Chapter 8

Dengue and dengue haemorrhagic

fever
James W LeDuc, Karin Esteves, Norman G. Gratz

Introduction
Dengue, a mosquito-borne viral disease, is becoming an increasing public
health problem in developing countries. This was recognized by the WHO,
which confirmed in a resolution at the forty-sixth World Health Assembly
in 1993 that dengue prevention and control should be among the priorities
of the Organization.
Dengue is caused by any of four antigenically and genetically distinct
viruses of the family Flaviviridae, named dengue-1, -2, -3 and -4 (Craven
1991). Humans are the primary vertebrate hosts of these viruses, although
there is evidence of a silent zoonotic cycle involving mosquitoes and monkeys
in parts of Asia and Africa (Rudnick 1965, Fagbami, Monath & Fabiyi 1977,
Rudnick 1984, Cordellier et al. 1983). The Aedes aegypti mosquito is the
principal urban vector (Clark & Casals 1965). Dengue fever is characterized by an abrupt onset of fever, headache, myalgia, loss of appetite, and
varying gastrointestinal symptoms, often accompanied by rash and bone or
joint pains (Rush 1789, Hammon, Rudnick & Sather 1960, Ehrenkranz et
al. 1971, San Juan Laboratories 1979, Kuberski et al. 1977, Lopez-Correa
et al. 1979). Symptoms persist for 37 days, and while convalescence may
be prolonged for several weeks, mortality is rare. Homologous immunity
is lifelong, but cross-protection to other dengue viruses is not elicited and,
indeed, there is evidence to suggest that heterologous antibodies may form
infectious immune complexes which may increase the severity of subsequent
infections (Halstead, Nimmannitya & Margiotta 1969, Halstead 1970,
1980, Kliks et al. 1989). Infected people are infective for feeding mosquitoes
during the febrile phase of their illness, and mosquitoes may then transmit
the virus to others at subsequent blood-meals after an incubation period of
approximately one week, depending upon ambient temperatures (Watts et
al. 1987). Infected female mosquitoes may also pass the virus on to progeny by transovarial transmission (Rosen et al. 1983, Khin & Than 1983,
Hull et al. 1984, Freier & Rosen 1987, 1988, Rosen 1987, 1988, Fauran,

220

Global Epidemiology of Infectious Diseases

Laille & Moreau 1990, Shroyer 1990, Mitchel & Miller 1990, de Souza
& Freier 1991, Bosio et al. 1992, Serufo et al. 1993). Mosquitoes infected
transovarially may transmit dengue virus to susceptible humans at their first
blood-meal. Once infected, mosquitoes remain infected for life and may
transmit dengue virus either during probing, interrupted blood feeding, or
when feeding to repletion. Thus, a single infected mosquito may transmit
the virus to several susceptible humans over its lifetime.
Dengue haemorrhagic fever (DHF) and dengue shock syndrome (DSS)
are more severe manifestations of dengue infection, and are most often
associated with infection of people with pre-existing antibodies to dengue,
either actively or passively acquired (Quintos et al. 1954, Hammon et al.
1960, Halstead, Nimmannitya & Margiotta 1969, Ramirez-Ronda et al.
1979, Halstead 1970, 1980). The clinical course follows that of classic
dengue fever initially; however, as the febrile phase subsides, the patients
condition rapidly deteriorates with signs of circulatory failure, neurological manifestations, and hypovolemic shock with potential fatal outcome if
prompt, proper, clinical management is not implemented. Haemorrhage
may be evident by petechiae, easy bruising, bleeding at injection sites, and
occasionally gastrointestinal bleeding, especially where gastric ulcers already
exist (Tsai et al. 1991).
There is no specific treatment for dengue fever. Although convalescence
may last several weeks following dengue infection, especially among adults,
disabling sequelae are not thought to follow infection. Dengue haemorrhagic fever and DSS require symptomatic treatment, coupled with careful
assessment of fluid and electrolyte balance for management of shock (World
Health Organization 1986, Craven 1991). Areas where clinicians are not
familiar with the proper treatment of DHF/DSS may have case fatality rates
substantially higher than those where physicians are experienced in the
proper treatment of the disease. Diagnosis is made on clinical grounds and
confirmed by laboratory demonstration of specific immune response, virus
isolation, or identification of dengue virus nucleic acid sequences in clinical
material (Russell & McCown 1972, Russell, Brandt & Dalrymple 1980,
Pan American Health Organization 1982, Repik et al. 1983, Henchal et al.
1986, Kerschner et al. 1986, World Health Organization 1986).
Dengue is the most widespread vector-borne virus disease in the world,
and DHF/DSS are rapidly increasing in incidence in many tropical areas
(Halstead 1992, 1993a, 1993b). This dramatic increase is a result of
changes in human lifestyle, increased international travel, and urban
crowding. Dengue has been known for at least 200 years, but DHF/DSS
has only been routinely recognized since the 1950s, although there is
evidence to suggest that DHF occurred as early as 1897 in north-eastern
Australia (Hare 1898, Craven 1991, Halstead 1993a). Following the end
of the Second World War, Asian cities experienced rapid growth and urban
crowding, and with these came increased habitat for the vector mosquito of
dengue, Aedes aegypti, which selectively breeds in water containers associated with human habitations. These breeding sites can include water jugs,

Dengue and dengue haemorrhagic fever

221

blocked drain spouts, decorative plant containers, discarded refuse that

may retain water, and many other items associated with modern life (Gratz
1993a). A pattern of dengue transmission then became evident, especially
in major cities of South East Asia, first with increasingly frequent epidemics
of classic dengue, then merging to continuous endemic transmission and
appearance of DHS/DSS, initially sporadically but becoming more common within the last 20 years, until these too are now endemic (Akiyama
1993, Gratz 1993b, Gubler 1993, Itoh 1993, Kojima 1993, Lam 1993).
This pattern has been seen in varying degrees throughout much of tropical
Asia, from the southern provinces of China to the west coast of India, and
south to northern Australia and many Pacific island nations.
The same evolutionary pattern is now being repeated in the urban
centres of Latin America (Gubler 1993, Halstead 1992, 1993a). In the
1960s, a hemisphere-wide Aedes aegypti eradication programme was in
existence, and nearly succeeded in total eradication of the vector. However, the decision was made to change the strategy from eradication to
control. As a result, Aedes aegypti has now reinfested virtually all areas
where it was once eliminated, and has even expanded to new habitats. As
a result of this reinfestation, transmission of dengue has recently occurred
in almost all parts of Latin America and the Caribbean (Pan American
Health Organization 1993). It is currently limited to increasingly frequent
epidemics of dengue fever, often coupled with sporadic cases or outbreaks
of DHF/DSS. The pattern is, however, identical to that experienced in Asia
during the 1950s and 1960s, and it is clear that DHF/DSS will become
increasingly common in the years to come (Gubler 1993). Indeed, in 1994
a major outbreak of dengue and DHF occurred in north-eastern Brazil,
and some experts have suggested that as many as 300 000 cases might
have occurred.
In Africa, similar conditions exist, but reporting has been fragmentary,
and there are no reliable estimates of the number of persons infected or the
virus serotypes in circulation (Monath et al. 1974, Fagbami, Monath &
Fabiyi 1977, Roche et al. 1983). As a typical example, however, in 1993
an outbreak of dengue occurred in Comoros, during which an estimated
60 000 cases occurred (Pan American Health Organization 1993).

