Learning from Ethiopia's success in reducing maternal and neonatal mortality through a health systems lens.

BACKGROUND: This study aimed to enhance insights into the key characteristics of maternal and neonatal mortality declines in Ethiopia, conducted as part of a seven-country study on Maternal and Newborn Health (MNH) Exemplars. METHODS: We synthesised key indicators for 2000, 2010 and 2020 and contextualised those with typical country values in a global five-phase model for a maternal, stillbirth and neonatal mortality transition. We reviewed health system changes relevant to MNH over the period 2000-2020, focusing on governance, financing, workforce and infrastructure, and assessed trends in mortality, service coverage and systems by region. We analysed data from five national surveys, health facility assessments, global estimates and government databases and reports on health policies, infrastructure and workforce. RESULTS: Ethiopia progressed from the highest mortality phase to the third phase, accompanied by typical changes in terms of fertility decline and health system strengthening, especially health infrastructure and workforce. For health coverage and financing indicators, Ethiopia progressed but remained lower than typical in the transition model. Maternal and neonatal mortality declines and intervention coverage increases were greater after 2010 than during 2000-2010. Similar patterns were observed in most regions of Ethiopia, though regional gaps persisted for many indicators. Ethiopia's progress is characterised by a well-coordinated and government-led system prioritising first maternal and later neonatal health, resulting major increases in access to services by improving infrastructure and workforce from 2008, combined with widespread community actions to generate service demand. CONCLUSION: Ethiopia has achieved one of the fastest declines in mortality in sub-Saharan Africa, with major intervention coverage increases, especially from 2010. Starting from a weak health infrastructure and low coverage, Ethiopia's comprehensive approach provides valuable lessons for other low-income countries. Major increases towards universal coverage of interventions, including emergency care, are critical to further reduce mortality and advance the mortality transition.

Effectiveness of a multi-country implementation-focused network on quality of care: Delivery of interventions and processes for improved maternal, newborn and child health outcomes.

The Network for Improving Quality of Care for Maternal, Newborn and Child Health (QCN) aims to work through learning, action, leadership and accountability. We aimed to evaluate the effectiveness of QCN in these four areas at the global level and in four QCN countries: Bangladesh, Ethiopia, Malawi and Uganda. This mixed method evaluation comprised 2-4 iterative rounds of data collection between 2019-2022, involving stakeholder interviews, hospital observations, QCN members survey, and document review. Qualitative data was analysed using a coding framework developed from underlying theories on network effectiveness, behaviour change, and QCN proposed theory of change. Survey data capturing respondents' perception of QCN was analysed with descriptive statistics. The QCN global level, led by the WHO secretariat, was effective in bringing together network countries' governments and global actors via providing online and in-person platforms for communication and learning. In-country, various interventions were delivered in 'learning districts', however often separately by different partners in different locations, and pandemic-disrupted. Governance structures for quality of care were set-up, some preceding QCN, and were found to be stronger and better (though often externally) resourced at national than local levels. Awareness of operational plans and network activities differed between countries, was lower at local than national levels, but increased from 2019 to 2022. Engagement with, and value of, QCN was perceived to be higher in Uganda and Bangladesh than in Malawi or Ethiopia. Capacity building efforts were implemented in all countries-yet often dependent on implementing partners and donors. QCN stakeholders agreed 15 core monitoring indicators though data collection was challenging, especially for indicators requiring new or parallel systems. Accountability initiatives remained nascent in 2022. Global and national leadership elements of QCN have been most effective to date, with action, learning and accountability more challenging, partner or donor dependent, remaining to be scaled-up, and pandemic-disrupted.

"We don't trust all data coming from all facilities": factors influencing the quality of care network data quality in Ethiopia.

BACKGROUND: Good quality data are a key to quality health care. In 2017, WHO has launched the Quality of Care Network (QCN) to reduce maternal, newborn and stillbirth mortality via learning and sharing networks. Guided by the principle of equity and dignity, the network members agreed to implement the programme in 2017-2021. OBJECTIVE: This paper seeks to explore how QCN has contributed to improving data quality and to identify factors influencing quality of data in Ethiopia. METHODS: We conducted a qualitative study in selected QCN facilities in Ethiopia using key informant interview and observation methods. We interviewed 40 people at national, sub-national and facility levels. Non-participant observations were carried out in four purposively selected health facilities; we accessed monthly reports from 41 QCN learning facilities. A codebook was prepared following a deductive and inductive analytical approach, coded using Nvivo 12 and thematically analysed. RESULTS: There was a general perception that QCN had improved health data documentation and use in the learning facilities, achieved through coaching, learning and building from pre-existing initiatives. QCN also enhanced the data elements available by introducing a broader set of quality indicators. However, the perception of poor data quality persisted. Factors negatively affecting data quality included a lack of integration of QCN data within routine health system activities, the perception that QCN was a pilot, plus a lack of inclusive engagement at different levels. Both individual and system capabilities needed to be strengthened. CONCLUSION: There is evidence of QCN's contribution to improving data awareness. But a lack of inclusive engagement of actors, alignment and limited skill for data collection and analysis continued to affect data quality and use. In the absence of new resources, integration of new data activities within existing routine health information systems emerged as the most important potential action for positive change.

