• View in gallery

    Time series of monthly dengue cases. (AF) DENV-1, DENV-2, DENV-3, DENV-4, serotype unavailable and all cases. DENV = dengue virus.

  • View in gallery

    Ratio of Queen Sirikit National Institute of Child Health (QSNICH) and Thailand Ministry of Public Health (MOPH) dengue cases.

  • View in gallery

    Mean age of primary and post-primary cases by serotype, and for untyped and all cases. Primary mean age for each group is shown in dark purple, red, light green, dark blue, orange, and black, respectively. Secondary mean age for each group is shown in light purple, light pink, dark green, light blue, brown, and gray, respectively. D1 = DENV-1; D2 = DENV-2; D3 = DENV-3; D4 = DENV-4.

  • View in gallery

    Percentage of dengue cases those were post-primary over time by serotype. DENV-1 to DENV-4, serotype unavailable and all cases.

  • View in gallery

    Catalytic model output of post-primary (second infections) proportion under a decrease in force of infection (FOI), known birth rates in Bangkok, and three times as many post-primary than primary infections being observed. Black dots reflect the model without Japanese encephalitis (JE) vaccination and blue dots reflect the model with JE vaccination. Estimates are shown from 1988 onward since estimates for FOI for the population ≤ 15 years old were possible for their whole lifespan.

  • View in gallery

    Proportion of primary and post-primary dengue cases that were dengue hemorrhagic fever (DHF) or dengue shock syndrome (DSS) over time.

  • View in gallery

    Proportion of dengue cases that were dengue hemorrhagic fever (DHF) or dengue shock syndrome (DSS) for primary and post-primary cases in different age groups.

  • 1.

    Gubler DJ, 2006. Dengue/dengue haemorrhagic fever: history and current status. Novartis Found Symp 277: 316; discussion 16–22, 71– 73, 251–253.

    • Search Google Scholar
    • Export Citation
  • 2.

    Nisalak A, Endy TP, Nimmannitya S, Kalayanarooj S, Thisayakorn U, Scott RM, Burke DS, Hoke CH, Innis BL, Vaughn DW, 2003. Serotype-specific dengue virus circulation and dengue disease in Bangkok, Thailand from 1973 to 1999. Am J Trop Med Hyg 68: 191202.

    • Search Google Scholar
    • Export Citation
  • 3.

    Guzman MG, Vazquez S, 2010. The complexity of antibody-dependent enhancement of dengue virus infection. Viruses 2: 26492662.

  • 4.

    Guzman MG, Kouri G, 2003. Dengue and dengue hemorrhagic fever in the Americas: lessons and challenges. J Clin Virol 27: 113.

  • 5.

    Bureau of Epidemiology MoPH, Thailand, 2015. Dengue Surveillance Reports. Available at: www.boe.moph.go.th/. Accessed April 17, 2015.

  • 6.

    Limkittikul K, Brett J, L'Azou M, 2014. Epidemiological trends of dengue disease in Thailand (2000–2011): a systematic literature review. PLoS Negl Trop Dis 8: e3241.

    • Search Google Scholar
    • Export Citation
  • 7.

    Yuill TM, Sukhavachana P, Nisalak A, Russell PK, 1968. Dengue-virus recovery by direct and delayed plaques in LLC-MK2 cells. Am J Trop Med Hyg 17: 441448.

    • Search Google Scholar
    • Export Citation
  • 8.

    Kuberski TT, Rosen L, 1977. A simple technique for the detection of dengue antigen in mosquitoes by immunofluorescence. Am J Trop Med Hyg 26: 533537.

    • Search Google Scholar
    • Export Citation
  • 9.

    Watts DM, Harrison BA, Nisalak A, Scott RM, Burke DS, 1982. Evaluation of Toxorhynchites splendens (Diptera: Culicidae) as a bioassay host for dengue viruses. J Med Entomol 19: 5459.

    • Search Google Scholar
    • Export Citation
  • 10.

    Klungthong C, Gibbons RV, Thaisomboonsuk B, Nisalak A, Kalayanarooj S, Thirawuth V, Nutkamhang N, Mammen MP, Jarman RG, 2007. Dengue virus detection using whole blood for RT-PCR and viral isolation. J Clin Microbiol 45: 24802485.

    • Search Google Scholar
    • Export Citation
  • 11.

    Jarman RG, Nisalak A, Anderson KB, Klungthong C, Thaisomboonsuk B, Kaneechit W, Kalayanarooj S, Gibbons RV, 2011. Factors influencing dengue virus isolation by C6/36 cell culture and mosquito inoculation of nested PCR-positive clinical samples. Am J Trop Med Hyg 84: 218223.

    • Search Google Scholar
    • Export Citation
  • 12.

    Clarke D, Casals J, 1955. Improved methods for hemagglutination studies with arthropod-borne viruses. Proc Soc Exp Biol Med 88: 9699.