Review of major studies and data available

Because of the variation in clinical presentation and lack of readily available, inexpensive diagnostic tests, dengue fever is generally not formally
reported. In some endemic countries, such as Thailand, DHF/DSS may be
systematically recorded and reports of such surveillance are periodically
published. Both WHO and the Rockefeller Foundation have monitored
formal country reports of DHF/DSS, and these reports formed the basis
of our estimates for incidence of disease (Halstead 1992, 1993b, Okabe
1993, Pan American Health Organization 1993). The published scientific
literature is generally limited to reports of specific outbreaks and most

222

Global Epidemiology of Infectious Diseases

often does not offer a quantitative historical perspective of dengue or

DHF/DSS. It is thus of limited usefulness in preparing burden of disease
analyses. The problem is further complicated by the fact that dengue
and DHF/DSS frequently occur as outbreaks or epidemics, often above a
background of endemic transmission. Consequently, larger outbreaks may
be recognized and reported; indeed, estimates of the number of cases may
be inflated, since many febrile diseases may be reported then as dengue.
Between outbreaks, uncomplicated dengue may be merged with all other
causes of febrile illness, and thus there may be general underreporting of
the disease.
The most comprehensive cost-benefit analysis of dengue and DHF/DSS
was prepared by Shepard & Halstead (1993), and their estimates are
incorporated into the present model whenever possible.
It is generally believed that millions of people suffer from dengue
fever each year in Asia, Africa, the Pacific Islands, and the Americas (Pan
American Health Organization 1993). It is virtually impossible, however,
to document the global impact of dengue fever given the general absence
of surveillance. For the purposes of this publication, we have focused on
the more severe complications of dengue infection, DHF/DSS, and severe
dengue fever cases. These conditions often require hospitalization and
form the basis for the limited number of officially reported cases. Thus,
our calculations are based on reported numbers of cases whenever possible.
As these estimates reflect only the most clinically severe cases, they almost
certainly underestimate the true burden of disease. We also used average
incidence rates whenever possible. These were usually determined based
on the number of cases reported over the past decade. This is justified
since dengue and DHF are usually epidemic diseases, where the number
of cases reported during an outbreak may well be 100 times greater than
during inter-epidemic periods. Averaging over a 10-year period serves
to compensate for these wide annual fluctuations. The trend, however,
in nearly all endemic areas, and especially in the Americas, is towards
more frequent and larger outbreaks. Unfortunately, our estimates fail to
reflect this upward trend, and thus also lead to an underestimate of the
true disease burden.

International Classification of Diseases

The specific categories related to dengue in the ninth and tenth revisions of
the International Classification of Diseases (ICD) are listed in Box 8.1.
Ninth revision
Uncomplicated dengue fever, also listed as breakbone fever, is coded as
061 in the ninth revision of the International Classification of Diseases.
Alternative codings could include 066: other arthropod-borne viral diseases, and 780.6: fever of unknown origin. Dengue haemorrhagic fever is
coded as 065.4, with the following synonymsBangkok, Chikungunya,

Dengue and dengue haemorrhagic fever

Box 8.1

International Classification of Diseases (ICD) codes

related to dengue and dengue haemorrhagic fever

ICD 9
061

223

ICD 10
Uncomplicated dengue fever
(also called breakbone fever)

A90

Classical dengue fever

A91

Dengue haemorrhagic fever

A9099 Anthropod-borne viral

haemorrhagic fevers and viral
haemorrhagic fevers

mosquito-borne, Philippine, Singapore, South-East Asia and Thailand

haemorrhagic feverall sharing the same code, even though Chikungunya is both antigenically and taxonomically a completely different virus.
Other classifications could result from coding based on clinical disease
rather than viral etiology, and might include 785.5: shock, hypovolaemic
without mention of trauma, and 448.9: haemorrhage, capillary, diseases
of capillaries, other, and unspecified.
Tenth revision
Classical dengue fever is coded as A90 in the tenth revision of the International Classification of Diseases, and dengue haemorrhagic fever is
classified as A91. The codes for arthropod-borne viral fevers and viral
haemorrhagic fevers are A90A99. Related codes that might include
dengue or dengue haemorrhagic fever include B33: other viral diseases,
not elsewhere classified, and B34: viral infection of unspecified site. Codes
based on clinical disease rather than specific etiology might include: D65:
disseminated intravascular coagulation (defibrination syndrome); D69:
purpura and other haemorrhagic conditions, especially D69.9: haemorrhagic condition, unspecified; R04: haemorrhage from respiratory passages,
especially R04.8: haemorrhage from other sites in respiratory passages;
R21: rash and other nonspecific skin eruption; R50: fever of unknown
origin; R57: shock, not elsewhere classified, especially R57.1: hypovolaemic shock, and R57.9: shock, unspecified; and H58: haemorrhage, not
elsewhere classified.

Estimated Burden of Dengue and Dengue

Haemorrhagic Fever
The most comprehensive data available for dengue and DHF/DSS originate
in South East Asia, in the Other Asia and Islands region. We have based
our estimates on data from this region, then tried to estimate the situation
in other areas by comparison with Other Asia and Islands.

224

Global Epidemiology of Infectious Diseases

Other Asia and Islands

Population at risk. Countries within the region were reviewed to exclude
those nations not thought to be at risk of severe dengue/DHF. An estimate
of 86.4 per cent of the regional population was considered at risk of this
disease. Countries or areas considered not at risk included: Bhutan, Democratic Peoples Republic of Korea, Mongolia, Nepal, Republic of Korea, and
Hong Kong (total population 92 652 000). The age-specific and sex-specific
population at risk was determined by multiplying the regional age-specific
and sex-specific population estimates by 0.864.
Overall incidence rates. The average number of reported cases in the
decade prior to 1992 was calculated based on statistics summarized in
recent publications (Halstead 1993a, 1993b) or those obtained by Gratz
(unpublished data). Those averages were used to determine country-specific
annual incidence rates per 1000 population based upon population estimates
for those countries.
These incidence rates were then used to estimate an overall incidence
rate for the regional populations at risk of severe dengue/DHF. The overall incidence rate among the population at risk was estimated to be 0.3
per 1000. Because of general underreporting and differences in reporting
criteria, this estimate is likely to be low. Furthermore, it is based on a 10year average, and in most parts of this region the incidence of dengue and
DHF/DSS is increasing, another reason to suspect that the estimate is low.
Nonetheless, there are no better estimates available for this or any other
region, and we therefore have based our calculations on this estimate.
Sex-specific incidence rates. Published reports indicate that female cases
of severe dengue/DHF exceed male cases by a ratio of 1.2:1.0 (Shepard
& Halstead 1993). Thus, we estimated the contribution of male cases at
an incidence rate of 0.27 per 1000, and females at 0.33 per 1000. These
rates were used to determine the total estimated number of cases.
Age-specific incidence rates. Published reports indicate that approximately
95 per cent of severe dengue/DHF occurs in children under 15 years old
(Shepard & Halstead 1993). Among these children, 75 per cent of cases
are estimated to occur in the 514 year age group, and 25 per cent in the
04 year age group. We estimated the contribution from the remaining age
groups as 3 per cent for 1544 year, 1.5 per cent for 4559 year, and 0.5
per cent for 60 years and over. We found no published reports to validate
the estimates for adults.
Estimated number of cases. Age-specific incidence rates were used to estimate
the age-specific number of cases by multiplying the fraction by the total
sex-specific estimated number of cases.
Estimated number of deaths. Case fatality rates for dengue/DHF were