Influences on policy-formulation, decision-making, organisation and management for maternal, newborn and child health in Bangladesh, Ethiopia, Malawi and Uganda: The roles and legitimacy of a multi-country network.

The Network for Improving Quality of Care for Maternal, Newborn and Child Health (QCN) is intended to facilitate learning, action, leadership and accountability for improving quality of care in member countries. This requires legitimacy-a network's right to exert power within national contexts. This is reflected, for example, in a government's buy-in and perceived ownership of the work of the network. During 2019-2022 we conducted iterative rounds of stakeholder interviews, observations of meetings, document review, and hospital observations in Bangladesh, Ethiopia, Malawi, Uganda and at the global level. We developed a framework drawing on three models: Tallberg and Zurn which conceptualizes legitimacy of international organisations dependent on their features, the legitimation process and beliefs of audiences; Nasiritousi and Faber, which looks at legitimacy in terms of problem, purpose, procedure, and performance of institutions; Sanderink and Nasiritousi, to characterize networks in terms of political, normative and cognitive interactions. We used thematic analysis to characterize, compare and contrast institutional interactions in a cross-case synthesis to determine salient features. Political and normative interactions were favourable within and between countries and at global level since collective decisions, collaborative efforts, and commitment to QCN goals were observed at all levels. Sharing resources and common principles were not common between network countries, indicating limits of the network. Cognitive interactions-those related to information sharing and transfer of ideas-were more challenging, with the bi-directional transfer, synthesis and harmonization of concepts and methods, being largely absent among and within countries. These may be required for increasing government ownership of QCN work, the embeddedness of the network, and its legitimacy. While we find evidence supporting the legitimacy of QCN from the perspective of country governments, further work and time are required for governments to own and embed the work of QCN in routine care.

Individual interactions in a multi-country implementation-focused quality of care network for maternal, newborn and child health: A social network analysis.

The Network for Improving Quality of Care for Maternal, Newborn and Child Health (QCN) was established to build a cross-country platform for joint-learning around quality improvement implementation approaches to reduce mortality. This paper describes and explores the structure of the QCN in four countries and at global level. Using Social Network Analysis (SNA), this cross-sectional study maps the QCN networks at global level and in four countries (Bangladesh, Ethiopia, Malawi and Uganda) and assesses the interactions among actors involved. A pre-tested closed-ended structured questionnaire was completed by 303 key actors in early 2022 following purposeful and snowballing sampling. Data were entered into an online survey tool, and exported into Microsoft Excel for data management and analysis. This study received ethical approval as part of a broader evaluation. The SNA identified 566 actors across the four countries and at global level. Bangladesh, Malawi and Uganda had multiple-hub networks signifying multiple clusters of actors reflecting facility or district networks, whereas the network in Ethiopia and at global level had more centralized networks. There were some common features across the country networks, such as low overall density of the network, engagement of actors at all levels of the system, membership of related committees identified as the primary role of actors, and interactions spanning all types (learning, action and information sharing). The most connected actors were facility level actors in all countries except Ethiopia, which had mostly national level actors. The results reveal the uniqueness and complexity of each network assessed in the evaluation. They also affirm the broader qualitative evaluation assessing the nature of these networks, including composition and leadership. Gaps in communication between members of the network and limited interactions of actors between countries and with global level actors signal opportunities to strengthen QCN.

Implementation of the 'Optimising the Health Extension Program' Intervention in Ethiopia: A Process Evaluation Using Mixed Methods.

