liverpool School of Tropical Medicine (LSTM), who work together and are responsible for the supply side and Advocates Coalition for Development and Environment (ACODE) that is responsible for the demand side.
The tools combined by CODES include; LQAS, Bottleneck analysis, Causal analysis, Continuous Quality Improvement (CQI), Community Dialogues based on Citizen Report Cards and U reports. Figure 1 shows how the tools were supposed to come together and reinforce each other. Lot Quality Assurance Sampling (LQAS) surveys were used to generate district specific data on service provision and healthcare use at community and health facility levels [20]. The surveys focused on availability of essential healthcare inputs; healthcare seeking behavior, coverage and quality of high impact interventions or practices at both community and health facility levels. The LQAS was also used to classify Supervision Areas as high or low performing relative to four predetermined standards set in consultation with program managers [21]. The data were then aggregated at district level and presented in the form of percentages with regard to coverage of interventions.
Based on the data generated through LQAS a bottleneck analysis tool was then used to assess health system constraints. Using the Tanahasi Model the DHMTs with support from CFI/LSTM identified bottlenecks in the operation of the health system; analyzed the constraining factors responsible for bottlenecks and selected effective measures for improving service performance and quality. Each DHMT identified five key bottlenecks affecting quality delivery of interventions for managing diarrhea, pneumonia, immunization and malaria at health facility and community (VHT) level. ACODE conducted a similar analysis of high-priority demand-side bottlenecks in one randomly selected community per Supervisory Area, for each district. All the information was then fed into the Tanahasi model which generated graphic displays that were presented to the district health teams, planners and policy makers for discussion.
For each of the selected bottlenecks an in-depth analysis was conducted to find the causes of the bottlenecks and of the management issues contributing to the bottlenecks. The causal analysis tool enabled DHMTs to determine areas needing management capacity-building and to reprioritize schedules for supervisory visits guided by the performance of the different supervision areas. Possible solutions and strategies were then identified and incorporated into the district annual health work plans with more resources being targeted to low performing Supervision Areas of the district [21].
Continuous Quality improvement (CQI) was used to aid the implementation of the identified priorities. CQI is meant to involve managers and service providers in continuous improvement of work processes to achieve better outcomes using the Plan, Do, Study, and Act (PDSA) cycles of the Quality Improvement (QI) Model approach to address priority problems at district and Health Facility (HF) levels. District and Health Facility Quality Improvement Teams were set up in all the pilot health facilities and were trained, mentored and supported by the implementing partner. Two Peer-to-peer learning [22, 23] workshops were held to facilitate DHMTs to share and learn from each other’s experiences and innovations.
In order to engage and empower communities in monitoring health service provision and to demand for quality services, Citizen Report Cards (CRCs) were developed with factual information generated from the LQAS and qualitative surveys to facilitate community dialogues. The CRCs were initially piloted and then revised based on the feedback before rollout. Each district had a Citizen Report Card that comprised of two A4 sheets presenting district specific quantitative and qualitative information on immunization coverage, utilization of services for pneumonia, diarrhea and malaria for children under five, availability of essential commodities, availability of human resources, geographical accessibility, initial, adequate and continuous utilization of health service for children under five. Communities during the Community Dialogues based their opinions, action plans and strategies on the information published in the CRCs. The CRCs also enabled the local communities to make their own appraisals, analyses, and plans and to monitor and evaluate the results. Additional file 1: Figure S1 shows a sample of a CRC.
Community dialogues were intended to mobilize and galvanize communities to demand effective and quality health service delivery and to promote accountability through the use of SMS media platforms such as U-Report [24]. Equity was a consideration in organizing Community Dialogues thus enabling underserved communities to articulate their barriers to access and utilization. The Community dialogue brought together 70 to 100 people comprised of parents/caretakers of children under 5 years, community leaders (political, cultural, religious and social), VHTs and health workers. The dialogues lasted 2 days. In order to get the participants to commit to the 2 days and to keep time breakfast and lunches were served at the venues. On the first day after the initial introductions the group was divided into four breakout sessions comprised of; caretakers, VHTs, community leaders and health workers. Health workers held their sessions at nearby health units while the other three shared a venue. During these sessions each of the breakout groups brainstormed on the problems that they face with health care utilization and delivery; discussed which of the problems required government assistance to address and which problems they could as communities address with relatively little if any assistance from outside groups or funders. As part of this process each group was required to name the problem, the cause, and the solutions, those responsible for addressing the problem and who would take responsibility and when the activity would begin. During the second day the groups were brought back together for an interface meeting between the community of users and health service providers. Together the groups identified overlapping action steps developed the previous day and action steps identified by each of the groups that required the activities and buy-in of other groups. All participants then voted on a dialogue wide contract and elected a CODES committee to oversee the activities. In order to get the communities to own the process CODES Committees were established with committee representatives from each of the villages represented to ensure follow through of the action plans developed. Post dialogue follow-up was conducted to establish the improvements that had occurred as a result of this intervention. During the dialogues the participants were also introduced to U-Report an SMS monitoring tool that was both free and anonymous. Participants were asked to register and use it as a means of providing feedback. Following the community dialogues ACODE sent out U-Report surveys that asked respondents about their general participation in any of the post-dialogue activities.