Dengue and dengue haemorrhagic fever

225

reviewed for countries within the region where information was available. Case fatality rates clustered into two groups: those with rates that
averaged 0.5 per cent, and those with rates that averaged 5.2 per cent.
We included Malaysia, Singapore, Taiwan, Thailand and Viet Nam as
low threat areas, and Bangladesh, Cambodia, Indonesia, Lao Peoples
Democratic Republic, Myanmar, Papua New Guinea, Philippines, and
Sri Lanka, and all islands as high threat areas, even though reliable
estimates of disease attributable to dengue/DHF do not exist for some of
these countries. A total of 27.7 per cent of the at-risk population lived in
low-threat areas, and 72.3 per cent in high-threat areas. We calculated
the estimated number of age-specific and sex-specific cases in these areas
by multiplying these factors by the total estimated number of cases. We
did not have sufficient information to divide populations into risk areas in
the other regions examined. The number of deaths in high-threat areas
is estimated by multiplying the estimated number of cases in that area by
the average case fatality rate, 5.2 per cent. The number of deaths in low
threat areas was estimated by multiplying the estimated number of cases
in that area by the average case fatality rate, 0.5 per cent. These values
were summed to estimate the total number of deaths for the region, and
are shown in Table 8.1.
India
Population at risk. Based on very limited information presented at a meeting
on dengue held in 1994 in Pune, India, and cosponsored by the National
Institute of Virology, the Rockefeller Foundation and the World Health
Organization, we estimated that roughly 80 per cent of the population of
Table 8.1
Sex and age
(years)

Estimates of the burden of dengue haemorrhagic fever,

Other Asia and Islands
Regional population (thousands)

Population at risk
(thousands)

Estimated
number of cases

Estimated
number of deaths

04
514
1544
4559
60+

43 763
84 032
160 825
34 138
20 208

37 811
72 604
138 953
29 495
17 460

19 002
57 005
2 400
1 200
400

736
2 208
93
46
15

Total

342 966

296 323

80 007

3 099

04
514
1544
4559
60+

41 988
80 217
159 611
35 090
22 662

36 278
69 307
137 904
30 318
19 580

22 994
68 983
2 905
1 452
484

891
2 672
112
56
19

Total

339 568

293 387

96 818

3 750

Males

Females

226

Global Epidemiology of Infectious Diseases

India is at risk of infection with dengue/DHF (World Health Organization 1994c).

Overall incidence rates. We found no reliable estimates of dengue incidence
in India; however, it was apparent from discussions during the meeting that
the problem may approach or exceed the magnitude of that seen in the
Other Asia and Islands Region. Consequently, we used the same estimate
for overall incidence as applied for that region, 0.3 per 1000.
Sex-specific incidence rates. We used the same sex-specific incidence rates
as used for the Other Asia and Islands Region.
Age-specific incidence rates. We used the same age-specific incidence rates
as used for the Other Asia and Islands Region.
Estimated number of cases. Age-specific incidence rates were used to estimate
the age-specific number of cases by multiplying the fraction by the total
sex-specific estimated number of cases.
Estimated number of deaths. A presentation made at the meeting on dengue indicated that there is a wide spread need for educational programmes
for medical staff on the proper management of DHF/DSS patients (World
Health Organization 1994c). Existing mortality rates attributable to DHF
may be approximately the same as in areas considered as high threat in the
Other Asia and Islands Region. We therefore used a case fatality rate of 5.2
per cent, realizing that additional information is clearly needed to form the
basis for a more accurate estimate.
The number of deaths was estimated by multiplying the estimated number of cases by the case fatality rate, 5.2 per cent. Estimates for the disease
burden attributable to DHF in India are shown in Table 8.2.
China
Population at risk. Only the southern provinces of China are at risk for
dengue/DHF; dengue does not occur in areas where hard freezes occur in
winter. Consequently, we estimated that 10 per cent of the total population
of China is at risk for dengue/DHF.
Overall incidence rates. The incidence of dengue/DHF in southern China
is less than in India or the Other Asia and Islands Regions, but nonetheless
thought to be substantial. We found no quantitative estimates of dengue/
DHF incidence rates in China, so we estimated an incidence rate of 0.2
per 1000, somewhat less than that used for the Other Asia and Islands
Region. One might suspect that the incidence in southern China would
approximate that seen in adjacent areas of the Lao Peoples Democratic
Republic and Viet Nam.

Dengue and dengue haemorrhagic fever

Table 8.2
Sex and age
(years)

227

Estimates of the burden of dengue haemorrhagic fever, India

Regional population (thousands)

Population at risk
(thousands)

Estimated
number of cases

Estimated
number of deaths

04
514
1544
4559
60+

59 789
101 752
200 525
47 567
29 768

47 831
81 402
160 420
38 054
23 814

22 541
67 624
2 847
1 424
475

1 172
3 516
148
74
25

Total

439 401

351 521

94 911

4 935

04
514
15-44
4559
60+

56 679
95 263
183 242
46 005
28 924

45 343
76 210
146 594
36 804
23 139

25 714
77 142
3 248
1 624
541

1 337
4 011
169
84
28

Total

410 113

328 090

108 270

5 630

Males

Females

Sex-specific incidence rates. We used the same ratio of sex-specific incidence

rates as used for the Other Asia and Islands Region; thus the incidence rate
for males was 0.18 per 1000 and for females 0.22 per 1000.
Age-specific incidence rates. We used the same age-specific incidence rates
as for the Other Asia and Islands Region.
Estimated number of cases. Age-specific incidence rates were used to estimate
the age-specific number of cases by multiplying the fraction by the total
sex-specific estimated number of cases.
Estimated number of deaths. We used an intermediate case fatality rate of
3.7 per cent for the entire population at risk. The number of deaths was
estimated by multiplying the estimated number of cases by this rate. These
estimates are presented in Table 8.3.
Sub-Saharan Africa
Population at risk. There is very little quantitative information regarding
the actual population at risk of severe dengue/DHF in Africa. Periodic
reports have been published that clearly indicate that dengue viruses are
present in many different parts of Africa, and recent outbreaks of dengue
have occurred in Comoros and among United Nations personnel deployed
to Somalia (Halstead 1992, World Health Organization 1993). Few serological surveys have been attempted and interpretation of those done is
hampered by the presence of other flaviviruses, which cross-react with
dengue in many serological tests. Dengue viruses are present, particularly
around urban centres where Aedes aegypti may be abundant. Based on very

