Spatiotemporal Dynamics of Dengue Epidemics, Southern Vietnam

Hoang Quoc Cuong, Nguyen Thanh Vu, Bernard Cazelles, Maciej F. Boni, Khoa T.D. Thai, Maia A. Rabaa, Luong Chan Quang, Cameron P. Simmons, Tran Ngoc Huu, and Katherine L. Anders

Author affiliations: Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam (H.Q. Cuong, M.F. Boni, K.T.D. Thai, C.P. Simmons, K.L. Anders); Pasteur Institute, Ho Chi Minh City (H.Q. Cuong, N.T. Vu, L.C. Quang, T.N. Huu); University of Oxford, Oxford, UK (M.F. Boni, C.P. Simmons, K.L. Anders); Unités Mixtes de Recherche, Paris, France (B. Cazelles); L'Unité Mixte Internationale, Bondy, France (B. Cazelles); University Medical Centre, Rotterdam, the Netherlands (K.T.D. Thai); Pennsylvania State University, University Park, Pennsylvania, USA (M.A. Rabaa); Monash University, Melbourne, Victoria, Australia (K.L. Anders)

Abstract

An improved understanding of heterogeneities in dengue virus transmission might provide insights into biological and ecologic drivers and facilitate predictions of the magnitude, timing, and location of future dengue epidemics. To investigate dengue dynamics in urban Ho Chi Minh City and neighboring rural provinces in Vietnam, we analyzed a 10-year monthly time series of dengue surveillance data from southern Vietnam. The per capita incidence of dengue was lower in Ho Chi Minh City than in most rural provinces; annual epidemics occurred 1–3 months later in Ho Chi Minh City than elsewhere. The timing and the magnitude of annual epidemics were significantly more correlated in nearby districts than in remote districts, suggesting that local biological and ecologic drivers operate at a scale of 50–100 km. Dengue incidence during the dry season accounted for 63% of variability in epidemic magnitude. These findings can aid the targeting of vector-control interventions and the planning for dengue vaccine implementation.

Dengue is a growing international public health problem for which a licensed vaccine, therapeutic drugs, and effective vector control programs are lacking. The increasing number of cases is associated with an expanding geographic range and increasing intensity of transmission in affected areas (1,2). The dynamics of dengue in disease-endemic areas are characterized by strong seasonality and multiannual epidemic peaks (3), with substantial interannual and spatial heterogeneity in the magnitude of seasonal epidemics (4). Extrinsic factors, including climatic and environmental variables, have been hypothesized to drive annual seasonality; intrinsic factors associated with human host demographics, population immunity, and the virus, drive the multiannual dynamics (5–7). Analyses from Southeast Asia have demonstrated multiannual oscillations in dengue incidence (8–10), which have been variably associated with macroclimatic weather cycles (exemplified by the El Niño Southern Oscillation) in different settings and with changes in population demographics in Thailand (11). In Thailand, a spatiotemporal analysis showed that the multiannual cycle emanated from Bangkok out to more distant provinces (9).
Knowledge of spatial and temporal patterns in dengue incidence at a subnational level is relevant for 2 main reasons: it can provide insights into the biological and ecologic mechanisms that drive transmission, and it might facilitate predictions of the magnitude, timing, and location of future dengue epidemics. For both of these reasons, detailed spatial resolution is useful because aggregated datasets can obscure some of the factors that influence the timing and size of individual local epidemics.
In southern Vietnam, dengue occurs year-round; a marked seasonal peak occurs during the rainy months of June–December, and the number of cases has been increasing over the past 15 years (12). As in many dengue-endemic settings, the dengue surveillance system in Vietnam relies on passive reporting of clinically diagnosed dengue in hospitalized patients. Vector control is the primary tool available for dengue prevention and control. In Vietnam, vector control is pursued through a targeted approach of low-volume space spraying of households around clusters of reported dengue cases. This strategy faces limitations in timeliness and sensitivity because of the reliance on and response to case reports for hospitalized patients only. A predictive epidemiologic tool that enables prioritization of limited resources for the most cost-effective reduction in cases would be highly valued in dengue-endemic settings.
To investigate spatial and temporal trends for dengue in southern Vietnam, we used a monthly time series of dengue surveillance data over 10 years, disaggregated to the district level. We analyzed the periodicity of dengue incidence, determined whether annual epidemics consistently originate in and spread from Ho Chi Minh City (HCMC) or another location, and characterized the differences in the magnitude and timing of epidemics among provinces and districts.