  • 13.

    Russell PK, Nisalak A, Sukhavachana P, Vivona S, 1967. A plaque reduction test for dengue virus neutralization antibodies. J Immunol 99: 285290.

    • Search Google Scholar
    • Export Citation
  • 14.

    Kuno G, Gubler DJ, Santiago de Weil NS, 1985. Antigen capture ELISA for the identification of dengue viruses. J Virol Methods 12: 93103.

  • 15.

    Innis BL, Nisalak A, Nimmannitya S, Kusalerdchariya S, Chongswasdi V, Suntayakorn S, Puttisri P, Hoke CH Jr, 1989. An enzyme-linked immunosorbent assay to characterize dengue infections where dengue and Japanese encephalitis co-circulate. Am J Trop Med Hyg 40: 418427.

    • Search Google Scholar
    • Export Citation
  • 16.

    World Health Organization, 1986. Dengue Hemorrhagic Fever: Diagnosis, Treatment and Control. Geneva, Switzerland: World Health Organization.

    • Search Google Scholar
    • Export Citation
  • 17.

    Cummings DAT, Iamsirithaworn S, Lessler JT, McDermott A, Prasanthong R, Nisalak A, Jarman RG, Burke DS, Gibbons RV, 2009. The impact of the demographic transition on dengue in Thailand: insights from a statistical analysis and mathematical modeling. PLoS Med 6: e1000139.

    • Search Google Scholar
    • Export Citation
  • 18.

    Rodriguez-Barraquer I, Buathong R, Iamsirithaworn S, Nisalak A, Lessler J, Jarman RG, Gibbons RV, Cummings DA, 2014. Revisiting Rayong: shifting seroprofiles of dengue in Thailand and their implications for transmission and control. Am J Epidemiol 179: 353360.

    • Search Google Scholar
    • Export Citation
  • 19.

    Goh KT, 1997. Dengue—a re-emerging infectious disease in Singapore. Ann Acad Med Singapore 26: 664670.

  • 20.

    Guzman MG, Kouri G, Bravo J, Valdes L, Vazquez S, Halstead SB, 2002. Effect of age on outcome of secondary dengue 2 infections. Int J Infect Dis 6: 118124.

    • Search Google Scholar
    • Export Citation
  • 21.

    Kongsomboon K, Singhasivanon P, Kaewkungwal J, Nimmannitya S, Mammen MP Jr, Nisalak A, Sawanpanyalert P, 2004. Temporal trends of dengue fever/dengue hemorrhagic fever in Bangkok, Thailand from 1981 to 2000: an age-period-cohort analysis. Southeast Asian J Trop Med Public Health 35: 913917.

    • Search Google Scholar
    • Export Citation
  • 22.

    Wilder-Smith A, Yoksan S, Earnest A, Subramaniam R, Paton NI, 2005. Serological evidence for the co-circulation of multiple dengue virus serotypes in Singapore. Epidemiol Infect 133: 667671.

    • Search Google Scholar
    • Export Citation
  • 23.

    Thai KT, Nishiura H, Hoang PL, Tran NT, Phan GT, Le HQ, Tran BQ, Nguyen NV, de Vries PJ, 2011. Age-specificity of clinical dengue during primary and secondary infections. PLoS Negl Trop Dis 5: e1180.

    • Search Google Scholar
    • Export Citation
  • 24.

    Adams B, Holmes EC, Zhang C, Mammen MP Jr, Nimmannitya S, Kalayanarooj S, Boots M, 2006. Cross-protective immunity can account for the alternating epidemic pattern of dengue virus serotypes circulating in Bangkok. Proc Natl Acad Sci USA 103: 1423414239.

    • Search Google Scholar
    • Export Citation
  • 25.

    Fried JR, Gibbons RV, Kalayanarooj S, Thomas SJ, Srikiatkhachorn A, Yoon IK, Jarman RG, Green S, Rothman AL, Cummings DA, 2010. Serotype-specific differences in the risk of dengue hemorrhagic fever: an analysis of data collected in Bangkok, Thailand from 1994 to 2006. PLoS Negl Trop Dis 4: e617.