228

Table 8.3
Sex and age
(years)

Global Epidemiology of Infectious Diseases

Estimates of the burden of dengue haemorrhagic fever, China

Regional population (thousands)

Population at risk
(thousands)

Estimated
number of cases

Estimated
number of deaths

04
514
1544
4559
60+

60 243
96 997
306 305
72 674
48 980

6 024
9 700
30 631
7 267
4 898

2 502
7 505
316
158
53

93
278
12
6
2

Total

585 199

58 520

10 534

390

04
514
15-44
4559
60+

57 946
90 401
284 078
64 403
51 666

5 795
9 040
28 408
6 440
5 167

2 866
8 598
362
181
60

106
318
13
7
2

Total

548 494

54 849

12 067

446

Males

Females

crude estimates, we considered that 20 per cent of the residents of Africa

were at risk of dengue/DHF, even though we could locate no confirmed
record of DHF in sub-Saharan Africa.
Overall incidence rates. The limited information currently available indicates that dengue/DHF is less abundant than in India or the Other Asia
and Islands Region. We estimated the incidence to be 0.1 per 1000, with
little factual basis for this figure.
Sex-specific incidence rates. We used the same ratio of sex-specific incidence
rates as for the Other Asia and Islands Region, thus we used an incidence
rate of 0.09 per 1000 males and 0.11 per 1000 females.
Age-specific incidence rates. We used the same age-specific incidence rates
as for the Other Asia and Islands Region.
Estimated number of cases. Age-specific incidence rates were used to estimate
the age-specific number of cases by multiplying the fraction by the total
sex-specific estimated number of cases.
Estimated number of deaths. We used a high case fatality rate of 5.2 per
cent, based on the fact that the incidence rate is relatively low, thus giving
clinicians little experience with the disease, and realizing that virtually no
special training has been provided to clinicians in Africa for management
of DHF/DSS. We used only a single case fatality rate for all of the SubSaharan Africa Region. The estimated number of deaths was calculated by

Dengue and dengue haemorrhagic fever

229

multiplying the case fatality rate by the estimated number of cases. These
estimates are presented in Table 8.4.
Latin America and the Caribbean
Population at risk. Dengue and DHF/DSS are rapidly expanding throughout Latin America and the Caribbean, closely following the reinfestation
of most urban centres and many rural areas by the vector mosquito of
dengue, Aedes aegypti. As a result, the general trend has been to increasing
incidence of disease, especially over the past decade. Limited quantitative
information is available to support specific country estimates, although the
Pan American Health Organization has attempted to monitor dengue and
DHF for more than a decade (Pan American Health Organization 1993).
The most dramatic outbreak of dengue and DHF/DSS occurred in Cuba in
1981, primarily between the months of June and October, when 344 203
cases of dengue were reported, of which 116 151 were hospitalized (Kouri,
Guzman & Bravo 1986). Of the hospitalized cases, 9203 were considered
serious, and an additional 1109 were considered very serious. A total
of 158 deaths were recorded (the case fatality rate was 1.5 per cent of
serious and very serious cases, or 0.14 per cent of hospitalized cases).
While Cuba has subsequently virtually eliminated Aedes aegypti mosquito
populations, the dramatic potential impact of epidemic dengue or DHF
on a nations health care system is nonetheless evident from these figures.
Indeed, at the peak of the Cuban epidemic, approximately 10 000 cases
were hospitalized per week, enough to swamp even the most sophisticated
medical system. Considering the expanding dengue incidence throughout
most of Latin America and the Caribbean in both urban and rural settings,
Table 8.4
Sex and age
(years)

Estimates of the burden of dengue haemorrhagic fever,

Sub-Saharan Africa
Regional population (thousands)

Population at risk
(thousands)

Estimated
number of cases

Estimated
number of deaths

04
514
1544
4559
60+

47 484
70 258
103 764
20 308
10 508

9 497
14 052
20 753
4 062
2 102

1 079
3 236
136
68
23

56
168
7
4
1

Total

252 322

50 464

4 542

236

04
514
15-44
4559
60+

47 030
69 818
106 257
22 117
12 730

9 406
13 964
21 251
4 423
2 546

1 035
1 536
2 338
487
280

54
80
122
25
15

Total

257 952

51 590

5 675

295

Males

Females

230

Global Epidemiology of Infectious Diseases

we have estimated that 60 per cent of the population of this region is at

risk of severe dengue/DHF.
Overall incidence rates. While the incidence of dengue is less in Latin
America and the Caribbean than in Asia, it is nonetheless significant and
increasing. From 1988 to 1992, an average of 102 162 cases of dengue
fever were reported in the region. During that same period, an average
of 2100 DHF cases were reported annually, with a mean of 31 deaths
attributed to the disease (Pan American Health Organization 1993), as
shown in Table 8.5.
Based on these reported rates, we have estimated the annual incidence
of DHF to be 0.006 per 1000.
Sex-specific incidence rates. We used the same ratio of sex-specific incidence
rates as for the Other Asia and Islands Region; thus the rate for males was
0.0056 per 1000 and for females 0.0063 per 1000.
Age-specific incidence rates. We used the same age-specific incidence rates
as for the Other Asia and Islands Region.
Estimated number of cases. Although dengue fever is rapidly increasing in
most countries of the region, DHF remains relatively uncommon, and thus
we used the same annual incidence rate for all populations at risk.
Estimated number of deaths. We used a case fatality rate of 1.5 per cent
based on the average number of deaths reported from 1988 to 1992. A
summary of estimates for the population at risk, number of cases and
deaths is found in Table 8.6.
Middle Eastern Crescent, Formerly Socialist Economies
of Europe, and Established Market Economies
In these regions, dengue and DHF/DSS are generally limited to tourists
travelling to and from endemic areas in other regions, although transmisTable 8.5

Reported cases of dengue fever, dengue haemorrhagic fever,

and deaths attributable to dengue haemorrhagic fever in
Latin America and the Caribbean, 1988 to 1992

Year

Dengue fever

Dengue haemorrhagic fever

Deaths

1988
1989
1990
1991
1992

47 783
89 138
116 389
155 543
101 958

86
2 682
3 646
1 760
2 319

6
33
62
31
23

Average

102 162

2 099

31

Source: Pan American Health Organization (1993).

