A region-structured model for early warning of Plasmodium vivax malaria transmission risk in the Republic of Korea

Abstract

Background

Malaria cases in the Republic of Korea decreased during the coronavirus disease 2019 pandemic but surged in 2023. Current models inadequately address spatial heterogeneity in transmission dynamics. This study aimed to address this by designing a region-structured model considering spatial heterogeneity based on regional malaria data from high-risk areas.

Methods

Malaria-risk areas were identified using data from the Korea Disease Control and Prevention Agency (KDCA), with eight regions designated as hotspots. The population heterogeneity of the model by region was represented using the “Who Acquires Infection From Whom” matrix. The model was calibrated using 2014–2018 KDCA civilian malaria-case data. The reproduction number ($R_t$) of each region was then calculated using the estimated parameters and predicted malaria dynamics.

Results

In the hotspots, the value of $R_t$ rose along with the number of long-latency patients, followed by an increase in short-latency patients. The points where $R_t$ exceeded and fell below one varied by region. Ganghwa-gun exhibited the longest period ($R_t>1$), whereas Deokyang-gu had the shortest. Maximum $R_t$ values ranged from 1.1 in Deokyang-gu to 2.7 in Ganghwa-gun. A criterion was established to estimate the timing of $R_t>1$ based on the weekly cumulative incidence per 100,000 people.

Conclusion

This study constructed a region-structured model reflecting spatial heterogeneity using actual data. By estimating $R_t$ and an easily accessible index for each region, the model provides an indicator that assists in implementing effective malaria management policies at the regional level.