Dengue outbreaks in high-income area, Kaohsiung City, Taiwan, 2003-2009

Abstract

Kaohsiung City, a modern metropolis of 1.5 million persons, has been the focus of dengue virus activity in Taiwan for several decades. The aim of this study was to provide a temporal and spatial description of dengue virus epidemiology in Kaohsiung City by using data for all laboratory-confirmed dengue cases during 2003-2009. We investigated age- and sex-dependent incidence rates and the spatiotemporal patterns of all cases confirmed through passive or active surveillance. Elderly persons were at particularly high risk for dengue virus-related sickness and death. Of all confirmed cases, ≈75% were detected through passive surveillance activities; case-patients detected through active surveillance included immediate family members, neighbors, and colleagues of confirmed case-patients. Changing patterns of case clustering could be due to the effect of unmeasured environmental and demographic factors.

Figure 1

Kaohsiung City, Taiwan (22°38′N, 120°16′E), indicating the 11 districts and 463 administrative units (Li) of the city and the main entry points for international travel and commerce. Insets show location of Taiwan in Southeast Asia (box) and of Kaohsiung City in Taiwan (gray shading).

Figure 2

Monthly average temperature, rainfall, and adult index (AI) for Aedes aegypti and Ae. albopictus mosquitoes, Kaohsiung City, Taiwan, 2003–2009. AI was calculated as number of adult female mosquitoes captured per number of inspected premises.

Figure 3

Epidemic curve of confirmed cases of dengue virus (DENV) infection (N = 2,087), by week of onset, Kaohsiung City, Taiwan, 2003–2009. Predominant serotypes (DENV-1–3) and numbers of confirmed cases are shown.

Figure 4

Age-specific distribution of case-patients with confirmed dengue virus infection (N = 2,087) detected by passive and active surveillance systems, Kaohsiung City, Taiwan, 2003–2009. Cases detected through passive surveillance were suspected dengue virus infections reported by health care facilities to Taiwan Center for Disease Control; cases detected through active surveillance were reported from airport screenings and by community contacts of case-patients.

Figure 5

Local indicator of spatial association (LISA) cluster maps of incidence rates for dengue virus infection during each epidemic period, Kaohsiung City, Taiwan, 2003–2009. High-value outlier, high-incidence Li (smallest administrative unit within each of 11 districts in Kaohsiung City) surrounded by low-incidence Lis; not significant, 0 spatial autocorrelation presented; Hot spot, high-incidence Li surrounded by high-incidence Lis. Hot-spot Lis circled with dashed lines are those that overlap with clusters of high residential density; hot-spot or high-value outlier Lis circled with solid lines are those that overlap with clusters of low residential density.