Dengue and dengue haemorrhagic fever

Table 8.6
Sex and age
(years)

231

Estimates of the burden of dengue haemorrhagic fever,

Latin America and the Caribbean
Regional population (thousands)

Population at risk
(thousands)

Estimated
number of cases

Estimated
number of deaths

Males
04
514
1544
4559
60+

28 721
52 124
104 286
22 249
14 231

22 977
41 699
83 429
17 799
11 385

236
707
30
15
5

2
7
0
0
0

Total

221 611

177 289

993

10

04
514
15-44
4559
60+

27 676
50 744
104 085
23 355
16 824

22 141
40 595
83 268
18 684
13 459

267
800
34
17
6

4
12
1
0
0

Total

222 684

178 147

1 122

17

Females

sion may exist in a few population centres. For example, in the Middle
Eastern Crescent Region, there is growing evidence that active dengue
transmission probably occurs in some locations. Nonetheless, the number
of cases in these regions is relatively small, and no quantitative information
exists upon which incidence estimates could be based.

Validity of incidence estimates

Clearly the estimate of incidence rates is the single most critical value in
calculating the global burden of disease. Unfortunately, the data from which
this estimate must be drawn, at least for dengue/DHF, are very weak. The
incidence of dengue/DHF depends upon several factors, including the abundance of potential mosquito vectors, especially Aedes aegypti, the number
of susceptible humans in the population at risk, and the presence of dengue
viruses circulating in or introduced into those populations. Other factors,
especially environmental and sociological, also influence estimates of the
incidence of dengue/DHF. The fact that distinct viruses may cause dengue,
the confusion among clinicians about what actually distinguishes dengue
from DHF, and the limitations in availability of laboratory confirmation,
all contribute to the ambiguity of incidence estimates. Further, dengue and
DHF tend to occur in epidemics, so that tremendous numbers of cases may
occur during one period, followed by subsequent years when relatively
few cases may be recorded. Finally, as with all diseases, the estimates are
only as good as the surveillance and reporting systems from which the
estimates are drawn. In general, for dengue and DHF, these are not well
developed. Between outbreaks, there is likely to be underreporting, dengue

232

Global Epidemiology of Infectious Diseases

being grouped with undifferentiated febrile illness or haemorrhagic diseases

of undetermined origin. During outbreaks, there may be overreporting,
since all febrile illnesses may be reported as dengue, and any febrile disease
with bleeding manifestations as DHF.
To overcome these potential problems, we have concentrated our efforts
at estimation on incidence in the Other Asia and Islands Region, where
dengue and DHF are best known clinically and historically, and where we
feel that the reporting systems are most reliable. To overcome the large
annual variations in incidence of dengue and DHF, we have attempted to
average the incidence rates for the most recent 10-year period for which
complete information is available. The actual number of reported DHF
cases by selected countries are presented in Table 8.7. We have attempted
to calculate the overall regional incidence rates by taking into account the
specific rates for those countries for which information is available, and
estimating for other countries in the region where DHF is known to occur,
but no published information on incidence rates is available. Whenever
possible, we used these same techniques in the other areas considered;
however, the data are generally much less reliable for the other regions.
Consequently, we used the rate for the Other Asia and Islands Region
as a reference point, then determined whether other regions had greater,
equal, or lower incidence rates, based on informal reporting and published
information from recent outbreaks. The gross estimates are summarized
in Table 8.8.
We suspect that the final product of this exercise grossly underestimates
the true burden of DHF. The question that remains to be answered is the
order of magnitude that this underestimation represents. Attempts are
in progress to improve dengue diagnosis and surveillance in hopes of
generating more accurate information on the true global burden of this
increasingly important disease.
In reviewing the data available to estimate incidence, a striking trend
is consistently seen. Dengue and DHF are increasing in virtually every
location where they are found, and are rapidly appearing in new areas,
following infestations by the vector mosquito, Aedes aegypti. This rapid
increase is most notable in the Other Asia and Islands Region, where DHF
has become endemic and a constant threat in many countries, and in Latin
America and the Caribbean, where dengue epidemics are increasingly
common and DHF is seen with increasing frequency. Given that the estimates for incidence were based on 10-year averages, they are most likely
underestimates of the true incidence of disease. The global burden of this
disease is likely to be much greater in the coming years.
Other considerations in estimating disease burden
The focus of this discussion has been on severe dengue and DHF, conditions likely to require hospitalization and thus be captured by a national
surveillance system. Uncomplicated dengue fever has not been considered
although it is acknowledged that there are many more dengue fever patients

Cases

5 909
4 665
13 875
12 710
13 875
12 710
22 765
47 573
10 362
13 043
21 120
17 620

196 2270

1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992

Total

6 994

231
255
491
382
460
600
1 039
1 527
464
458
578
509

Deaths

Indonesia

16 152

486
3 052
790
702
384
1 403
2 025
1 448
2 401
2 071
741
649

Cases

211

14
36
10
5
11
9
8
2
14
38
39
25

Deaths

Malaysia

38 822

1 524
1 706
2 856
2 323
2 666
2 191
7 292
1 181
1 196
6 318
8 055
1 514

Cases

1 372

90
49
83
39
134
111
222
65
52
182
305
40

Deaths

Myanmar

13 979

123
305
1 684
2 545
2 096
687
859
2 922
305
588
1 865
n/a

Cases

654

8
31
130
89
210
30
27
68
14
27
20
n/a

Deaths

Philippines

9 535

133
216
205
86
126
354
436
245
944
1 733
2 179
2 878

Cases

180

0
0
2
0
2
0
0
0
0
4
6
4

Deaths

Singapore

681 013

25 641
22 250
30 022
69 597
80 076
29 030
170 630
26 926
69 204
113 855
43 782
n/a

Cases

3 689

194
159
231
451
542
206
896
189
280
422
119
n/a

Deaths

Thailand

Dengue haemorrhagic fever reported, selected countries, Other Asia and Islands, 19821992

Year

Table 8.7

1 009 640

35 323
39 806
149 519
30 498
45 107
46 266
354 517
85 160
40 205
37 569
94 630
51 040

