Predictive Indicators for Forecasting Epidemic of Dengue/Dengue Haemorrhagic Fever Through Epidemiological, Virological and Entomological Surveillance Nguyen Thi Kim Tien, Do Quang Ha, Tran Khanh Tien, Luong Chan Quang Pasteur Institute, 167, Pasteur Street, Ho Chi Minh City, Vietnam By
Abstract Dengue/dengue haemorrhagic fever is the leading cause of death from the declared infectious diseases in children in southern Vietnam. The dengue epidemic in 1998 was an extensive one, with the morbidity rate of 455 cases/100 000 pop. and mortality rate of 1.27/100 000 pop. DEN-3 was the predominant serotype. However, in 1999, there was a lower morbidity (80.7 cases/100 000) and a lower mortality (0.23 deaths/100 000). The circulation of DEN-3 virus decreased and that of DEN-4 virus emerged. In both the years, most of the cases and deaths occurred in children under 15 years of age. Shock accounted for 15.3% and 13.5%, respectively, of all DF/DHF cases in 1998 and 1999, of which 2.5% and 2.1% died. The epidemic curve reached its peak in the rainy season (June to October). An excessive increase in the number of cases compared with the average of the previous five years and an increased circulation of a new dengue virus serotype during the first quarter may be used as indicators to predict larger epidemics in the year. There was no correlation of entomological indices with the epidemiological situation in this study. Key words: Dengue fever, Dengue haemorrhagic fever, Epidemics, Vietnam
Introduction Dengue fever/dengue haemorrhagic fever (DF/DHF) continues to be a major public health problem in Vietnam, especially in the southern region(1). In the southern region, with a population of over 27 million and high temperatures throughout the year, DF/DHF is the leading cause of high mortality in children from the declared infectious diseases. The annual DF/DHF 44
morbidity rates fluctuate from 100 to 450 cases/100 000 pop. and the number of deaths ranges from 60 to 220 per year. In 1998, an extensive DHF epidemic occurred in southern Vietnam ever since its first appearance in 1963. A total of 123 997 cases (accounting for 52% of all country’ s cases) and 347 deaths (accounting for 83% Dengue Bulletin – Vol 23, 1999
Predictive Indicators for Forecasting Epidemic of DF/DHF
of all country’ s deaths) were reported. In 1999, the DHF prevalence and number of deaths decreased remarkably. Throughout the two consecutive epidemics of DF/DHF, an assessment was made of the epidemiological, virological and entomological surveillance indicators that may be used to predict large outbreaks early enough to mount an effective control programme. An analysis of the same is presented in this paper.
Entomological surveillance The adult mosquito density index and Breteau index were used in the study.
Results Morbidity and mortality Table 1 includes the DF/DHF cases and deaths reported during the years 19931999. It is apparent that the number of cases and deaths in 1998 were two times more than the average during the previous five years (1993-1997)(455 cases compared with 217 cases and 1.27 deaths as compared with 0.62 deaths) while in 1999, the number of cases and deaths were less than a half of the average reported during 1993-1997. However, the case-fatality rate showed a decreasing trend from 1993 onwards (from 0.41% to 0.28%). This was possibly due toimprovements in the clinical management of DF/DHF cases.
Methods and materials Epidemiological surveillance The number of cases and deaths were reported from hospitals according to the clinical and laboratory criteria prescribed by the World Health Organization(2). The current surveillance system was used for data collection, reporting and feedback . Data of cases and deaths were reported weekly or monthly from hospitals to district preventive medicine centres and then to the Provincial Preventive Medicine Centre. From here, these data were sent to the Pasteur Institute, Ho Chi Minh City, and finally to the Ministry of Health.
Seasonal variation Figures 1 and 2 show the distribution of morbidity and mortality in 1993-1997, 1998 and 1999. DF/DHF cases occurred throughout the year, but the peak season was from June to October which is the rainy season in Vietnam. From the figures, we can identify that during the early months of 1998, the epidemic curve was much higher than that of 1993-1997 (the average number during the previous five years). From this, one could predict a large DF/DHF epidemic occurring in 1998. In comparison, during the early months of 1999, one could predict that there would be no big epidemic in this year.
Serological and virological surveillance Virus isolation was done at the Pasteur Institute by using C6/36 cells (Aedes albopictus) with blood samples collected from acute DF/DHF patients. Specific IgM antibody of dengue was detected in patients’ sera by using Mac-ELISA tests. Serum samples were collected from the patients at least 5 days after the onset of the illness.