    • Search Google Scholar
    • Export Citation
 
 
 

 

 
 
 

 

 

 

 

 

 

Forty Years of Dengue Surveillance at a Tertiary Pediatric Hospital in Bangkok, Thailand, 1973–2012

View More View Less
  • Department of Virology, Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand; Department of Epidemiology, Johns Hopkins School of Public Health, Baltimore, Maryland; Queen Sirikit National Institute of Child Health, Ministry of Public Health, Bangkok, Thailand; Division of Infectious Diseases and Immunology, Department of Medicine, University of Massachusetts Medical School, Worcester, Massachusetts

Long-term observational studies can provide valuable insights into overall dengue epidemiology. Here, we present analysis of dengue cases at a pediatric hospital in Bangkok, Thailand, during a 40-year period from 1973 to 2012. Data were analyzed from 25,715 hospitalized patients with laboratory-confirmed dengue virus (DENV) infection. Several long-term trends in dengue disease were identified including an increase in mean age of hospitalized cases from an average of 7–8 years, an increase after 1990 in the proportion of post-primary cases for DENV-1 and DENV-3, and a decrease in the proportion of dengue hemorrhagic fever and dengue shock syndrome cases in primary and post-primary cases over time. Exploratory mechanistic analysis of these observed trends considered changes in diagnostic methods, demography, force of infection, and Japanese encephalitis vaccination as possible explanations. Thailand is an important setting for studying DENV transmission as it has a “mature” dengue epidemiology with a strong surveillance system in place since the early 1970s. We characterized changes in dengue epidemiology over four decades, and possible impact of demographic and other changes in the human population. These results may inform other countries where similar changes in transmission and population demographics may now or may soon be occurring.

Introduction

Dengue virus (DENV) is the causative agent of dengue disease. DENV has four antigenically distinct serotypes (DENV-1, DENV-2, DENV-3, and DENV-4). DENV infection results in a broad spectrum of clinical presentations from asymptomatic infection to undifferentiated febrile illness, dengue fever (DF), and dengue hemorrhagic fever (DHF) including dengue shock syndrome (DSS).1 Severe clinical manifestations are more likely to occur with repeated infections with a different serotype from the initial infecting serotype.24 There is no specific treatment, although with proper diagnosis and management, fatality rates in Thailand are less than 1%.5,6 Thailand has had sustained transmission of all four DENV serotypes since 1958.2

The epidemiological and clinical manifestations of DENV infections are the result of complex interactions among virologic, immunologic, and ecologic factors. Many questions remain about how these factors interact to result in dengue disease patterns. Small but important differences in these contributing factors in different environments and years make it difficult to draw general conclusions from short-term studies. Thus, long-term observational studies can provide invaluable insights into overall dengue behavior.

Since 1962, Queen Sirikit National Institute of Child Health (QSNICH) and the Armed Forces Research Institute of Medical Sciences (AFRIMS) have engaged in collaborative studies on dengue. An earlier report from this collaboration presented findings from 27 years of dengue surveillance from 1973 to 1999, highlighting serotype-specific epidemiological and clinical patterns of disease.2 Here, we analyze an additional 13 years covering 40 continuous years of dengue surveillance from 1973 to 2012. This unique dataset reveals several long-term trends in dengue disease including changes in age distribution of hospitalized cases, relative proportions of primary and post-primary infections, and disease severity.

Methods

Ethics statement.

The retrieval and analysis of coded preexisting data in this study was approved by the Institutional Review Boards of Walter Reed Army Institute of Research and QSNICH. Blood samples from passive surveillance were originally collected at QSNICH for public health purposes. All data were analyzed anonymously.

Description of the dataset.

The analyzed data were from public health samples collected during passive surveillance of hospital-attended dengue cases from 1973 to 2012 at QSNICH, a 420-bed tertiary care pediatric hospital located in Bangkok, Thailand, which serves as a Thailand Ministry of Public Health (MOPH) dengue referral center for Bangkok. Acute and convalescent blood samples from clinically suspected dengue patients at QSNICH were tested for evidence of DENV infection at AFRIMS. Dengue testing was performed using virologic and serologic techniques varying during different time periods, as listed in Table 1. Acute blood samples were tested by viral isolation and/or hemi-nested reverse transcriptase polymerase chain reaction (RT-PCR) as previously described.711 Acute and convalescent blood samples were tested by dengue serological assays as previously described.1215 Primary infection refers to the first DENV infection in an individual and was determined serologically by dengue hemagglutination inhibition assay and/or dengue IgM/IgG capture enzyme-linked immunosorbent assay (ELISA) according to published criteria.15 Post-primary infection refers to any DENV infection subsequent to primary infection and was also determined serologically.15 Clinical classification was based on World Health Organization guidelines applicable at the time of hospitalization.16

Table 1

Dengue diagnostic methods used during study period

MethodologyYear
Viral isolation and identification by direct viral plaque on LLC-MK2 cell line1973 to 1977
Hemagglutination inhibition assay1973 to present
Plaque reduction neutralization test1973 to present
Fluorescent antibody assay1973 to 1987
Immunofluorescent antibody identification of serotypes1973 to 1983
Mosquito inoculation in Toxorhynchites splendens1977 to present
Antigen-capture ELISA for serotype identification1981 to present
IgM/IgG-capture ELISA1987 to present
Hemi-nested serotype-specific PCR assay1995 to present

ELISA = enzyme-linked immunosorbent assay; PCR = polymerase chain reaction.

