A Two-Pronged Analytical Approach to Transmission Dynamics with Spatial Heterogeneity: Epidemic Curve Profiling and Mobility-Adjusted Rt Estimation

Background. Capturing spatial heterogeneity in disease transmission is essential for effective and timely public health interventions. However, conventional methods for estimating the effective reproduction number (Rt) often assume homogeneous mixing and may not adequately capture spatial connectivity—such as human mobility—potentially obscuring local transmission risks. Shape-based analyses of epidemic curves characterize timing and recurrence of waves but remain descriptive and cannot quantify transmission intensity. Because each approach offers only a partial view of regional dynamics, an integrative framework is needed to capture both epidemic structure and dynamics. Methods. We present an integrative approach that couples mobility-adjusted Rt estimation with shape-based clustering of epidemic growth curves. Using COVID-19 case and mobility data from 17 South-Korean provinces during the first wave (February–April 2020), we estimate Rt within a sliding window that incorporates inter-regional mobility matrices derived from telecommunication records, thereby weighting cross-regional transmission potential. Simultaneously, we apply dynamic time warping and hierarchical clustering to align and group epidemic curves based on morphological similarity. Results. Our analysis reveals three distinct regional outbreak profiles: (1) prolonged, staggered outbreaks in high-mobility metropolitan areas; (2) synchronized, sharp surges followed by rapid suppression in southeastern provinces centered on the Daegu superspreading event; and (3) brief, self-contained outbreaks in isolated regions such as Sejong. Regions with comparable Rt trajectories sometimes displayed markedly different curve shapes, indicating that neither Rt nor curve morphology alone fully captures spatial transmission dynamics. In metropolitan areas, the mobility-adjusted Rt rose earlier than visible case surges, suggesting its utility as an early indicator of latent transmission pressure. Conclusions. By coupling structural shape analysis with mobility-aware Rt estimation, our integrative framework provides a richer, more actionable understanding of subnational epidemic dynamics. This dual perspective enables earlier detection of transmission risks and supports geographically targeted interventions. Our methodology is broadly applicable to emerging diseases where both spatial connectivity and temporal dynamics drive transmission.