Am. J. Trop. Med. Hyg., 74(4), 2006, pp.

684–691
Copyright © 2006 by The American Society of Tropical Medicine and Hygiene

VOLUME REPLACEMENT IN INFANTS WITH DENGUE HEMORRHAGIC

FEVER/DENGUE SHOCK SYNDROME
NGUYEN THANH HUNG,* NGUYEN TRONG LAN, HUAN-YAO LEI, YEE-SHIN LIN, LE BICH LIEN,
KAO-JEAN HUANG, CHIOU-FENG LIN, DO QUANG HA, VU THI QUE HUONG, LAM THI MY, TRAI-MING YEH,
JYH-HSIUNG HUANG, CHING-CHUAN LIU, AND SCOTT B. HALSTEAD
Department of Dengue Hemorrhagic Fever, Children’s Hospital No. 1, Ho Chi Minh City, Vietnam; Department of Microbiology and
Immunology, Department of Medical Technology, and Department of Pediatrics, College of Medicine, National Cheng Kung
University, Tainan, Taiwan, Republic of China; Arbovirus Laboratory, Pasteur Institute, Ho Chi Minh City, Vietnam; Department of
Pediatrics, University of Medicine and Pharmacy, Ho Chi Minh City, Vietnam; Division of Research and Diagnosis, Department of
Health, Center for Disease Control, Taipei, Taiwan, Republic of China; Uniformed Services University of the Health Sciences,
Bethesda, Maryland

Abstract. Volume replacement was studied prospectively in 208 infants with dengue hemorrhagic fever/dengue
shock syndrome (DHF/DSS). The mean volume of intravenous fluid used was 110.4 mL/kg administered over a mean
period of 25.8 hours. The mean volumes of intravenous fluid replacement in infants with DSS was significantly higher
than in those with non-shock DHF (129.8 mL/kg versus 102.1 mL/kg; P ⳱ 0.001). Patients with DSS had significantly
higher proportional requirements for dextran and blood transfusions than non-shock infants. Recurrent shock, pro-
longed shock, and acute respiratory failure were recorded in 8, 6, and 13 patients, respectively. Four patients with DSS
died of severe complications. Intravenous fluid replacement with special care to avoid fluid overload requires careful
attention to established indications for use of colloidal solutions and blood transfusions. To improve case fatality rates,
special efforts need to be directed to infants with DHF/DSS accompanied by severe complications.

INTRODUCTION and effective replacement of plasma loss with crystalloid, col-

Dengue infections result in a disease continuum that in-
cludes syndromes varying in severity and prognosis. These
include dengue fever (DF), dengue hemorrhagic fever
(DHF), and dengue shock syndrome (DSS), the most severe
form of DHF. All are caused by dengue viruses that belong to
the family Flaviviridae as four serotypes (DEN-1, DEN-2,
DEN-3 and DEN-4).1,2 In the past 15 years, there has been a
dramatic increase in the global incidence of dengue and DHF/
DSS. More than 2.5 billion people are now at risk in more
than 100 countries worldwide, and every year approximately
50 million infections occur, including 500,000 cases of DHF
and DSS.3 Nearly 95% of cases are among children less than
15 years of age, while ⱖ 5% of all DHF/DSS cases occur in
infants. In contrast to DHF/DSS cases in children older than
one year of age, which occur during secondary dengue virus
infections, almost all of DHF/DSS cases in infants occur dur-
ing primary dengue virus infections.4,5
The main hallmark that differentiates DHF from DF is
clinically significant vascular permeability usually between
the third and the sixth days of illness, which results in plasma
leakage from the intravascular compartment to extravascular
spaces.2 In less severe cases (non-shock DHF grades I and II),
plasma leakage is mild to moderate, and many patients re-
cover spontaneously or shortly after administration of intra-
venous fluid. In more severe cases (DSS grades III and IV),
there are large plasma losses, hypovolemic shock ensues, and
it can progress rapidly to profound shock. The patient in
shock may die within 12–24 hours if appropriate treatment is
not promptly administered. Volume replacement is the main-
stay for treatment of DHF/DSS.2,6 The current World Health
Organization (WHO) guidelines for volume replacement
were based on the studies of DHF/DSS in children.2 Early
* Address correspondence to Nguyen Thanh Hung, Department of
Dengue Hemorrhagic Fever, Children’s Hospital No. 1, 341 Su Van
Hanh Street, District 10, Ho Chi Minh City, Vietnam. E-mail:
hungdhf@hcm.fpt.vn
loidal solutions results in favorable outcome in most pa-
tients.2,6 In a recent randomized double-blind comparison of
four intravenous fluids (dextran, gelatin, Ringer’s lactate
[RL], and saline) for initial resuscitation of 230 Vietnamese
children older than one year of age with DSS, there was no
clear advantage to using any of the four fluids. The most
significant factor determining clinical response was the pulse
pressure at presentation.7
Case fatality rates of DHF/DSS in children of < 1% to 5%
have been reported from centers experienced in fluid resus-
citation.6–9 DHF/DSS is less common in infancy but when it
does occur the risk of dying is higher than in older chil-
dren.10,11 Volume replacement in infants with DHF/DSS is a
challenging management problem. There are few published
studies on this topic. In attempt to support evidence-based
management of DHF/DSS infants with the goal of reducing
mortality in this group, we present our experience with vol-
ume replacement in infants with DHF/DSS.
MATERIALS AND METHODS
Study design. This was part of a prospective study on DHF
in infants that had focused on clinical and immunologic as-
pects and management.5,12 The study period was from August
1997 to December 2002 at the Department of Dengue Hem-
orrhagic Fever of Children’s Hospital No. 1 in Ho Chi Minh
City, a tertiary referral pediatric hospital in southern Viet-
nam. The Department of Dengue Hemorrhagic Fever has 85
in-patient beds and is staffed with 10 doctors and 25 nurses.
Patients. Two hundred seventy-two infants less than 12
months of age admitted to the Department of Dengue Hem-
orrhagic Fever with clinical diagnosis of DHF according to
the 1997 criteria of the WHO were recruited into the study
after parental or guardian informed consent was obtained.2
Dengue virus infections in the patients were studied by 1)
viral envelope and membrane (E/M)-specific capture IgM en-
zyme-linked immunosorbent assay (ELISA) and/or nonstruc-
tural protein 1 serotype-specific IgG ELISA at the Center for