The QSNICH data were compared with reported dengue cases collected from Bangkok by the MOPH from 1973 to 2012. The MOPH passive sentinel surveillance system collected information on clinically confirmed dengue cases of all ages, of which the majority, but not all, presented to hospitals. Infections among these patients were generally not serotyped, although age and some disease severity information were available. The QSNICH cases will be included in this MOPH total, although all the QSNICH cases have been laboratory confirmed.

Statistical analysis.

Descriptive statistics were used to summarize dengue case data and trends over time. Statistical analyses were performed using SPSS for Windows version 19.0 (SPSS Inc., Chicago, IL) and R version 3.0.2 (R Foundation for Statistical Computing, Vienna, Austria).

To evaluate mechanisms for the changes in transmission trends over time, we used an age-stratified catalytic model of DENV infection to simulate the impact of changes in birth rate and force of infection (FOI) on the primary/post-primary distributions of cases (see Supplemental Information). The model uses Bangkok birth rates and previous estimates of decreasing FOI during this time period,17 assumes two infections, and presumes that post-primary (here second) infections were 1.5 times more likely to result in hospitalization than primary infections (the trends are insensitive to this proportion—more detail in the Supplemental Information). We modeled primary and post-primary infections in this time period.

Results

Data were analyzed from 25,715 patients with laboratory-confirmed DENV infection admitted to QSNICH from 1973 to 2012. Ninety-nine percent of the patients in this dataset were 15 years old or younger.

Case numbers and serotype distributions.

Of 25,715 dengue cases, both primary/post-primary infection status and serotype determination were available for 12,090 (47%) patients. The majority of the 53% of the cases did not have a serotype, with only 2.7% missing primary secondary classification. Serotype detection rates increased with the use of diagnostic PCR (since 1995) from 30% to 70–80%. Figure 1 shows the number of cases of each serotype, untyped cases, and total cases over time. Of the cases with serotype information, 38.5% were DENV-1, 26.9% were DENV-2, 24.6% were DENV-3, and 10% were DENV-4. From 1973 to 1986, DENV-2 was generally the predominant serotype. In 1987, a large DENV-3 epidemic occurred. DENV-1 was predominant for much of the period from 1988 to 2012 except in 1993–1994, 1995–1999, and 2010–2011 when DENV-4, DENV-3, and DENV-2 predominated, respectively. DENV-4 incidence appeared to peak every 8–12 years, although with generally much smaller epidemics compared with the other serotypes.

Figure 1.
Figure 1.

Time series of monthly dengue cases. (AF) DENV-1, DENV-2, DENV-3, DENV-4, serotype unavailable and all cases. DENV = dengue virus.

Citation: The American Society of Tropical Medicine and Hygiene 94, 6; 10.4269/ajtmh.15-0337

There was an increase in the number of cases over the entire study period (P < 0.01), but no significant increase from 1987 onwards (when IgG/IgM ELISA was introduced). Figure 2 shows the ratio (with binomial 95% confidence intervals [CIs]) of QSNICH to MOPH dengue cases for each year. This ratio was around 0.1 in the 1970s, 0.15 in the 1980s, peaked at 0.25–0.3 in the 1990s, then steadily decreased until 2012 when it was approximately 0.06. No relationship was seen between this ratio and the proportion of cases of each serotype in each year.

Figure 2.
Figure 2.

Ratio of Queen Sirikit National Institute of Child Health (QSNICH) and Thailand Ministry of Public Health (MOPH) dengue cases.

Citation: The American Society of Tropical Medicine and Hygiene 94, 6; 10.4269/ajtmh.15-0337

Age distribution of cases.

The mean age of cases by serotype and primary/post-primary status over 40 years is shown in Figure 3. The CIs have become smaller in recent years due to the increase in the proportion of cases with known serotype. There were fluctuations in this mean age over time, particularly in recent years. However, an increase in the mean age of cases occurred for most serotypes and for all serotypes combined for both primary and post-primary infections. The increase in mean age is clear through 2005 for primary cases (linear association of age with year, P values < 0.01) and until the end of the data for post-primary (P value < 0.01). For primary cases, we observe a decrease from 2005 until 2012, with a suggestion of this downturn in the post-primary cases.

Figure 3.
Figure 3.

Mean age of primary and post-primary cases by serotype, and for untyped and all cases. Primary mean age for each group is shown in dark purple, red, light green, dark blue, orange, and black, respectively. Secondary mean age for each group is shown in light purple, light pink, dark green, light blue, brown, and gray, respectively. D1 = DENV-1; D2 = DENV-2; D3 = DENV-3; D4 = DENV-4.