Dengue Bulletin – Vol 23, 1999
45
Predictive Indicators for Forecasting Epidemic of DF/DHF Table 1. Number of DF/DHF cases and deaths reported from 1993 to 1999 Year 1993 1994 1995 1996 1997 Average of 1993-1997 1998 Decrease in 1998 over 1993-97 1999 Increase in 1999 over 1993-97 No. of cases 43048 25901 51897 55974 77370 50838 123997 234.9% 21951 43.2% Cases/ 100,000 pop. 172.7 104.1 203.7 288.0 319.9 217.7 455.7 209.3% 80.7 37.1% No. of deaths 177 61 165 178 222 160.60 347 216% 65 40.5% Deaths/ 100,000 pop. 0.71 0.25 0.65 0.67 0.83 0.62 1.27 204.2% 0.23 37% CFR (%) 0.41 0.24 0.32 0.32 0.29 0.31 0.28 90.3% 0.29 93.6%
Figure 1. Monthly distribution of reported DF/DHF cases in Southern Vietnam, 1993-1997, 1998 and 1999 30 25 20 15 ' Thousand cases
' '
'
10 5 ' ) ! ' ) ! ' ) ! ' ) ! ' ) !
' ! ! ) )
! )
! )
! )
' ! )
' ! )
0 Jan
Feb
Mar
Apr
May
Jun
Jul ' 1998
! 1993-1997
Aug Sep )1999
Oct
Nov
Dec
46
Dengue Bulletin – Vol 23, 1999
Predictive Indicators for Forecasting Epidemic of DF/DHF
Figure 2. Monthly distribution of reported DHF deaths in Southern Vietnam, 1993-1997, 1998 and 1999 80 70 60 50 40 30 20 10 ' ) ! ' ' ! ) Feb ) ! Mar ' ! ) Apr ) ! May ) Jun Jul ' ! ' ! ! ) ) Aug ! ) Sep ' ' Death cases '
' ! ) Oct ' ! ) Nov ! ) ' Dec
0 Jan
1998 ! 1993-1997 '
) 1999
Age distribution and disease severity DF/DHF cases in 1998 and 1999 occurred mostly in children under 15 years of age (i.e. 90% of the total cases). Shock accounted for 15.3% and 13.5% of all cases of DF/DHF in 1998 and 1999 respectively, while 2.5% and 2.1% of the shock cases died.
new serotype which had emerged since 1994, was the predominant serotype. So, at that time, a large epidemic was predicted to occur in that year. Figure 3 shows the correlation of the dengue serotypes isolated with the number of DHF cases reported during 1993-1999. The positive isolation rate of DEN-3 increased gradually from 1994 to 1997 (from 1% to 4%), and got its peak (8.2%) among other serotypes in 1998, resulting in the largest epidemic this year (455 cases/100 000 pop). In 1999, the circulation of DEN-3 was lower: 3.6%. DEN-1 continued to decrease from 1.4% in 1998 to 1% in 1999. DEN-2 increased from 1.7% (1998) to 3.4% (1999). Surprisingly, DEN-4 emerged in 1997, and increased gradually with a positive rate of 0.32% in 1998 and 2% in 1999. The emergence of this serotype may predict a large epidemic breaking out in the near future. 47
Virus isolation Table 2 shows the virus isolation rate distributed by month in 1998 and 1999. Most of the dengue viruses were isolated during June - October. The mean positive isolation rate was 11.6% in 1998 and 10% in 1999. In 1998, even in January and February, the isolation rates were high, being 18.9% and 10.6% respectively. DEN-3, the
Dengue Bulletin – Vol 23, 1999
Predictive Indicators for Forecasting Epidemic of DF/DHF Table 2. Monthly distribution of positive virus isolation rate in 1998 and 1999 Year 1998 D1 D2 D3 D4 Positive tests/ tests % 1999 D1 D2 D3 D4 Positive tests/ tests % Jan 1 3 12 1 17/ 90 18.9 0 0 0 0 0/37 15/ 142 10.6 0 0 0 0 0/21 0/ 10 0 0 0 1 0 1/35 3/ 57 5.3 0 1 0 0 1/34 5/ 103 4.9 0 0 2 0 2/37 25/ 244 10.2 3 2 1 2 8/46 25/ 179 14.0 0 5 1 2 8/64 25/ 104 24.0 1 3 4 5 13/65 2/ 78 2.6 0 3 0 2 5/47 Feb 3 6 6 Mar 0 0 0 Apr 0 0 3 May 0 0 5 Jun 5 2 18 Jul 2 6 17 Aug 6 0 19 Sep 0 0 2 Oct 0 4 15 1 20/ 161 12.4 0 1 6 0 7/70 Nov 0 0 4 2 6/ 53 11.3 2 3 5 1 11/56 0/ 15 0 0 2 1 0 3/74 Dec 0 0 0 Total 17 (1.4) 21 (1.7) 101 (8.2) 4 (0.32) 143/ 1236 11.6 6 (1) 20 (3.4) 21 (3.6) 12 (2) 59/586
