Elsevier

The Lancet

Volume 369, Issue 9571, 28 April–4 May 2007, Pages 1452-1459
The Lancet

Articles
Burden of symptomatic dengue infection in children at primary school in Thailand: a prospective study

https://doi.org/10.1016/S0140-6736(07)60671-0Get rights and content

Summary

Background

Dengue viruses are a major cause of morbidity and mortality in tropical and subtropical areas. Our aim was to assess prospectively the burden of dengue-related illness in children in Thailand.

Methods

We did a prospective study in a cohort of children at primary school in northern Thailand from 1998 to 2002. We assessed the burden of dengue illness as disability-adjusted life years (DALYs) and patient costs per illness.

Findings

Dengue accounted for 328 (11%) of the 3056 febrile cases identified in 2114 children during the study period. The mean burden of dengue was 465·3 (SD 358·0; range 76·5–954·0) DALYs per million population per year, accounting for about 15% of DALYs lost to all febrile illnesses (3213·1 [SD 2624·2] DALYs per million per year). Non-hospitalised patients with dengue illnesses represented a substantial proportion of the overall burden of disease, with 44–73% of the total DALYs lost to dengue each year due to such illness. The infecting dengue serotype was an important determinant of DALYs lost: DEN4 was responsible for 1% of total DALYs lost, DEN1 for 9%, DEN2 for 30%, and DEN3 for 29%.

Interpretation

Use of prospective data to estimate the burden of disease shows that most DALYs lost to dengue illness were the result of non-hospitalised illnesses of long duration. Thus, inclusion of non-hospitalised cases is critical to accurately assess the total burden of dengue illness.

Introduction

Dengue fever and dengue haemorrhagic fever are important causes of morbidity and mortality in tropical and subtropical regions of the world.1, 2, 3, 4 Since there is no effective treatment for dengue infection, prevention efforts rely on mosquito control and the development of tetravalent dengue vaccines.5, 6, 7 Several population-based studies have assessed the burden of disease due to dengue and found the effect of dengue on both financial and other factors to be substantial.2, 8, 9, 10, 11 These studies mainly used surveillance data derived from hospitalised cases, reported cases of dengue haemorrhagic fever, and deaths attributed to dengue; data on non-hospitalised cases and less severe dengue disease were not available. Therefore, these calculations probably underestimate the true burden of disease, given that classic dengue fever represents about 90% of symptomatic dengue cases and most of these are non-hospitalised cases.12

Use of disability-adjusted life years (DALYs) facilitates the assessment of the effect of illness and premature death on an individual and societal level in a manner that does not necessitate financial valuations of life and health.13, 14 Previous endeavours to describe DALYs lost to dengue have been variable with respect to source population, case ascertainment, parameter estimates, and geographical region. Nonetheless, these studies have concluded that dengue is associated with a sizeable burden of illness, relative to other diseases.

Our aim was to use prospectively collected data to assess the burden of dengue illness in children at primary school in Thailand. Such an approach captures and includes data for all febrile illnesses, thus permitting the comparison of dengue and non-dengue cases as well as hospitalised and non-hospitalised cases; these data should therefore represent the full burden of dengue illness in this cohort.

Section snippets

Participants

The study methods and determination of acute, symptomatic dengue virus infections in the cohort have been described previously.12, 15 Briefly, the study was done in Kamphaeng Phet province in northern Thailand. In January, 1998, 2214 children were recruited from grades one to five (children aged 5 years to 15 years) at 12 local primary schools. Enrolment criteria for study participation were attendance at a study school, enrolment in first to sixth grades thereafter, and parental written

Results

Between 1998 and 2002, 3056 cases of febrile illnesses were seen in 2119 children, of which 328 (11%) were confirmed dengue infections (table 1). Of these dengue infections, 52 cases progressed to dengue haemorrhagic fever. There were no deaths attributable to dengue infection during this period. The incidence of dengue infection in the cohort ranged from 0·9% in 2000 to 8·4% in 2001. The number of dengue and non-dengue febrile illnesses varied greatly by year, with a notable peak in non-dengue

Discussion

Our estimates of the burden of dengue illness in an endemic area over 5 years confirm previous reports that suggest that dengue virus constitutes a substantial burden to the health and finances of populations living in affected areas. Most DALYs lost to dengue illness were the result of illnesses of long duration that did not require hospitalisation, suggesting that previous estimates were conservative because they excluded non-hospitalised dengue cases.

Previous studies have used surveillance

References (25)

  • DJ Gubler et al.

    Emergence of epidemic dengue/dengue hemorrhagic fever as a public health problem in the Americas

    Infect Agents Dis

    (1993)
  • MG Guzman et al.

    Enhanced severity of secondary dengue-2 infections: death rates in 1981 and 1997 Cuban outbreaks

    Rev Panam Salud Publica

    (2002)
  • Cited by (147)

    • Flaviviruses including Zika virus

      2023, Molecular Medical Microbiology, Third Edition
    View all citing articles on Scopus
    View full text