Tracking events to inform a health systems model of maternal healthcare outcomes in urban settings Lessons learned from Lubumbashi, DRC.

Rationale: In urban settings, provision and use of maternal healthcare services is affected by complex dynamics and events, such as natural disasters, traffic, policies, and community activities. By systematically tracking such events, we can develop a health systems’ model for simulating how disruptions influence provision and access to maternal healthcare. The model will inform strategies that anticipate system shocks and ensure accessibility to timely urban maternal healthcare services, thus improving health outcomes. This study presents the first phase of model development, describing documentation of events at micro, meso and macro levels in Lubumbashi (DRC), sharing lessons on data collection and the impact of events on maternal healthcare, which will serve as a foundation for model development.Method: Since August 2024, primary data collection of events has been carried out by trained investigators. Information was collected at health facilities, city, regional and national levels from various sources (media, government announcement, healthcare facilities, etc.). Data is entered on a KoboToolbox form and weekly meetings are organized to review this data and evaluate the data collection process. Descriptive analysis is performed, providing an overview of events and their foreseen effect on maternal healthcare provision and use.Result: We documented 156 events over 40 weeks, with a median of 2 events per week. Most events occurred in Lubumbashi city (n=52) and at national level (n=48). Disruption events included health policy changes (n=38), infectious disease outbreaks (n=15), election campaigns (n=12), and demonstrations (n=12). Most events lasted one day (n=55), with sources of information from healthcare facilities (n=83), government announcements (n=75) and media (n=36). Events had direct and indirect impacts on healthcare, e.g., early discharge of patients from hospitals due to doctors’ strike, geographical inaccessibility to services due to blocked bridge, etc. Lessons learned include the importance of real-time data entry, documenting lag in effect, and reflections on event-health system links.Conclusion: Collecting event data is time-consuming and requires a skilled team with tools adaptability, and task allocation. This vital information helps us understand how events disrupt maternal healthcare services and will be used to develop a model for testing strategies that promote system resilience to shocks and continued healthcare access.