ABSTRACT
A dynamic school-based cohort of 2-15 year-olds was established in Long Xuyen, Viet Nam to provide epidemiological data for a dengue vaccine efficacy trial. Active surveillance of febrile episodes identified clinically-suspected dengue and acute and convalescent sera were collected. IgG seroconversion between annual seroprevalence surveys identified sub-clinical infections. In 2004, 2190 children were enrolled with 3239, 3146, and 3081 present each year from 2005 to 2007 consecutively. In all, 627 children had a total of 690 clinically-suspected dengue episodes (394 hospitalisations, 296 outpatients) with 284-310 (41.2-45.0%) laboratory-confirmed depending on testing. Dengue serotype 2 was predominant in 2004 and 2005, and serotype 1 in 2006 and 2007. The acute dengue disease incidence rate per 1000 person-years ranged from 16.9 in 2005 to 40.4 in 2007. The average annual incidence of primary dengue infection (IgG seroconversion in previously naïve children) was 11.4% and the symptomatic to asymptomatic primary infection ratio ranged from 1:3-1:6. Study withdrawal rate, a feasibility indicator for conducting efficacy trials, was low: 4.2% per year when excluding children who changed schools. Our 2004-2007 results confirm the high transmission of dengue in children in Long Xuyen and demonstrate the suitability of this study site for a large scale efficacy trial.