Citation: The American Society of Tropical Medicine and Hygiene 94, 6; 10.4269/ajtmh.15-0337

If the 40-year study period is divided into 10-year intervals, the mean age (with 95% CI) for each successive decade was 5.13 (4.49, 5.73), 4.31 (3.93, 4.61), 5.26 (4.98, 5.47), and 5.95 (5.45, 6.17) years for primary infections; 7.34 (7.19, 7.47), 7.55 (7.46, 7.63), 8.18 (8.09, 8.24), and 8.53 (8.39, 8.59) years for post-primary infections; and 7.10 (6.97, 7.25), 7.09 (7.00, 7.17), 7.62 (7.55, 7.70), and 8.10 (8.00, 8.18) for all infections. For comparison, we looked at the mean age of dengue cases under 15 years of age in the MOPH system for the last three decades (for which MOPH data were available). The mean age (with 95% CI) was 7.04 (7.00, 7.08), 8.04 (8.00, 8.08), and 8.84 (8.80, 8.88), respectively. These ages were similar to the mean ages of the QSNICH post-primary cases for the last three decades, and showed the same increasing trend.

Primary and post-primary infections.

Only 15.2% of dengue cases were primary whereas 82.1% were post-primary (2.7% could not be determined due to lack of appropriate convalescent sample). The distribution of primary and post-primary infections differed by serotype. DENV-1 and DENV-3 were more likely to be primary cases than DENV-2 and DENV-4 (Figure 4). Over the course of the 40-year study period, the proportion of DENV-2 and DENV-4 cases that were post-primary remained consistently high. Interestingly, the proportion of DENV-1 and DENV-3 cases that were post-primary increased over the entire study period and from 1987 onwards (linear associations, all P values < 0.001). Given the smaller proportion of cases that were typed in the early years, this increase was particularly obvious from 1990 onwards, until the mid-2000s. After this time, there was a possible decrease/stabilization in the proportion of post-primary cases (Figure 4).

Figure 4.
Figure 4.

Percentage of dengue cases those were post-primary over time by serotype. DENV-1 to DENV-4, serotype unavailable and all cases.

Citation: The American Society of Tropical Medicine and Hygiene 94, 6; 10.4269/ajtmh.15-0337

Exploring the mean age and increase in post-primary proportion trends.

The catalytic model showed that with only changes in births and a decrease in the FOI as simulated here, it is possible to recreate an increase in the proportion of cases that are secondary from 1990 to 2000 followed by a leveling off/decrease (Figure 5). This trend is consistent with what was actually observed in recent years for DENV-1 and DENV-3 (Figure 4); however, the modeled increase is smaller than that actually observed. With Japanese encephalitis (JE) vaccination included in the model, the increase in the proportion of post-primary cases was predicted to last until 2005, followed by a similar decrease.

Figure 5.
Figure 5.

Catalytic model output of post-primary (second infections) proportion under a decrease in force of infection (FOI), known birth rates in Bangkok, and three times as many post-primary than primary infections being observed. Black dots reflect the model without Japanese encephalitis (JE) vaccination and blue dots reflect the model with JE vaccination. Estimates are shown from 1988 onward since estimates for FOI for the population ≤ 15 years old were possible for their whole lifespan.

Citation: The American Society of Tropical Medicine and Hygiene 94, 6; 10.4269/ajtmh.15-0337

Disease severity over time.

Figure 6 shows the proportion of all primary and post-primary dengue cases classified as DHF or DSS over time. For primary infections, despite large CIs, it appears there was a decrease in this percentage over 40 years from almost all cases being DHF or DSS in 1973 to around 50% of cases in the 2000s. For post-primary infections, the percentage of DHF/DSS cases was also close to 100% initially, remained fairly steady at around 80% throughout the 1980s, 1990s, and early 2000s, and then sharply decreased in 2005 with an increase again in the most recent years. Figure 7 shows the proportion of DHF/DSS for primary and post-primary infections by age. The decreasing linear relationship of disease severity with time is significant for both primary and secondary cases (P values < 0.001). In primary infections, no clear relationship with age was apparent. In post-primary infections, there were also large CIs, but the highest proportion of DHF/DSS cases occurred in the 5–9 year age group, with a nonsignificant suggestion of a slight decrease thereafter.

Figure 6.
Figure 6.

Proportion of primary and post-primary dengue cases that were dengue hemorrhagic fever (DHF) or dengue shock syndrome (DSS) over time.

Citation: The American Society of Tropical Medicine and Hygiene 94, 6; 10.4269/ajtmh.15-0337

Figure 7.
Figure 7.

Proportion of dengue cases that were dengue hemorrhagic fever (DHF) or dengue shock syndrome (DSS) for primary and post-primary cases in different age groups.

