The building blocks of community health systems: a systems framework for the design, implementation and evaluation of iCCM programs and community-based interventions.

INTRODUCTION: Almost all sub-Saharan African countries have adopted some form of integrated community case management (iCCM) to reduce child mortality, a strategy targeting common childhood diseases in hard-to-reach communities. These programs are complex, maintain diverse implementation typologies and involve many components that can influence the potential success of a program or its ability to effectively perform at scale. While tools and methods exist to support the design and implementation of iCCM and measure its progress, these may not holistically consider some of its key components, which can include program structure, setting context and the interplay between community, human resources, program inputs and health system processes. METHODS: We propose a Global South-driven, systems-based framework that aims to capture these different elements and expand on the fundamental domains of iCCM program implementation. We conducted a content analysis developing a code frame based on iCCM literature, a review of policy documents and discussions with key informants. The framework development was guided by a combination of health systems conceptual frameworks and iCCM indices. RESULTS: The resulting framework yielded 10 thematic domains comprising 106 categories. These are complemented by a catalogue of critical questions that program designers, implementers and evaluators can ask at various stages of program development to stimulate meaningful discussion and explore the potential implications of implementation in decentralised settings. CONCLUSION: The iCCM Systems Framework proposed here aims to complement existing intervention benchmarks and indicators by expanding the scope and depth of the thematic components that comprise it. Its elements can also be adapted for other complex community interventions. While not exhaustive, the framework is intended to highlight the many forces involved in iCCM to help managers better harmonise the organisation and evaluation of their programs and examine their interactions within the larger health system.

Early Experiences in the Integration of Non-communicable Diseases into Emergency Primary Health Care, Beni Region, Democratic Republic of the Congo.

Background: Health services in humanitarian crises increasingly integrate the management of non-communicable diseases into primary care. As there is little description of such programs, this case study aims to describe the initial implementation of non-communicable disease management within emergency primary care in the conflict-affected Beni Region of Democratic Republic of the Congo (DRC). Objectives: We implemented and evaluated a primary care approach to hypertension and diabetes management to assess the feasibility of patient monitoring, early clinical and programmatic outcomes, and costs, after seven months of care. Methods: We designed clinical and programmatic modules for diabetes and hypertension management for clinical officers and the use of patient cards and community health workers to improve adherence. We used cohort analysis (April to October 2018), time-trend analysis, semi-structured interviews, and costing to evaluate the program. Findings: Increases in consultations for hypertension (incidence rate ratio [IRR] 13.5, 95% CI 5.8-31.5, p < 0.00) and diabetes (IRR 3.6, 95% CI 1-12.9, p < 0.05) were demonstrated up to the onset of violence and an Ebola epidemic in August 2018. Of 833 patients, 67% were women of median age 56. Nearly all were hypertensives (88.7%) and newly diagnosed (95.9%). Treatment adherence, defined as attending ≥2 visits in the seven month period, was demonstrated by 45.4% of hypertension patients. Community health workers had contact with 3.2-3.8 patients per month. Respondents stated that diabetes care remained fragmented with insulin and laboratory testing located outside of primary care. Program and management costs were 115 USD per person per treatment course. Conclusions: In an active conflict setting, we demonstrated that non-communicable disease care can be well-organized through clinical training and cohort analysis, and adherence can be addressed using patient-held cards and monitoring by community health workers. Nearly all diagnoses were new, emphasizing the need to establish self-management. Insecurity reduced access for patients but care continued for a subset of patients during the Ebola epidemic.

Testing a simplified tool and training package to improve integrated Community Case Management in Tanganyika Province, Democratic Republic of Congo: a quasi-experimental study.

BACKGROUND: Integrated community case management (iCCM) is a strategy to train community health workers (relais communautaires or RECOs in French) in low-resource settings to provide treatment for uncomplicated malaria, pneumonia, and diarrhea for children 2-59 months of age. The package of Ministry of Public Health tools for RECOs in the Democratic Republic of Congo that was being used in 2013 included seven data collection tools and job aids which were redundant and difficult to use. As part of the WHO-supported iCCM program, the International Rescue Committee developed and evaluated a simplified set of pictorial tools and curriculum adapted for low-literate RECOs. METHODS: The revised training curriculum and tools were tested in a quasi-experimental study, with 74 RECOs enrolled in the control group and 78 RECOs in the intervention group. Three outcomes were assessed during the study period from Sept. 2015-July 2016: 1) quality of care, measured by direct observation and reexamination; 2) workload, measured as the time required for each assessment - including documentation; and 3) costs of rolling out each package. Logistic regression was used to calculate odds ratios for correct treatment by the intervention group compared to the control group, controlling for characteristics of the RECOs, the child, and the catchment area. RESULTS: Children seen by the RECOs in the intervention group had nearly three times higher odds of receiving correct treatment (adjusted odds ratio aOR = 2.9, 95% confidence interval CI = 1.3-6.3, P = 0.010). On average, the time spent by the intervention group was 10.6 minutes less (95% CI = 6.6-14.7, P < 0.001), representing 6.2 hours of time saved per month for a RECO seeing 35 children. The estimated cost savings amounts to over US$ 300 000 for a four-year program supporting 1500 RECOs. CONCLUSION: This study demonstrates that, at scale, simplified tools and a training package adapted for low-literate RECOs could substantially improve health outcomes for under-five children while reducing implementation costs and decreasing their workload. The training curriculum and simplified tools have been adopted nationally based on the results from this study.

