Drivers and Barriers to Improved Data Quality and Data-Use Practices: An Interpretative Qualitative Study in Addis Ababa, Ethiopia.

INTRODUCTION: An objective of the Information Revolution Roadmap of Ethiopia's Health Sector Transformation Plan was to improve health management information system (HMIS) data quality and data use at the point of health service delivery. We aimed to assess drivers of and barriers to improving HMIS data quality and use, focusing on key Information Revolution strategies including Connected Woreda, capacity building, performance monitoring teams, and motivational incentives. METHODS: We conducted an interpretative qualitative study across all 11 health centers in 3 subcities of Addis Ababa, Ethiopia: Yeka, Akaki-Kaliti, and Ledeta. A total of 40 key informant interviews and 6 focus group discussions with a total of 43 discussants were conducted. We coded information gathered line-by-line and grouped responses under thematic codes as they emerged. Findings were triangulated and validated. RESULTS: Our findings indicate that the main drivers of data quality and use at the point of service delivery were the use of the Connected Woreda strategy and its tools, capacity-building activities including mentorship, performance monitoring-team activities that led to active leadership engagement, and motivational incentives for data producers and users. Barriers to optimal data-use practices were the use of duplicative data collection tools at health facilities, under-developed health information system infrastructure, inadequate health information technician staffing and capacity limitations at the health facility level, insufficient leadership commitment, and unfavorable health worker attitudes toward data. DISCUSSION: Improvements in quality and use of HMIS data at health facilities are expected to result in delivering better-quality health services to the community as data enable health workers to identify gaps in health care, fix them, and monitor improvements. Future investments should focus on strengthening the promising data-use practices, resolving bottlenecks caused by duplicative data collection tools, enhancing individual and institutional capacity, addressing suboptimal health worker attitudes toward data, and overcoming infrastructure and connectivity challenges.

Lessons Learned From the Capacity-Building and Mentorship Program to Improve Health Information Systems in 11 Districts of Ethiopia.

INTRODUCTION: Health information systems (HIS) performance in Ethiopia is currently insufficient, and improvements are required to ensure that decision making is data driven. We share our experiences from the early-stage implementation of a package of HIS capacity-strengthening interventions as part of an innovative academic-government collaboration that addresses challenges in HIS performance. METHODS: We used routine program data to assess HIS performance using the Performance of Routine Information System Management (PRISM) assessment tools. The assessment employed a pre-post design and was conducted in a total of 24 selected health facilities (6 hospitals and 18 health centers) from 11 districts in Ethiopia at project baseline (2018) and midline (2020). RESULTS: Source document completeness rate reached less than 80% for the majority of the assessed data elements. Improvements were observed in quarterly report completeness (26% vs. 83%) and timeliness (17% vs. 48%). Though data inaccuracies are noted for all assessed data elements in 2020, the majority (83%) of skilled birth attendance and HIV reports (68%) fall within the acceptable range of reporting accuracy. The identification of performance-related problems, using performance monitoring team (PMT) meetings, improved between 2018 and 2020 (67% vs 89%). Similar improvements were also observed in developing action plans to solve identified problems via the PMT (52% in 2018 vs. 89% in 2020). Data use for planning and target setting (65% in 2018 vs. 90% in 2020), reviewing performance (58% in 2018 vs. 60% in 2020), and supervision (51% in 2018 vs. 53% in 2020) all improved among assessed health facilities. DISCUSSION: This study showed that a capacity-building and mentorship program that engages experts from multiple disciplines and sectors can improve the quality and use of health data. This partnership enabled engagement between government and academic stakeholders and allowed for a more robust exchange of resources and expertise toward HIS improvement.

Critical barriers to sustainable capacity strengthening in global health: a systems perspective on development assistance.