Citation: The American Society of Tropical Medicine and Hygiene 94, 6; 10.4269/ajtmh.15-0337

The proportion of cases of each serotype (DENV-1, DENV-2, DENV-3, DENV-4 and untyped) which are DHF/DSS is 0.66 (95% CI: 0.62, 0.69), 0.75 (0.68, 0.75), 0.71 (0.66, 0.75), 0.77 (0.58, 0.90), and 0.77 (0.74, 0.79) for primary cases and 0.69 (0.66, 0.70), 0.80 (0.79, 0.82), 0.73 (0.70, 0.75), 0.74 (0.71, 0.77), and 0.85 (0.84, 0.86) for secondary cases.

Discussion

All four DENV serotypes circulated continuously in Bangkok from 1973 to 2012, causing both DF and DHF in children. Our evaluation of serotype-specific dengue disease patterns over a 40-year period furthers our understanding of the relationships among serotype, severity, and epidemiology of dengue.

Since 1973, the mean age of dengue cases increased from 7.12 to 8.14 years. A similar increase was noted in cases under 15 years of age from the MOPH, and is consistent with an increase seen throughout Thailand.17 The observed increase in mean age is consistent with a decrease in the FOI as has been shown in previous works.17,18 During the time of mean age increase (and implied decrease in FOI), there was also an increase in the proportion of cases that were post-primary for DENV-1 and DENV-3. This increase is counterintuitive to what one would expect in children with a decreasing FOI. One would expect to see a decrease in post-primary proportion, as both primary and post-primary infections would occur at an older age, and therefore, a greater proportion of post-primary than primary cases would be too old for a pediatric hospital. In fact, the opposite was observed. However, we demonstrated that a catalytic model accounting for the differential population level susceptibility to infection over time due to changing FOI and birth rates predicted a period of around 10 years (1990–2000) during which a disproportionate number of people were primed for a post-primary infection compared with the number of people newly entering the population as susceptible to a primary infection. This may have led temporarily to an increase in the post-primary proportion; however, the model does not show an increase as large as that observed, suggesting other processes might be at play, an area that warrants further exploration. In the model, as this population aged and the birth rates stabilized, the effect diminished (around 2000) and the post-primary proportion stabilized as would be expected with a decrease in FOI. The simulation of JE vaccination in the model continued this increase for another 5 years also consistent with observation. The consistency of our model with observed data highlights the importance of understanding all facets of demography and other possible contributing factors when considering trends in case counts. The suggestion of a decrease in mean age in the last few years was not explored here, and will be of interest to follow with data from subsequent years.

An overall decrease in disease severity occurred during the 40-year study period. There are several plausible, nonmutually exclusive explanations for the decreasing severity over time. A difference in disease severity with age is a phenomenon that has been noted previously, for example, in Cuba, Thailand, and Singapore,1923 which could be consistent with the observed trends; however, we do not observe such a clear relationship with age and severity here and observe a fairly small shift in age of cases. Secondly, in 1999, a new handbook was published with recommendations for hospitalization of suspected dengue cases.16 These changes in hospitalization criteria could have led to a greater number of less severe cases being hospitalized, and could therefore account for the observed trends in disease severity. In addition, the decrease in disease severity in post-primary cases could have been the result of some primary cases that are more likely to be less severe, being categorized as post-primary in the era of JE vaccination. Finally, our findings and previous literatures show that DENV-1 (and DENV-3 to some extent) is less associated with severe disease than the other serotypes. The decrease in in severity over time would be consistent with the increase in cases of these serotypes. The factors associated with disease severity deserve continued focus, so we are able to fully understand observed changes, particularly after the introduction of a vaccination.

This study demonstrated serotype cycling over the years in Bangkok as has been well documented previously.2,24,25 DENV-1 and DENV-2 predominated for long periods of time, and conversely, DENV-4 has never reached the same peaks in case numbers as the other serotypes. It is not clear whether differences in the predominant serotypes were due to differences in FOI or to differences in their ability to cause disease upon infection. In recent years, all four serotypes have more consistently circulated at the same time. However, the ability to determine serotype has increased, which may have influenced the observed trend toward more serotypes circulating at one time.

There was a slight increase in overall case numbers over time. However, this becomes nonsignificant when we consider only the time period with consistent serological testing. This suggests that this trend was due to reporting/testing changes rather than a true increase in cases numbers. In the last decade, QSNICH dengue cases have been around 6% of the MOPH cases. Over time, this proportion has fluctuated, with a particularly high proportion during the 1990s. This does not appear to be related to the serotype distributions at each time, but could be due to changes in the surveillance system. This is important to bear in mind when using each dataset to make conclusions about trends in dengue epidemiology.

An updated analysis of trends of dengue in Bangkok is timely for understanding what may happen in the future, with or without the introduction of a dengue vaccine. Thailand is an important setting for studying DENV transmission as it has a “mature” dengue epidemiology with a strong surveillance system in place since the early 1970s. Description and understanding of how this epidemiology has developed over the years and interacted with other changes in the population such as demography is important for other countries where similar changes in transmission and population demographics may now or may soon be occurring.