Cases

7 455

408
361
1 798
368
399
511
1 566
826
289
255
403
271

Deaths

Viet Nam

Dengue and dengue haemorrhagic fever

233

234

Table 8.8

Global Epidemiology of Infectious Diseases

Crude estimate for the global burden of disease attributable

to dengue haemorrhagic fever, 1994

Region

Cases

Deaths

Other Asia and Islands

India
China
Sub-Saharan Africa
Latin America and Caribbean

180 000
205 000
23 000
11 000
3 000

7 000
11 000
1 000
600
100

Total

422 000

19 700

for each case of DHF. Consequently, the nearly half a million cases of DHF
estimated here are likely to reflect 550 million or more dengue infections
annually. Certainly in epidemic years, this could well be an underestimation. The cost of uncomplicated dengue is perhaps most noticeable
among the adult population, where the disease is more often clinically
apparent, frequently leading to acute illness of about a weeks duration,
and occasionally with prolonged convalescence. Those at greatest risk are
residents of areas recently infested with vector mosquitoes and exposed
for the first time as adults to dengue, as is happening now in many parts
of Latin America, and travellers (including deployed military personnel,
business people, tourists and others). No attempt has been made to quantify
either the economic or health impact of dengue on these groups, but it is
undoubtedly substantial and growing.
Risk factors
Risk factors for dengue and DHF are primarily those associated with the
potential to be fed upon by infected mosquitoes. Because the primary
vector feeds mostly during daylight hours, and is well adapted to the
urban environment, the socioeconomic conditions that favour the spread
of other vector-borne disease, such as malaria, are not as important with
dengue. Certainly lower-income neighbourhoods may be at increased risk
of infection, especially if litter and refuse that may retain sufficient water
to support mosquito larval development are abundant; however, this mosquito species breeds equally efficiently in flower pots and other ornamental
containers that hold water, and may be abundant in any area, including
middle-income or upper-income neighbourhoods, or in public places such
as schools and parks. This, coupled with its day-biting activity, means that
people may easily be exposed to dengue-infected mosquitoes during their
daily travels, even though their personal dwellings have screens and they
live in a nice neighbourhood.
A second major risk factor is prior infection with a dengue virus, since
heterologous anti-dengue antibodies may enhance dengue infection and
predispose an individual to much greater risk of DHF on subsequent infection. Similarly, maternally-acquired antibodies, as titres wane and are no
longer able fully to inactivate dengue virus, place infants at increased risk

Dengue and dengue haemorrhagic fever

235

of antibody-enhanced dengue infection leading to increased risk of DHF.

Other risk factors include the human population density, and thus the
availability of susceptible individuals to sustain transmission, residence in
tropical environments, where vector species survive best, and residence in
an area where infected travellers may introduce the virus and precipitate
epidemic transmission.
The risk of dying from dengue or DHF depends to a large extent on
the skills of local attending physicians and nurses. Properly trained and
experienced health care professionals can expertly manage critically ill
patients and reduce mortality rates to below 1 per cent; however, those not
familiar with the disease may find mortality rates of 5 per cent or greater
among their seriously ill patients.

Other diseases
Two other vector-borne diseases frequently occur under conditions similar
to those which support dengue virus transmission: yellow fever and Chikungunya fever. Yellow fever is well known as a historical scourge of the
tropics; thousand of individuals died, especially in the temperate and tropical zones of the Americas, often seriously limiting national development
projects, as was the case with the Panama Canal. What is often overlooked
today is the fact that the same conditions that originally supported urban
yellow fever transmission in the Americas are once again developing, with
widespread and abundant populations of vector mosquitoes, Aedes aegypti,
and large, susceptible human populations. All that is needed to initiate
urban transmission is the introduction of infected humans into a situation
where they may be fed upon by vector mosquitoes, and an outbreak could
ensue. Currently, active yellow fever transmission is limited to endemic
areas of sub-Saharan Africa and northern South America; however, with
the frequency of air travel to and from virtually all parts of the world, it is
not inconceivable that an individual could be infected in a remote endemic
area, and shortly thereafter be fed upon by potential vector mosquitoes in
a major urban centre far removed from the site of original infection.
Chikungunya is not as well known as yellow fever, but it too is transmitted by Aedes aegypti, and may cause massive outbreaks in urban settings.
This virus is not known to cause fatal human infection, but it does cause
a serious febrile illness similar to dengue fever, which may incapacitate
infected individuals for about a week. Large epidemics have been reported
in India, Indonesia and the Philippines, and the virus is known to be present in much of Africa.

Cost-effectiveness of interventions
At present, countries where dengue and DHF are endemic spend millions
of dollars on vector control. This includes direct health care costs for
treatment and indirect costs resulting from lost tourism and lost work.

236

Global Epidemiology of Infectious Diseases

The latter arises not solely in relation to infected individuals; since those
infected are often children, parents may have to stop working while they
care for their sick child. The average hospital stay for DHF patients is
5 to 10 days. Shepard & Halstead (1993) have summarized recent cost
estimates for dengue epidemics and found the total for direct health care
interventions, associated vector control activities, and indirect costs to be
from several million to over US$100 million per outbreak. Unfortunately,
the frequency of such outbreaks is increasing as new areas are infested
with Aedes aegypti.

Future research priorities

Substantial research efforts are required to manage the problem of dengue/
DHF. An immediate need exists for expanded medical training to teach
physicians and nurses the proper management of DHF patients, especially
in areas where the disease is just emerging and clinicians have little experience in managing these seriously ill patients. Significant reductions in
the mortality rates in endemic areas can be realized with minimal costs
through such efforts, and clearly this should be among the highest priority
activities in dengue control efforts. Likewise, support is required in the
clinical laboratory to ensure that dengue/DHF is properly and promptly
diagnosed. Again, this can be accomplished with relatively little investment
since common laboratory techniques are used, and all that may be required
is provision of the appropriate diagnostic reagents. Accurate diagnosis
should then feed into a well-organized surveillance and reporting system
that could allow health managers to monitor disease trends and initiate
outbreak control efforts at the first indication of a pending epidemic.
Several organizations are attempting to develop a vaccine for dengue/
DHF. A tetravalent vaccine would be especially beneficial. Since humans
represent the primary amplifying host of these viruses, a fully protective
vaccine that was widely administered could interrupt transmission and
significantly reduce the risk of infection. This is an especially challenging
task, however, since four distinct viruses may cause the disease, and the
confounding factor of immune-enhancement where there is partial protection, as would be elicited in a vaccine against only one, two or three
of the four dengue viruses, may in fact place individuals at greater risk of
DHF than unvaccinated persons. A live, attenuated, tetravalent vaccine
for dengue viruses has entered into Phase 2 clinical trials, and appears to
be safe and immunogenic.
At present, the only avenue for effective prevention of dengue/DHF is
through vector control. Unfortunately, sustained vector control is costly,
and as a consequence, frequently limited in success. Research is needed into
better ways of controlling Aedes aegypti populations. These must not only
be economically feasible, but also sustainable. Further, they need to take
into account the movement of large segments of the populations from rural
areas to urban centres in many affected countries. Current control efforts

Dengue and dengue haemorrhagic fever

237

are often directed towards routine pest management, and may include
community involvement coupled with emergency insecticide applications
when outbreaks are recognized. These efforts have met with marginal success, and it is clear that there is room for improvement. Historically, Aedes
aegypti was controlled by highly regimented mosquito control programmes
that involved both source reduction and effective use of insecticides (Soper
et al. 1943). Such programmes are, however, extremely expensive and difficult to maintain. Today, some countries have succeeded in controlling
vector mosquito populations using these techniques (for example, Cuba
after the 1981 outbreak), but in most cases countries have relied on insecticide applications after outbreaks are recognized, generally with much
less success. It is clear that innovative measures are needed, which will be
identified only through multidisciplinary research efforts that consider both
the sociological aspects of the local conditions leading to vector breeding,
and of the biology of the vector mosquito itself.
Equally challenging are the characteristics of the viruses that cause
dengue. Wide variations in the pathogenicity of dengue viruses have been
observed. Some strains of the same serotype may cause outbreaks with
little or no serious illness, while others may elicit significant numbers of
severe or fatal DHF cases. The molecular basis for this variation must be
determined. Fortunately, the complete nucleotide sequence is now available for strains of all four dengue serotypes, and infectious clones have
been produced for some serotypes. These tools should help in the future
definition of virulence factors of dengue viruses.