Background: Development assistance for health (DAH) is an important mechanism for funding and technical support to low-income countries. Despite increased DAH spending, intractable health challenges remain. Recent decades have seen numerous efforts to reform DAH models, yet pernicious challenges persist amidst structural complexities and a growing number of actors. Systems-based approaches are promising for understanding these types of complex adaptive systems. This paper presents a systems-based understanding of DAH, including barriers to achieving sustainable and effective country-driven models for technical assistance and capacity strengthening to achieve better outcomes Methods: We applied an innovative systems-based approach to explore and map how donor structures, processes, and norms pose challenges to improving development assistance models. The system mapping was carried out through an iterative co-creation process including a series of discussions and workshops with diverse stakeholders across 13 countries. Results: Nine systemic challenges emerged: 1) reliance on external implementing partners undermines national capacity; 2) prioritizing global initiatives undercuts local programming; 3) inadequate contextualization hampers program sustainability; 4) decision-maker blind spots inhibit capacity to address inequities; 5) power asymmetries undermine local decision making; 6) donor funding structures pose limitations downstream; 7) program fragmentation impedes long-term country planning; 8) reliance on incomplete data perpetuates inequities; and 9) overemphasis on donor-prioritized data perpetuates fragmentation. Conclusions: These interconnected challenges illustrate interdependencies and feedback loops manifesting throughout the system. A particular driving force across these system barriers is the influence of power asymmetries between actors. The articulation of these challenges can help stakeholders overcome biases about the efficacy of the system and their role in perpetuating the issues. These findings indicate that change is needed not only in how we design and implement global health programs, but in how system actors interact. This requires co-creating solutions that shift the structures, norms, and mindsets governing DAH models.

The effects of switching from 10 to 5-dose vials of MR vaccine on vaccination coverage and wastage: A mixed-method study in Zambia.

INTRODUCTION: Vaccines procured for low-income countries are often packaged in multi-dose vials to reduce program costs. To avoid wastage, health workers may refrain from opening a vial if few children attend an immunization session, possibly leading to lower coverage. Lowering the number of doses in a vial may increase coverage and reduce wastage. METHODS: We used a mixed methods approach to measure the effects of switching from conventional 10-dose measles containing vaccine (MCV) vials to 5-dose MCV vials on coverage and open vial wastage in 14 districts purposely selected from two provinces in Zambia. The districts were paired based on the number of health facilities and the average size of the health facility catchment population. One district from each pair was randomly allocated to receive 5-dose vials while the other continued with the conventional vials. We applied propensity score matched difference-in-difference analysis to estimate intervention effects on coverage using pre-intervention household survey and post-intervention household survey after 11 months of the intervention. The intervention effects on wastage rates were estimated from multivariate analysis of the administrative data. Key informant interviews were conducted to better understand health workers' behavior and preferences at baseline, midline and endline, and analyzed using thematic analysis techniques. RESULTS: MCV coverage rates increased across both arms for both doses. A five percentage-point intervention effect was detected for MCV1 and 3.5 percentage-point effect for MCV2. The MCV wastage rate was 47% lower in facilities using 5-dose vials (16.2%) versus 10-dose vials (30.5%). Healthcare workers reported being more willing to open a 5-dose vial than a 10-dose vial for one child, as they were less concerned about wastage. DISCUSSION: Switching 10-dose MCV vials to 5-dose vials improved coverage, decreased wastage, and improved willingness to open a vial. These findings can contribute to strategies for reducing missed opportunities for vaccination.

Mobile-Based Nutrition and Child Health Monitoring to Inform Program Development: An Experience From Liberia.

BACKGROUND: Implementing complex nutrition and other public health projects and tracking nutrition interventions, such as women's diet and supplementation and infant and young child feeding practices, requires reliable routine data to identify potential program gaps and to monitor trends in behaviors in real time. However, current monitoring and evaluation practices generally do not create an environment for this real-time tracking. This article describes the development and application of a mobile-based nutrition and health monitoring system, which collected monitoring data on project activities, women's nutrition, and infant and young child feeding practices in real time. PROGRAM DESCRIPTION: The Liberia Agricultural Upgrading Nutrition and Child Health (LAUNCH) project implemented a nutrition and health monitoring system between April 2012 and June 2014. The LAUNCH project analyzed project monitoring and outcome data from the system and shared selected behavioral and programmatic indicators with program managers through a short report, which later evolved into a visual data dashboard, during program-update meetings. The project designed protocols to ensure representativeness of program participants. FINDINGS: LAUNCH made programmatic adjustments in response to findings from the monitoring system; these changes were then reflected in subsequent quarterly trends, indicating that the availability of timely data allowed for the project to react quickly to issues and adapt the program appropriately. Such issues included lack of participation in community groups and insufficient numbers of food distribution points. Likewise, the system captured trends in key outcome indicators such as breastfeeding and complementary feeding practices, linking them to project activities and external factors including seasonal changes and national health campaigns. CONCLUSION: Digital data collection platforms can play a vital role in improving routine programmatic functions. Fixed gathering locations such as food distribution points represent an opportunity to easily access program participants and enable managers to identify strengths and weaknesses in project implementation. For programs that track individuals over time, a mobile tool combined with a strong database can greatly improve efficiency and data visibility and reduce resource leakages.