Universal versus conditional day 3 follow-up for children with non-severe unclassified fever at the community level in the Democratic Republic of the Congo: A cluster-randomized, community-based non-inferiority trial.

BACKGROUND: The World Health Organization's integrated community case management (iCCM) guidelines recommend that all children presenting with uncomplicated fever and no danger signs return for follow-up on day 3 following the initial consultation on day 1. Such fevers often resolve rapidly, however, and previous studies suggest that expectant home care for uncomplicated fever can be safely recommended. We aimed to determine if a conditional follow-up visit was non-inferior to a universal follow-up visit for these children. METHODS AND FINDINGS: We conducted a cluster-randomized, community-based non-inferiority trial among children 2-59 months old presenting to community health workers (CHWs) with non-severe unclassified fever in Tanganyika Province, Democratic Republic of the Congo. Clusters (n = 28) of CHWs were randomized to advise caregivers to either (1) return for a follow-up visit on day 3 following the initial consultation on day 1, regardless of illness resolution (as per current WHO guidelines; universal follow-up group) or (2) return for a follow-up visit on day 3 only if illness continued (conditional follow-up group). Children in both arms were assessed again at day 8, and classified as a clinical failure if fever (caregiver-reported), malaria, diarrhea, pneumonia, or decline of health status (development of danger signs, hospitalization, or death) was noted (failure definition 1). Alternative failure definitions were examined, whereby caregiver-reported fever was first restricted to caregiver-reported fever of at least 3 days (failure definition 2) and then replaced with fever measured via axillary temperature (failure definition 3). Study participants, providers, and investigators were not masked. Among 4,434 enrolled children, 4,141 (93.4%) met the per-protocol definition of receipt of the arm-specific advice from the CHW and a timely day 8 assessment (universal follow-up group: 2,210; conditional follow-up group: 1,931). Failure was similar (difference: -0.7%) in the conditional follow-up group (n = 188, 9.7%) compared to the universal follow-up group (n = 230, 10.4%); however, the upper bound of a 1-sided 95% confidence interval around this difference (-∞, 5.1%) exceeded the prespecified non-inferiority margin of 4.0% (non-inferiority p = 0.089). When caregiver-reported fever was restricted to fevers lasting ≥3 days, failure in the conditional follow-up group (n = 159, 8.2%) was similar to that in the universal follow-up group (n = 200, 9.1%) (difference: -0.8%; 95% CI: -∞, 4.1%; p = 0.053). If caregiver-reported fever was replaced by axillary temperature measurement in the definition of failure, failure in the conditional follow-up group (n = 113, 5.9%) was non-inferior to that in the universal follow-up group (n = 160, 7.2%) (difference: -1.4%; 95% CI: -∞, 2.5%; p = 0.012). In post hoc analysis, when the definition of failure was limited to malaria, diarrhea, pneumonia, development of danger signs, hospitalization, or death, failure in the conditional follow-up group (n = 108, 5.6%) was similar to that in the universal follow-up group (n = 147, 6.7%), and within the non-inferiority margin (95% CI: -∞, 2.9%; p = 0.017). Limitations include initial underestimation of the proportion of clinical failures as well as substantial variance in cluster-specific failure rates, reducing the precision of our estimates. In addition, heightened security concerns slowed recruitment in the final months of the study. CONCLUSIONS: We found that advising caregivers to return only if children worsened or remained ill on day 3 resulted in similar rates of caregiver-reported fever and other clinical outcomes on day 8, compared to advising all caregivers to return on day 3. Policy-makers could consider revising guidelines for management of uncomplicated fever within the iCCM framework. TRIAL REGISTRATION: ClinicalTrials.gov NCT02595827.

Universal versus conditional three-day follow up visit for children with uncomplicated fever at the community level: design of a cluster-randomized, community-based, non-inferiority trial in Tanganyika, Democratic Republic of Congo.