Conclusions
Dengue, and especially DHF/DSS, are significant global diseases that are
increasing in incidence and economic importance. Four distinct viruses
may cause classic dengue fever, and while infections with one serotype
leads to lifelong homologous protection, there is no lasting cross-protection from other dengue serotypes; indeed, subsequent dengue infections
are associated with increased risk of DHF/DSS. All four dengue viruses
are transmitted primarily by the urban mosquito vector, Aedes aegypti, a
species that is well adapted to life in close association with humans, and
one whose distribution and abundance is rapidly expanding throughout
tropical and subtropical cities, suburbs and villages of the world. Vector
control is expensive and inefficient, and as a result, conditions are prime
for epidemic dengue virus transmission in these areas. The result is that
today much of the worlds population is at risk of dengue, either as an
immediate threat in areas where the primary vector is abundant and dengue
viruses are endemic, or to travellers to these areas.
We have attempted to estimate the global burden of disease attributable to severe dengue and DHF/DSS, based on reported numbers of cases.
The greatest number of cases and deaths are thought to occur in India
and the Other Asia and Islands Region; however, the fastest growing

238

Global Epidemiology of Infectious Diseases

region in terms of dengue fever is the Latin America and the Caribbean
Region. Little information is available to document the burden of disease
for the Sub-Saharan Africa Region. These analyses conclude that nearly
half a million cases of dengue sufficiently severe to reach formal reporting occur annually. Of these, approximately 20 000, or about 4 per cent,
end in death. These values clearly underestimate the true impact of severe
dengue because:
they are based on reported cases only, and this information is fragmen-

tary and incomplete;

they are based on annual estimates averaged over a decade and thus

minimize the impact of the increasing incidence of DHF/DSS in some

regions;
in some regions, little information is available on which to base any

estimate.

References
Akiyama J (1993) Review of multicentre project to control DHF vectors by community participation in WHO, SEARO. Tropical Medicine, 35:277283.
Bosio CF et al. (1992) Variation in the efficiency of vertical transmission of dengue1 virus by strains of Aedes albopictus (Diptera: Culicadae). Journal of Medical
Entomology, 29:985989.
Clark DH, Casals J (1965) Arboviruses, Group B. In: Horsfall FL, Tamm I, eds. Viral
and rickettsial infections of man. Philadelphia, Lippincott, pp. 606658.
Cordellier R et al. (1983) Circulation silvatique du virus Dengue 2 en 1980, dans
les savannes sub-soudaniennes du Cte dIvoire [Silvian circulation of the Dengue 2 virus in 1980 in the savannahs of Cte d'Ivoire]. Cahier ORSTOM Srie
Entomologie Mdicale Et Parasitologie, 21:165172.
Craven RB (1991) Flaviviruses. In: Belshe RB ed. Textbook of human virology, 2nd
ed., St. Louis, MO, Mosby-Year Book, Inc. pp. 633662.
de Souza M, Freier JE (1991) Vertical transmission of dengue 1 virus by Haemagogus
equinus mosquitoes. Journal of the American Mosquito Control Association, 7:
11820.
Ehrenkranz NJ et al. (1971) Pandemic dengue in Caribbean countries and the southern
United States: past, present and potential problems. New England Journal of
Medicine, 285:14601469.
Fagbami AH, Monath TP, Fabiyi A (1977) Dengue virus infections in Nigeria: a
survey for antibodies in monkeys and humans. Transactions of the Royal Society
of Tropical Medicine and Hygiene, 71:6065.
Fauran P, Laille M, Moreau JP (1990) Etude sur la transmission verticale des virus
de la dengue dans le Pacifique Sud [Study on the vertical transmission of the
dengue virus in the South Pacific]. Bulletin de la Socit de Pathologie Exotique,
83:311316.

Dengue and dengue haemorrhagic fever

239

Freier JE, Rosen L (1987) Vertical transmission of dengue viruses by mosquitoes of

the Aedes scutellaris group. American Journal of Tropical Medicine and Hygiene,
37:640647.
Freier JE, Rosen L (1988) Vertical transmission of dengue viruses by Aedes mediovittatus. American Journal of Tropical Medicine and Hygiene, 39:218222.
Gratz N (1993a) Lessons of Aedes aegypti control in Thailand. Medical and Veterinary Entomology, 7:110.
Gratz N (1993b) What must we do to effectively control Aedes aegypti? Tropical
Medicine, 35: 243251.
Gubler DJ (1993) Emergent and resurgent arboviral diseases as Public Health Problems. Concepts in Virology, 24:257273.
Halstead SB (1970) Observations related to the pathogenesis of dengue hemorrhagic
fever: VI hypotheses and discussion. Yale Journal of Biology and Medicine, 42:
350362.
Halstead SB (1980) Immunological parameters of togavirus disease syndromes. In:
Schlesinger RW, ed. The Togaviruses. New York, Academic Press, pp.107174.
Halstead SB (1992) The XXth century dengue pandemic: need for surveillance and
research. World Health Statistics Quarterly, 45:292298.
Halstead SB (1993a) Dengue virology: fifty years of problems and progress. Concepts
in Virology, 25:275281.
Halstead SB (1993b) Global epidemiology of dengue: health systems in disarray.
Tropical Medicine, 35:137146.
Halstead SB, Nimmannitya S, Margiotta MR (1969) Dengue and chikungunya virus
infection in man in Thailand, 19621964: II observations on disease in outpatients.
American Journal of Tropical Medicine and Hygiene, 18:972983.
Hammon WMcD et al. (1960) New hemorrhagic fevers of children in the Philippines and Thailand. Transactions of the Association of American Physicians,
73:140155.
Hammon WMcD, Rudnick A, Sather GE (1960) Viruses associated with epidemic
hemorrhagic fevers of the Philippines and Thailand. Science, 131:11021103.
Hare GE (1898) The 1897 epidemic of dengue in North Queensland. Australasian
Medical Gazette, 17:98107.
Henchal EA et al. (1986) Identification of an antigenic and genetic variant of
dengue 4 virus from the Caribbean. American Journal of Tropical Medicine
and Hygiene, 35:393400.
Hull B et al. (1984) Natural transovarial transmission of dengue 4 virus in Aedes
aegypti in Trinidad. American Journal of Tropical Medicine and Hygiene, 33:
12481250.
Itoh T (1993) Control of DF/DHF vector, Aedes Mosquito, with insecticides.
Tropical Medicine, 35:259267.
Kerschner JH et al. (1986) Genetic and epidemiologic studies of dengue type 2
viruses by hybridization using synthetic deoxyoligonucleotides as probes. Journal
of General Virology, 67:26452661.