0
0
2.85
3.94
5.4
17.39
12.5
20.00
10.63
10.0
19.64
4.05
10.07
Figure 3. Correlation of positive dengue virus isolation rate with DF/DHF morbidity in Southern Vietnam, 1993-1999 500 400 300 Cases/100000 % Dengue (+) 20
)
15
) 200
& D3
10
) 100 0
+ & 1993
& + 1994
& + 1995
& ) + 1996
& ) + 1997 D1
& ) + 1998 D4 D2
5
) + 1999
0
Cases ) DEN-1 + DEN-2 & DEN-3
DEN-4
48
Dengue Bulletin – Vol 23, 1999
Predictive Indicators for Forecasting Epidemic of DF/DHF
Serological surveillance The IgM antibody anti-dengue was identified from the patients’ sera in all months of the years 1998 and 1999. The mean seropositive rate was over 50% in 1998 and 44.3% in 1999. The seropositive result of Mac-ELISA tests during the preepidemic stage (first quarter) may predict alarge epidemic occurring during the succeeding months of the year.
for the period 1993-97, 1998 & 1999 are included in Figures 4 and 5. It is evident from Figures 4 and 5 that there was no correlation between the entomological data and the epidemiological situation: the curves of A. aegypti density index and Breteau index of 1998 (the year of the largest DHF epidemic) were lower than that of the years 1993-1997. However, the curve of 1998 was sometimes higher than that of 1999.
Entomological data Breteau and density indices at sentinel surveillance points in Southern Vietnam
Conclusion There was a large DHF epidemic in 1998 in southern Vietnam, with high numbers of cases
Figure 4. Breteau index at sentinel surveillance points in southern Vietnam, 1993-1997, 1998 & 1999
140 120 100 80 ) 60 # ' 40 20 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec # ) ) # ' ' ' # # ) # ) ' ' ' # ' # ) ) ) ) ' # # ' ) ) # ' ' ) #
' 1993-1997 #1998 )1999
Dengue Bulletin – Vol 23, 1999
49
Predictive Indicators for Forecasting Epidemic of DF/DHF
Figure 5. Density index of Aedes aegypti at sentinel surveillance points in southern Vietnam –1993-1997, 1998 & 1999 3.5 3 2.5 2 1.5 1
'
'
#
' # # ' ) ' # ) ) # ' ) # ' ) # ' )
' ' # )
'
'
)
) 0.5 # 0 Jan
) #
# ) # Apr May
)
Feb
Mar
Jun
Jul # 1998
Aug
Sep
Oct
Nov
Dec
'1994-1998
) 1999
and deaths; and DEN- 3 was the predominant serotype. In 1999, there was a great remission of DHF cases and deaths, but there was a resurgence of DEN-4, a new serotype, which has tended to increase gradually. Any emerging new serotype may be used as a predictive indicator. The excessive increase in the number of DHF cases and deaths in the first quarter of 1998 was noted to compare with the average morbidity of the previous five years. At the same time, there was an increased circulation of the new serotype
of dengue virus and the seropositive results of Mac-ELISA tests in the pre-epidemic stage, all these may be considered as predictive indicators for a larger DHF epidemic. There was no correlation between the entomological indices and the epidemiological situation in this study. References 1. Do Quang Ha and Trinh Quan Huan. Dengue activity in Vietnam and its control programme. Dengue Bulletin, WHO/SEARO, Vol 21, Dec 1997:35-40. 2. World Health Organization. Dengue haemorrhagic fever: Diagnosis, treatment, prevention and control. World Health Organization, Geneva, 1997.
50
Dengue Bulletin – Vol 23, 1999