ACKNOWLEDGMENTS

We wish to acknowledge the clinical staff at QSNICH, and the clinical, laboratory and data management staff at AFRIMS. We thank Siraporn Sawasdivorn, Director, QSNICH, for her support. We also thank three anonymous reviewers for their useful comments on the manuscript.

  • 1.

    Gubler DJ, 2006. Dengue/dengue haemorrhagic fever: history and current status. Novartis Found Symp 277: 316; discussion 16–22, 71– 73, 251–253.

    • Search Google Scholar
    • Export Citation
  • 2.

    Nisalak A, Endy TP, Nimmannitya S, Kalayanarooj S, Thisayakorn U, Scott RM, Burke DS, Hoke CH, Innis BL, Vaughn DW, 2003. Serotype-specific dengue virus circulation and dengue disease in Bangkok, Thailand from 1973 to 1999. Am J Trop Med Hyg 68: 191202.

    • Search Google Scholar
    • Export Citation
  • 3.

    Guzman MG, Vazquez S, 2010. The complexity of antibody-dependent enhancement of dengue virus infection. Viruses 2: 26492662.

  • 4.

    Guzman MG, Kouri G, 2003. Dengue and dengue hemorrhagic fever in the Americas: lessons and challenges. J Clin Virol 27: 113.

  • 5.

    Bureau of Epidemiology MoPH, Thailand, 2015. Dengue Surveillance Reports. Available at: www.boe.moph.go.th/. Accessed April 17, 2015.

  • 6.

    Limkittikul K, Brett J, L'Azou M, 2014. Epidemiological trends of dengue disease in Thailand (2000–2011): a systematic literature review. PLoS Negl Trop Dis 8: e3241.

    • Search Google Scholar
    • Export Citation
  • 7.

    Yuill TM, Sukhavachana P, Nisalak A, Russell PK, 1968. Dengue-virus recovery by direct and delayed plaques in LLC-MK2 cells. Am J Trop Med Hyg 17: 441448.

    • Search Google Scholar
    • Export Citation
  • 8.

    Kuberski TT, Rosen L, 1977. A simple technique for the detection of dengue antigen in mosquitoes by immunofluorescence. Am J Trop Med Hyg 26: 533537.

    • Search Google Scholar
    • Export Citation
  • 9.

    Watts DM, Harrison BA, Nisalak A, Scott RM, Burke DS, 1982. Evaluation of Toxorhynchites splendens (Diptera: Culicidae) as a bioassay host for dengue viruses. J Med Entomol 19: 5459.

    • Search Google Scholar
    • Export Citation
  • 10.

    Klungthong C, Gibbons RV, Thaisomboonsuk B, Nisalak A, Kalayanarooj S, Thirawuth V, Nutkamhang N, Mammen MP, Jarman RG, 2007. Dengue virus detection using whole blood for RT-PCR and viral isolation. J Clin Microbiol 45: 24802485.

    • Search Google Scholar
    • Export Citation
  • 11.

    Jarman RG, Nisalak A, Anderson KB, Klungthong C, Thaisomboonsuk B, Kaneechit W, Kalayanarooj S, Gibbons RV, 2011. Factors influencing dengue virus isolation by C6/36 cell culture and mosquito inoculation of nested PCR-positive clinical samples. Am J Trop Med Hyg 84: 218223.

    • Search Google Scholar
    • Export Citation
  • 12.

    Clarke D, Casals J, 1955. Improved methods for hemagglutination studies with arthropod-borne viruses. Proc Soc Exp Biol Med 88: 9699.

  • 13.

    Russell PK, Nisalak A, Sukhavachana P, Vivona S, 1967. A plaque reduction test for dengue virus neutralization antibodies. J Immunol 99: 285290.

    • Search Google Scholar
    • Export Citation
  • 14.

    Kuno G, Gubler DJ, Santiago de Weil NS, 1985. Antigen capture ELISA for the identification of dengue viruses. J Virol Methods 12: 93103.

  • 15.

    Innis BL, Nisalak A, Nimmannitya S, Kusalerdchariya S, Chongswasdi V, Suntayakorn S, Puttisri P, Hoke CH Jr, 1989. An enzyme-linked immunosorbent assay to characterize dengue infections where dengue and Japanese encephalitis co-circulate. Am J Trop Med Hyg 40: 418427.

    • Search Google Scholar
    • Export Citation
  • 16.

    World Health Organization, 1986. Dengue Hemorrhagic Fever: Diagnosis, Treatment and Control. Geneva, Switzerland: World Health Organization.

    • Search Google Scholar
    • Export Citation
  • 17.