240

Global Epidemiology of Infectious Diseases

Khin MM, Than KA (1983) Transovarial transmission of dengue 2 virus by Aedes

aegypti in nature. American Journal of Tropical Medicine and Hygiene, 32:
590594.
Kliks SC et al. (1989) Antibody-dependent enhancement of dengue virus growth
in human monocytes as a risk factor for dengue hemorrhagic fever. American
Journal of Tropical Medicine and Hygiene, 40:444451.
Kojima S (1993) A rapporteurs summary: vector control. Tropical Medicine, 35:
331334.
Kouri G, Guzman MG, Bravo J (1986) Haemorrhagic dengue in Cuba: history of
an epidemic. Bulletin of the Pan American Health Organization, 20:2430.
Kuberski T et al. (1977) Clinical and laboratory observations on patients with primary and secondary dengue type 1 infections with hemorrhagic manifestations
in Fiji. American Journal of Tropical Medicine and Hygiene, 26:775783.
Lam SK (1993) A rapporteurs summary: health system. Tropical Medicine, 35:335.
Lopez-Correa RH et al. (1979) The 1977 dengue epidemic in Puerto Rico: epidemiologic and clinical observations. In: Dengue in the Caribbean, 1977.
Washington, DC, Pan American Health Organization (Scientific Publication,
No. 375, pp. 6067).
Mitchell CJ, Miller BR (1990) Vertical transmission of dengue viruses by strains
of Aedes albopictus recently introduced into Brazil. Journal of the American
Mosquito Control Association, 6:251253.
Monath TP et al. (1974) Arbovirus studies in Nupeko Forest, a possible natural
focus of yellow fever virus in Nigeria: I. description of the area and serological
survey of humans and other vertebrate hosts. Transactions of the Royal Society
of Tropical Medicine and Hygiene, 68:3038.
Okabe N (1993) Situation on dengue fever and dengue haemorrhagic fever in the
western Pacific region. Tropical Medicine, 35:147160.
Pan American Health Organization (1982) Dengue laboratory diagnostic procedures, Washington, DC, Pan American Health Organization.
Pan American Health Organization (1993) Dengue in the Americas: an update.
Epidemiological Bulletin, 14:13.
Pan American Health Organization (1994) Dengue and dengue haemorrhagic
fever in the Americas: guidelines for prevention and control. Washington, DC,
Pan American Health Organization.
Quintos FN et al. (1954) Hemorrhagic fever observed among children in the
Philippines. Philippine Journal of Pediatrics, 3:19.
Ramirez-Ronda CH et al. (1979) Dengue hemorrhagic shock in the western hemisphere. Tropical and Geographical Medicine, 31:127131.
Repik PM et al. (1983) RNA finger-printing as a method for distinguishing dengue 1 virus strains. American Journal of Tropical Medicine and Hygiene, 32:
577589.
Roche JC et al. (1983) Isolement de 96 souches de virus dengue 2 partir de
moustiques capturs en Cte dIvoire et Haute-Volta [Isolation of 96 samples

Dengue and dengue haemorrhagic fever

241

of the dengue 2 virus from mosquitoes captured in Cte dIvoire and HauteVolta]. Annales de Virologie, 134E:233238.
Rosen L (1987) Mechanism of vertical transmission of the dengue virus in mosquitoes. Comptes Rendus de lAcadmie des Sciences Srie III, Sciences de la
Vie, 304:34750.
Rosen L (1988) Further observations on the mechanism of vertical transmission
of flaviviruses by Aedes mosquitoes. American Journal of Tropical Medicine
and Hygiene, 39:123126.
Rosen L et al. (1983) Transovarial transmission of dengue viruses by mosquitoes:
Aedes albopictu and Aedes aegypti. American Journal of Tropical Medicine
and Hygiene, 32:11081119.
Rudnick A (1965) Studies of the ecology of dengue in Malyasia: a preliminary
report. Journal of Medical Entomology, 2:203208.
Rudnick A (1984) The ecology of the dengue virus complex in Peninsular Malaysia.
In: Pang T and Pathmanathan R, eds. Proceedings of the International Conference
on Dengue/DHF, Kuala Lumpur, University of Malyasia Press, pp. 7.
Rush B (1789) An account of the bilious remitting fever as it appeared in Philadelphia in the summer and autumn of the year 1780. In: Medical inquiries and
observations, Philadelphia, pp.89100.
Russell PK, McCown JM (1972) Comparison of dengue-2 and dengue-3 virus
strains by neutralization tests and identification of subtype of dengue-3. American Journal of Tropical Medicine and Hygiene, 21:9799.
Russell PK, Brandt WE, Dalrymple JM (1980) Chemical and antigenic structure
of flaviviruses. In: Schlesinger RW, ed. The Togaviruses, New York, Academic
Press, pp. 503530.
San Juan Laboratories (1979) 1979 annual report. San Juan, Puerto Rico, Centers
for Disease Control.
Serufo JC et al. (1993) Isolation of dengue virus type 1 from larvae of Aedes
albopictus in Campos Altos City, State of Minas Gerais, Brazil. Memorias do
Instituto Oswaldo Cruz, 88:503-504.
Shepard DS, Halstead SB (1993) Dengue (with notes on yellow fever and Japanese
encephalitis). In: Jamison DT et al., eds. Disease control priorities in developing
countries, New York, Oxford University Press, 303320.
Shroyer DA (1990) Vertical maintenance of dengue-1 virus in sequential generations
of Aedes albopictus. Journal of the American Mosquito Control Association,
6:312314.
Soper FL et al. (1943) The organization of permanent nation-wide anti-Aedes
aegypti measures in Brazil, New York, Rockefeller Foundation.
Tsai CJ et al. (1991) Upper gastrointestinal bleeding in dengue fever. American
Journal of Gastroenterology, 86:3335.
Watts DM et al. (1987) Effect of temperature on the vector efficiency of Aedes
aegypti for dengue 2 virus. American Journal of Tropical Medicine and Hygiene,
36:143152.

242

Global Epidemiology of Infectious Diseases

World Health Organization (1986) Dengue haemorrhagic fever: diagnosis treatment and control, Geneva, World Health Organization.
World Health Organization (1993) Dengue epidemic. Weekly Epidemiological
Record, 68:191.
World Health Organization (1994a) Dengue and dengue haemorrhagic fever.
Weekly Epidemiological Record, 69:265266.
World Health Organization (1994b) Dengue and dengue haemorrhagic fever.
Weekly Epidemiological Record, 69:237239.
World Health Organization (1994c) International conference on dengue haemorrhagic fever. Geneva, World Health Organization (WHO/CDS/MIM/94.1).