Health system resilience: quantifying the dynamic impact of environmental shocks on health service utilization using an interrupted time series and time-series forecasting approach in Western Province, Zambia.

INTRODUCTION: In low- and lower-middle-income nations, advancements toward Universal Health Coverage (UHC) are progressively jeopardized by the combined effects of acute shocks, such as floods and pandemics, alongside chronic health system stressors. This study employs a multi-model time-series methodology to dynamically forecast health service utilization and assess the impacts of major shocks in Western Province, Zambia.

MATERIALS AND METHODS: We conducted a longitudinal ecological study utilizing 73 months of routine health data from 62 healthcare facilities (October 2017 to September 2023). We used Prophet and Hierarchical Time Series (HTS) models for forecasting and an Interrupted Time Series (ITS) analysis to quantify the impacts of a drought, the COVID-19 pandemic, and a double-peak flood event.

RESULTS: The Prophet model was the most accurate forecasting tool (MAPE = 7.22 %). The ITS analysis demonstrated that each shock distinctly affected health service utilization. The drought was associated with an immediate decline in utilization (level change = -0.105, p < 0.001), while the COVID-19 pandemic resulted in an immediate increase (level change = 0.069, p < 0.001). The 2023 flood event showed no immediate impact but was associated with a significant positive trend change (trend change = 0.019, p < 0.001).

DISCUSSION: This study provides innovative empirical evidence demonstrating that the interaction between environmental shocks and health system stressors dynamically affects health service utilization.

CONCLUSION: The findings underscore the significance of adopting multi-model predictive strategies to produce early warnings and to inform targeted, data-driven interventions to strengthen health system resilience.