Sampling intensity and African swine fever detection in South Korean wild boars: a nationwide longitudinal surveillance analysis, 2019–2025

Introduction African swine fever (ASF) has persisted in South Korean wild boar populations since 2019 despite ongoing control efforts. Because reported cases may reflect differences in surveillance effort, this study examined nationwide surveillance data from October 2019 to July 2025 to assess temporal and spatial patterns while accounting for surveillance volume. Methods Monthly county-level surveillance data were analyzed using descriptive statistics and generalized linear mixed-effects models to assess ASF detection, positivity, and surveillance volume. Sensitivity analyses addressed temporal dependence, and a composite surveillance-prioritization framework was evaluated. Results Among 300,887 wild boar samples tested, 3,530 were ASF-positive, corresponding to an overall positivity of 1.17%. Annual positivity declined from 5.77% in 2020 and 4.83% in 2021 to 0.64% in 2024 and 0.15% during January–July 2025, despite a substantial increase in surveillance volume. Overall, 72.1% of ASF-positive detections occurred during winter and spring. In the primary models, greater surveillance volume was associated with a higher probability of detecting at least one ASF-positive wild boar within a sampled county-month (OR = 2.78, 95% CI: 2.32–3.35; p < 0.001) but was not associated with all-pathway surveillance positivity (OR = 0.94, 95% CI: 0.81–1.10; p = 0.454). The positive-detection rate was also markedly lower in 2025 than in 2020 (IRR = 0.06, 95% CI: 0.03–0.11; p < 0.001). Sensitivity analyses showed that some surveillance-volume estimates were sensitive to explicit modeling of within-county temporal dependence, while complementary subsequent-detection models provided mixed evidence across lag periods. A county-level surveillance-prioritization framework identified areas where additional surveillance may be informative, although temporal back-testing showed that the composite score did not consistently outperform simpler recent-detection indicators. Discussion The observed ASF surveillance signal declined over time, although seasonal and geographic heterogeneity persisted. Accounting for surveillance volume, recent detection history, and temporal dependence may improve surveillance interpretation and support targeted resource allocation.