Study setting: service delivery
Uganda’s health services are devolved to District Local Governments (DLGs) and delivered by the district health services led by the District Health Teams (DHTs) [25]. Basic health services are provided through a referral structure consisting of; Village Health Teams (comprised of community volunteers), and health facilities (including Health Centres II, III and IV and hospitals). Most districts have NGOs and other agencies operating in the health sector but most of their interventions are managed vertically and are not always in line with district priorities. While district health office has control over funds allocated by the central government for health, most of this funding is usually ear-marked, leaving the district health managers with minimal “fiscal space” for re-allocation of resources to address needs identified at district level.
In 2009, the standard Health Facility Population Ratio for health centre IV, III, and II was 1: 100,000, 1: 20,000 and 1: 5000 respectively. The Actual situation for health centre IV, III and II was 1: 187,500, 1: 84,507 and 1: 14,940 respectively [26]. Districts are catering for a substantial proportion of the population with DHTs making decisions, planning and providing services for this proportion of the population.
Uganda represents an excellent real-world opportunity to examine the effects of developing a prioritized package of interventions. Uganda has developed a comprehensive National Child Survival Strategy 2010–2015 [27], with 15 interventions at family/community level, 12 at population level, and 17 at health facility level to be delivered by village health teams.
Study design
This was qualitative study. This design was chosen because it allows for in-depth exploration of the experiences of the DHTs and other partners in the adaptation, integration and implementation of the various tools [28, 29].
Study sites and selection criteria
The study was conducted in five districts in Uganda, Bukomansibi, Masaka, Buikwe, Mukono and Wakiso. The criteria used by the CODES project to select the districts included: high child mortality rates and the representation of both new and old districts. Bukomansimbi and Buikwe were both new districts, while Mukono, Wakiso and Masaka were old districts.
Study participants and sample size
Thirty eight participants were selected for the study. These included members of the DHTs, district planners, Chief Administrative Officers in the five pilot districts and officials from the two implementing partners. As this was a qualitative study the sample size was determined by data saturation point, a point at which further interviews generate no new information about the subject of investigation.
Study team
The research team consisted of a Ugandan Social Scientist experienced in qualitative research (AK), two Ugandan public health specialists (DKH) and (PW) with previous experience as Heads of a DHT, a Ugandan research assistant (ES), a Ugandan biostatistician (DB) and a Swedish health systems specialist (SSP) with experience in health systems including Uganda. No one on the team was working within the district health system.
Sampling method for participants
The research participants were selected using a purposive sampling method [30]. This method allowed for selection of participants based on their knowledge and participation in CODES related activities.
Participants’ recruitment procedure
The participants were invited to participate in the study through the DHO’s office. Appointments were then arranged by the research team and face-to-face interviews conducted.
Data collection method
The study was conducted from January 2012 to December 2013. Individual interviews (IDIs) were used to collect data. This was the method of choice because it allowed the participants to reflect on their individual experiences which would be more difficult in focus groups. All IDIs were conducted in English since all the participants were fluent in English. During the interviews the study participants were asked to reflect and describe [31] the processes they normally went through when planning and setting priorities for their districts; their experience with the CODES project, the usefulness of the tools; their willingness to continue using them and their perceived constraints. Probing was done to get a deeper understanding of the experiences [32, 33]. Audio recording of the interviews was done. Participants’ confidentiality was maintained. Each interview lasted at most 60 min.
Interview guide
A semi-structured interview guide was used for data collection. The guide was specifically developed with the aim of understanding from the perspective of the District Health Management Team members their experiences in the adoption and alignment of the CODES intervention within the district planning processes in the five pilot districts. The guide further explored their perspectives regarding the usefulness of the CODES intervention and their willingness to adopt the various tools as part of their routine efforts in district management capacity strengthening and empowering communities to utilize and demand quality health services.
Data analysis
Data from the audio recording device was transcribed verbatim. Data from the interviews were supplemented by observations during the implementation and by information extracted from the implementation reports and district plans [34]. Interview transcripts were read a number of times to get a general understanding of the material and the emerging themes coded. Thematic analysis [35] was used to distill the experiences of adoption and implementation of the intervention and the lessons learnt in the process.
Ethical considerations
Ethical clearance was obtained from Uganda National Council for Science and Technology (UNCST-SS 2548) to conduct this study. Permission to conduct the study was also sought from the District Health office in all the five districts. Individual consent was obtained from all the participants prior to being interviewed.