Saita Sayambhu, Maeakhian Sasithan, Silawan Tassanee
Faculty of Public Health, Thammasat University, Lampang 25190, Thailand.
Thammasat University Research Unit in One Health and Ecohealth, Thammasat University, Pathum Thani 12120, Thailand.
Trop Med Infect Dis. 2022 Aug 8;7(8):171. doi: 10.3390/tropicalmed7080171.
The efforts towards effective control of the COVID-19 pandemic may affect the incidence of dengue. This study aimed to investigate temporal variations and spatial clusters of dengue in Thailand before and during the COVID-19 pandemic. Reported dengue cases before (2011-2019) and during (2020-2021) the COVID-19 pandemic were obtained from the national disease surveillance datasets. The temporal variations were analyzed using graphics, a seasonal trend decomposition procedure based on Loess, and Poisson regression. A seasonal ARIMA model was used to forecast dengue cases. Spatial clusters were investigated using the local indicators of spatial associations (LISA). The cyclic pattern showed that the greatest peak of dengue cases likely changed from every other year to every two or three years. In terms of seasonality, a notable peak was observed in June before the pandemic, which was delayed by one month (July) during the pandemic. The trend for 2011-2021 was relatively stable but dengue incidence decreased dramatically by 7.05% and 157.80% on average in 2020 and 2021, respectively. The forecasted cases in 2020 were slightly lower than the reported cases (2.63% difference), whereas the forecasted cases in 2021 were much higher than the actual cases (163.19% difference). The LISA map indicated 5 to 13 risk areas or hotspots of dengue before the COVID-19 pandemic compared to only 1 risk area during the pandemic. During the COVID-19 pandemic, dengue incidence sharply decreased and was lower than forecasted, and the spatial clusters were much lower than before the pandemic.
为有效控制新冠疫情所做的努力可能会影响登革热的发病率。本研究旨在调查新冠疫情之前及期间泰国登革热的时间变化和空间聚集情况。从国家疾病监测数据集获取了新冠疫情之前（2011 - 2019年）和期间（2020 - 2021年）报告的登革热病例。使用图表、基于局部加权回归散点平滑法（Loess）的季节性趋势分解程序以及泊松回归分析时间变化。采用季节性自回归整合移动平均模型（ARIMA）预测登革热病例。使用空间自相关局部指标（LISA）研究空间聚集情况。周期性模式表明，登革热病例的最大峰值可能从隔年出现变为每两到三年出现一次。在季节性方面，疫情之前6月出现显著峰值，疫情期间该峰值推迟了一个月（7月）出现。2011 - 2021年的趋势相对稳定，但2020年和2021年登革热发病率分别平均大幅下降了7.05%和157.80%。2020年预测病例略低于报告病例（相差2.63%），而2021年预测病例远高于实际病例（相差163.19%）。LISA地图显示，新冠疫情之前有5至13个登革热风险区域或热点，而疫情期间只有1个风险区域。在新冠疫情期间，登革热发病率急剧下降且低于预测值，空间聚集情况也远低于疫情之前。
