Dengue in northeastern Brazil: a spatial and temporal perspective.

The state of Ceará (Brazilian Northeast) has a high incidence of dengue. Therefore, we aimed to characterize the temporal patterns and spatial distribution of dengue cases in Ceará during 2001-2019. A spatiotemporal ecological study was performed with secondary data. Time-trend analysis was performed using a segmented log-linear regression model to estimate the average annual percentage change (AAPC) and the annual percentage change (APC) in incidence of dengue. We also performed spatiotemporal analysis to identify the place, time, and relative risk (RR) of dengue clusters. There were 539,653 dengue cases. The AAPC reduced over time (-9.5%; 95% confidance interval [CI]: -18.3; -0.3). Three trends were identified-2001-2004: APC=-20.9% (95% CI: -65.1 to 44.8), 2005-2015: APC=7.9% (95% CI: -6.0 to 98.9), and 2016-2019: APC=-48.8% (95% CI: -83.0 to -6.1). During 2001-2007, 10 significant clusters were identified (RR=3.57-14.38: n=4 and RR=0.05-0.39: n=6). During 2008-2013, there was 1 cluster in the western region (RR= 3.40) and four other clusters (RR=0.02-0.15). The last period presented 5 high-RR clusters (RR=2.95-9.24). The low-RR clusters were located in the central-north, central-south, south, and northwest regions. However, the central-west region remained a high-RR cluster region throughout the study period. Dengue showed a decreasing incidence. During the epidemic years, the southern, eastern, and western regions presented high-risk clusters. Introduction of a new dengue serotype in a low-RR area can cause explosive outbreaks due to population susceptibility.