BACKGROUND: The current recommendation within integrated Community Case Management guidelines that all children presenting with uncomplicated fever and no danger signs be followed up after three days may not be necessary. Such fevers often resolve rapidly (usually within 48-96 h), and previous studies suggest that expectant home care for uncomplicated fever can be safely recommended. We aim to determine the non-inferiority of a conditional versus a universal follow-up visit for these children. METHODS: We are conducting a cluster-randomized, community-based, non-inferiority trial enrolling ~4300 children (ages 2-59 months) presenting to community health workers (CHWs) with uncomplicated fever in Tanganyika Province, Democratic Republic of the Congo. Clusters (n = 28) of CHWs are randomized to advise caretakers of such children to either 1) return for a follow-up visit on Day 3 following the initial consultation (Day 1), regardless of illness resolution (as per current guidelines) or 2) return for a follow-up visit on Day 3 only if the child's signs have not resolved. Enrolled children are followed up at Day 7 for a repeat assessment and recording of the primary outcome of the study, "failure", which is defined as having fever, diarrhea, pneumonia or decline of health status (e.g. hospitalization, presenting danger signs, or death). DISCUSSION: The results of this trial will be interpreted in conjunction with a similarly designed trial currently ongoing in Ethiopia. If a follow-up visit conditional on continued illness is shown to be non-inferior to current guidelines stipulating universal follow-up, appropriate updating of such guidelines could reduce time and human resource pressures on both providers and caregivers throughout communities of sub-Saharan Africa and South Asia. TRIAL REGISTRATION: This trial was registered at ClinicalTrials.gov ( NCT02595827 ) on November 2nd, 2015.

Update on mortality in the Democratic Republic of Congo: results from a third nationwide survey.

BACKGROUND: The humanitarian crisis in the Democratic Republic of Congo (DRC) has been among the world's deadliest in recent decades. We conducted our third nationwide survey to examine trends in mortality rates during a period of changing political, security, and humanitarian conditions. METHODS: We used a 3-stage, household-based cluster sampling technique to compare east and west DRC. Sixteen east health zones and 15 west zones were selected with a probability proportional to population size. Four east zones were purposely selected to allow historical comparisons. The 20 smallest population units were sampled in each zone, 20 households in each unit. The number and distribution of households determined whether they were selected using systematic random or random walk sampling. Respondents were asked about deaths of household members during the recall period: January 2006-April 2007. FINDINGS: In all, 14,000 households were visited. The national crude mortality rate of 2.2 deaths per 1000 population per month (95% confidence interval [CI] 2.1-2.3) is almost 70% higher than that documented for DRC in the 1984 census (1.3) and is unchanged since 2004. A small but significant decrease in mortality since 2004 in the insecure east (rate ratio: 0.96, P = .026) was offset by increases in the western provinces and a transition area in the center of the country. Nonetheless, the crude mortality rate in the insecure east (2.6) remains significantly higher than in the other regions (2.0 and 2.1, respectively). Deaths from violence have declined since 2004 (rate ratio 0.7, P = .02). CONCLUSIONS: More than 4 years after the official end of war, the crude mortality rate remains elevated across DRC. Slight but significant improvements in mortality in the insecure east coincided temporally with recent progress on security, humanitarian, and political fronts.

Mortality in the Democratic Republic of Congo: a nationwide survey.

BACKGROUND: Commencing in 1998, the war in the Democratic Republic of Congo has been a humanitarian disaster, but has drawn little response from the international community. To document rates and trends in mortality and provide recommendations for political and humanitarian interventions, we did a nationwide mortality survey during April-July, 2004. METHODS: We used a stratified three-stage, household-based cluster sampling technique. Of 511 health zones, 49 were excluded because of insecurity, and four were purposely selected to allow historical comparisons. From the remainder, probability of selection was proportional to population size. Geographical distribution and size of cluster determined how households were selected: systematic random or classic proximity sampling. Heads of households were asked about all deaths of household members during January, 2003, to April, 2004. FINDINGS: 19,500 households were visited. The national crude mortality rate of 2.1 deaths per 1000 per month (95% CI 1.6-2.6) was 40% higher than the sub-Saharan regional level (1.5), corresponding to 600,000 more deaths than would be expected during the recall period and 38,000 excess deaths per month. Total death toll from the conflict (1998-2004) was estimated to be 3.9 million. Mortality rate was higher in unstable eastern provinces, showing the effect of insecurity. Most deaths were from easily preventable and treatable illnesses rather than violence. Regression analysis suggested that if the effects of violence were removed, all-cause mortality could fall to almost normal rates. INTERPRETATION: The conflict in the Democratic Republic of Congo remains the world's deadliest humanitarian crisis. To save lives, improvements in security and increased humanitarian assistance are urgently needed.