    Cummings DAT, Iamsirithaworn S, Lessler JT, McDermott A, Prasanthong R, Nisalak A, Jarman RG, Burke DS, Gibbons RV, 2009. The impact of the demographic transition on dengue in Thailand: insights from a statistical analysis and mathematical modeling. PLoS Med 6: e1000139.

    • Search Google Scholar
    • Export Citation
  • 18.

    Rodriguez-Barraquer I, Buathong R, Iamsirithaworn S, Nisalak A, Lessler J, Jarman RG, Gibbons RV, Cummings DA, 2014. Revisiting Rayong: shifting seroprofiles of dengue in Thailand and their implications for transmission and control. Am J Epidemiol 179: 353360.

    • Search Google Scholar
    • Export Citation
  • 19.

    Goh KT, 1997. Dengue—a re-emerging infectious disease in Singapore. Ann Acad Med Singapore 26: 664670.

  • 20.

    Guzman MG, Kouri G, Bravo J, Valdes L, Vazquez S, Halstead SB, 2002. Effect of age on outcome of secondary dengue 2 infections. Int J Infect Dis 6: 118124.

    • Search Google Scholar
    • Export Citation
  • 21.

    Kongsomboon K, Singhasivanon P, Kaewkungwal J, Nimmannitya S, Mammen MP Jr, Nisalak A, Sawanpanyalert P, 2004. Temporal trends of dengue fever/dengue hemorrhagic fever in Bangkok, Thailand from 1981 to 2000: an age-period-cohort analysis. Southeast Asian J Trop Med Public Health 35: 913917.

    • Search Google Scholar
    • Export Citation
  • 22.

    Wilder-Smith A, Yoksan S, Earnest A, Subramaniam R, Paton NI, 2005. Serological evidence for the co-circulation of multiple dengue virus serotypes in Singapore. Epidemiol Infect 133: 667671.

    • Search Google Scholar
    • Export Citation
  • 23.

    Thai KT, Nishiura H, Hoang PL, Tran NT, Phan GT, Le HQ, Tran BQ, Nguyen NV, de Vries PJ, 2011. Age-specificity of clinical dengue during primary and secondary infections. PLoS Negl Trop Dis 5: e1180.

    • Search Google Scholar
    • Export Citation
  • 24.

    Adams B, Holmes EC, Zhang C, Mammen MP Jr, Nimmannitya S, Kalayanarooj S, Boots M, 2006. Cross-protective immunity can account for the alternating epidemic pattern of dengue virus serotypes circulating in Bangkok. Proc Natl Acad Sci USA 103: 1423414239.

    • Search Google Scholar
    • Export Citation
  • 25.

    Fried JR, Gibbons RV, Kalayanarooj S, Thomas SJ, Srikiatkhachorn A, Yoon IK, Jarman RG, Green S, Rothman AL, Cummings DA, 2010. Serotype-specific differences in the risk of dengue hemorrhagic fever: an analysis of data collected in Bangkok, Thailand from 1994 to 2006. PLoS Negl Trop Dis 4: e617.

    • Search Google Scholar
    • Export Citation

Author Notes

* Address correspondence to Hannah E. Clapham, Department of Epidemiology, Johns Hopkins School of Public Health, 615 N. Wolfe St., Baltimore, MD 21205. E-mail: Hannah.e.clapham@gmail.com† These authors contributed equally to this work.

Financial support: Derek A. T. Cummings received funding from the National Institute of General Medical Sciences of the National Institutes of Health under award no. U54 GM088491. Hannah E. Clapham, Justin T. Lessler, and Derek A. T. Cummings received funding from the National Institutes of Health under award no. 5R01AI102939-03. Anon Srikiatkhachorn received funding from the National Institute of Allergy and Infectious Diseases under award no. 5 P01 AI034533-22.

Authors' addresses: Ananda Nisalak, Chonticha Klungthong, Butsaya Thaisomboonsuk, Stefan Fernandez, Julia Reiser, Louis R. Macareo, and In-Kyu Yoon, Department of Virology, Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand, E-mails: anandan@afrims.org, chontichak@afrims.org, butsayat.fsn@afrims.org, stefan.fernandez.mil@mail.mil, julia.a.reiser@gmail.com, louis.macareo.mil@afrims.org, and inkyu.yoon.mil@afrims.org. Hannah E. Clapham, Justin T. Lessler, and Derek A. T. Cummings, Department of Epidemiology, Johns Hopkins School of Public Health, Baltimore, MD, E-mails: hannah.e.clapham@gmail.com, jlessler@jhsph.edu, and derek.cummings@jhu.edu. Siripen Kalayanarooj, Queen Sirikit National Institute of Child Health, Ministry of Public Health, Bangkok, Thailand, E-mail: siripenk@gmail.com. Anon Srikiatkhachorn, UMASS, Center for Infectious Disease and Vaccine Research, Worcester, MA, E-mail: anon.srikiatkhachorn@umassmed.edu.

Save