Dengue Literature Review

Taiwan

5 min read

Taiwan

Overview

Taiwan is a non-endemic but periodically hyperendemic dengue setting. Dengue outbreaks occur primarily in southern Taiwan — Tainan, Kaohsiung, and Pingtung — driven by Aedes aegypti populations in urban areas. Taiwan’s National Health Insurance Research Database (NHIRD), covering >99% of the population, has made it a valuable setting for large-scale epidemiological cohort studies on dengue outcomes.

Key Points from Literature

Epidemic History and Scale

  • Major outbreaks in Taiwan: 1981, 1987–88, 2001–02, and 2007; DENV-2 predominated in 2002; primary DENV-1 or DENV-3 in 2004–2007 (see Wan2012 - Autoimmunity in Dengue Pathogenesis)
  • Between 2002 and 2015, 63,814 laboratory-confirmed dengue cases were recorded in the NDDCC (Notifiable Disease Dataset of Confirmed Cases), concentrated in southern Taiwan (see Shih2023 - Autoimmune Disease Risk After Dengue).
  • The 2014–2015 dengue epidemics were the most severe in Taiwan’s history, causing over 58,000 confirmed cases. These epidemics greatly expanded the available case count for epidemiological study.
  • Dengue has not been endemic in Taiwan; overall incidence and seroprevalence remain low outside epidemic periods, meaning misclassification of controls as cases is a smaller concern than in hyperendemic countries.
  • Seasonal pattern: outbreaks typically start by importation from abroad in early summer, spread locally, and end in winter (see Wan2012 - Autoimmunity in Dengue Pathogenesis)
  • Dengue is primarily an adult disease in Taiwan: DF peaks in the 50–54 age range; DHF peaks in the 60–64 age range — in contrast to hyperendemic Southeast Asia where dengue is predominantly a childhood disease (see Wan2012 - Autoimmunity in Dengue Pathogenesis)

Laboratory Confirmation and Surveillance System

  • Suspected dengue cases must by law be reported to Taiwan CDC within 24 hours of clinical diagnosis; blood samples are submitted to approved laboratories for confirmation.
  • Before 2015, laboratory confirmation required virus isolation, RT-PCR, a 4-fold increase in paired IgG titres, or single-sample dengue-specific IgM/IgG — tests unavailable in most hospitals, creating confirmation delays.
  • From 2015 onwards, NS1 rapid antigen tests became widely available, substantially shortening confirmation time and improving diagnostic accuracy (see NS1 Antigen Detection).
  • A critical epidemiological finding: only 51.4% of patients hospitalised with a dengue discharge diagnosis between 2000–2010 were ultimately lab-confirmed in the NDDCC — implying that pre-2015 hospital-based dengue case series substantially overestimate true dengue counts (see Shih2023 - Autoimmune Disease Risk After Dengue).

NHIRD as a Research Resource

  • The NHIRD provides longitudinal insurance claims data for >99% of the Taiwanese population since 1995; encrypted individual IDs allow linkage to multiple national databases including the NDDCC and the Registry for Catastrophic Illness.
  • The Registry for Catastrophic Illness captures validated autoimmune disease diagnoses reviewed by specialist physicians, providing high-quality outcome ascertainment for cohort studies.
  • The combination of NHIRD + NDDCC has enabled population-based cohort studies with over 300,000 participants and mean follow-up exceeding 4 years (see Shih2023 - Autoimmune Disease Risk After Dengue).

Li2018: NHIRD Cohort Study of Autoimmune Disease Risk (2000–2010)

Li2018 - Increased Risk of Autoimmune Diseases in Dengue used the same NHIRD to examine autoimmune disease risk in dengue patients. Cohort: 12,506 hospitalised dengue patients (ICD-9 coded, 2000–2010) matched to 112,554 controls; 3-year follow-up. This study used the pre-2015 ICD-coded dengue case definition — the era before NS1 rapid antigen tests, when only 51.4% of hospital dengue discharges were ultimately lab-confirmed (established by Shih2023).

Key results:

  • Overall aHR 1.88 (95% CI 1.49–2.37) for any autoimmune disease — subsequently refuted by Shih2023 using lab-confirmed cases
  • Primary adrenocortical insufficiency was the most frequent autoimmune outcome (n=19; aHR 2.05)
  • ADEM aHR 3.80 — consistent with Shih2023’s aHR 2.72 for the same disease (the one cross-study convergence)
  • GBS non-significant (aHR 0.97) — also non-significant in Shih2023

The Li2018 and Shih2023 studies together demonstrate that Taiwan’s NHIRD, combined with the NDDCC (Notifiable Disease Dataset of Confirmed Cases), is a uniquely powerful research resource — but that the quality of the dengue case definition (ICD-coded vs. lab-confirmed) decisively determines the validity of the resulting risk estimates.

Contradictions & Debates

  • The Li2018 vs. Shih2023 methodological contrast illustrates a core tension in NHIRD-based dengue research: ICD-coded dengue over-diagnoses (51.4% confirmation rate pre-2015) inflates apparent autoimmune risk, while lab-confirmed dengue provides much narrower but more reliable estimates. Any NHIRD study using dengue diagnoses from before 2015 should be evaluated with this misclassification rate in mind.
  • Taiwan’s non-endemic status means dengue patterns here (epidemic, concentrated in urban south, predominantly introduced) may differ from hyperendemic countries in Southeast Asia and Latin America. Findings on autoimmune sequelae from Taiwan may not generalise to settings where dengue is endemic and most adults have prior exposure.

Sources

Graph View

Backlinks