An intervention called 'Optimising the Health Extension Program', aiming to increase care-seeking for childhood illnesses in four regions of Ethiopia, was implemented between 2016 and 2018, and it included community engagement, capacity building, and district ownership and accountability. A pragmatic trial comparing 26 districts that received the intervention with 26 districts that did not found no evidence to suggest that the intervention increased utilisation of services. Here we used mixed methods to explore how the intervention was implemented. A fidelity analysis of each 31 intervention activities was performed, separately for the first phase and for the entire implementation period, to assess the extent to which what was planned was carried out. Qualitative interviews were undertaken with 39 implementers, to explore the successes and challenges of the implementation, and were analysed by using thematic analysis. Our findings show that the implementation was delayed, with only 19% (n = 6/31) activities having high fidelity in the first phase. Key challenges that presented barriers to timely implementation included the following: complexity both of the intervention itself and of administrative systems; inconsistent support from district health offices, partly due to competing priorities, such as the management of disease outbreaks; and infrequent supervision of health extension workers at the grassroots level. We conclude that, for sustainability, evidence-based interventions must be aligned with national health priorities and delivered within an existing health system. Strategies to overcome the resulting complexity include a realistic time frame and investment in district health teams, to support implementation at grassroots level.

Quality antenatal care services delivery at health facilities of Ethiopia, assessment of the structure/input of care setting.

BACKGROUND: According to the Donabedian model, the assessment for the quality of care includes three dimensions. These are structure, process, and outcome. Therefore, the present study aimed at assessing the structural quality of Antenatal care (ANC) service provision in Ethiopian health facilities. METHODS: Data were obtained from the 2018 Ethiopian Service Availability and Readiness Assessment (SARA) survey. The SARA was a cross-sectional facility-based assessment conducted to capture health facility service availability and readiness in Ethiopia. A total of 764 health facilities were sampled in the 9 regions and 2 city administrations of the country. The availability of equipment, supplies, medicine, health worker's training and availability of guidelines were assessed. Data were collected from October-December 2017. We run a multiple linear regression model to identify predictors of health facility readiness for Antenatal care service. The level of significance was determined at a p-value < 0.05. RESULT: Among the selected health facilities, 80.5% of them offered Antenatal care service. However, the availability of specific services was very low. The availability of tetanus toxoid vaccination, folic acid, iron supplementation, and monitoring of hypertension disorder was, 67.7, 65.6, 68.6, and 75.1%, respectively. The overall mean availability among the ten tracer items that are necessary to provide quality Antenatal care services was 50%. In the multiple linear regression model, health centers, health posts and clinics scored lower Antenatal care service readiness compared to hospitals. The overall readiness index score was lower for private health facilities (ß = - 0.047, 95% CI: (- 0.1, - 0.004). The readiness score had no association with the facility settings (Urban/Rural) (p-value > 0.05). Facilities in six regions except Dire Dawa had (ß = 0.067, 95% CI: (0.004, 0.129) lower readiness score than facilities in Tigray region (p-value < 0.015). CONCLUSION: This analysis provides evidence of the gaps in structural readiness of health facilities to provide quality Antenatal care services. Key and essential supplies for quality Antenatal care service provision were missed in many of the health facilities. Guaranteeing properly equipped and staffed facilities shall be a target to improve the quality of Antenatal care services provision.

Factors associated with health care provider knowledge on abortion care in Ethiopia, a further analysis on emergency obstetric and newborn care assessment 2016 data.

BACKGROUND: Abortion is one of the major direct causes of maternal death, accounting for 7.9% globally. In Africa, 5.5 million women have unsafe abortions annually. Although maternal deaths due to complications of abortion have declined in Ethiopia, women still die from complications. Few studies have focused on providers' clinical knowledge. This study investigates the level of health workers' knowledge of comprehensive abortion care and its determinants in Ethiopia. METHODS: Data from the national emergency obstetric and newborn care (EmONC) assessment was used. A total of 3804 facilities that provided institutional deliveries in the 12 months before the assessment were included. Provider knowledge was assessed by interviewing a single provider from each facility. Criteria for selection included: having attended the largest number of deliveries in the last one or two months. A summary knowledge score was generated based on the responses to three knowledge questions related to immediate complications of unsafe abortion, how a woman should be clinically managed and what the counselling content should contain. The score was classified into two categories (< 50% and > =50%). Logistic regression was used to determine individual and facility-level factors associated with the summary knowledge score. RESULT: A total of 3800 providers participated and the majority were midwives, nurses and health officers. On average, providers identified approximately half or fewer of the expected responses. The multivariate model showed that midwives and nurses (compared to health officers), being female, and absence of training or practice of manual vacuum aspiration were associated with lower knowledge levels. Important facility level factors protective against low knowledge levels included employment in Addis Ababa, being male and having internet access in the facility. CONCLUSION: To increase knowledge levels among providers, pre- and in-service training efforts should be particularly sensitive to female providers who scored lower, ensure that more midlevel providers are capable of performing manual vacuum aspiration as well as provide special attention to providers in the Gambella.