684
 VOLUME REPLACEMENT IN INFANTS WITH DENGUE
Disease Control, Department of Health in Taipei, Taiwan13
or 2) capture IgM ELISA at the Pasteur Institute in Ho Chi
Minh City, Vietnam.14
Of 272 infants hospitalized with DHF a positive IgM-
capture ELISA result was obtained in 245 infants. Among
these infants 182 had non-shock DHF (grade I, 1 infant; grade
II, 181 infants) and 63 had DSS (grade III, 54 infants; grade
IV, 9 infants). The nutritional status and clinical and immu-
nologic aspects of these infants have been previously re-
ported.5,12 Serologic testing showed that almost all of the pa-
tients (95.3%) had primary dengue virus infections.5 Intrave-
nous fluid therapy was indicated in 208 infants: 145 infants
with non-shock DHF and 63 infants with DSS. We focused on
fluid management in these 208 infants. Ethical approval of the
study was obtained from the Scientific and Ethical Committee
of Children’s Hospital No. 1 in Ho Chi Minh City.
Investigations and observations. A full history, physical ex-
amination findings, management, and subsequent progress
were recorded on a standard data form for each patient. For
infants receiving intravenous fluid therapy, a fluid balance
sheet was used to record the type, rate, and quantities of fluid
administered, and to calculate the volume of intravenous fluid
per kilogram of body weight given per 24 hours.
On admission, a venous blood sample was obtained for
determination of hematocrit and complete blood cell count,
and a serum sample was stored for virus isolation or serologic
confirmation of dengue virus infection, in conjunction with a
second sample collected on the day of discharge. Complete
blood cell counts (H-2000 counter; Careside Inc., Culver City,
CA) were determined in the hospital laboratory. Chest radio-
graphic and/or ultrasound examinations were conducted as
clinically required.5
Management. Treatment of non-shock DHF was support-
ive and symptomatic. Management of DHF/DSS patients in
the study principally followed the current WHO guidelines
for volume replacement with some adaptations.2 Intravenous
fluid therapy was indicated when the patient had one or more
of the signs/symptoms: repeated vomiting, rapid liver enlarge-
ment, hematemesis, melena, lethargy, a high degree of hemo-
concentration, and a rapidly rising hematocrit. Ringer’s lac-
tate with or without 5% dextrose was started at a rate of 6–7
mL/kg of body weight/hour, then adjusted according to the
patient’s clinical condition, vital signs, hematocrit, and urine
output. For patients with DSS, plasma losses were immedi-
ately replaced with electrolyte or, in case of profound shock,
colloidal solutions. Use of intravenous fluid was continued to
replace further plasma losses to maintain effective circulation
for 24–48 hours. For patients with DSS grade III, RL was
started at a rate of 15–20 mL/kg/hour, while those with DSS
grade IV (profound shock) received RL at a rate of 20 mL/kg
over a 15-minute period, followed by colloidal solution (dex-
tran 40 or dextran 70) at a rate of 10–20 mL/kg/hour. Pulse,
blood pressure (BP), and respiratory rate were recorded ev-
ery 15–30 minutes until shock was overcome. When the con-
dition of the patient did not improve with RL infusion (vital
signs were still unstable [shock persists]), dextran was used at
a rate of 15–20 mL/kg/hour. A blood transfusion was only
indicated in patients with severe bleeding. Infants receiving
intravenous fluid therapy were closely observed around the
clock until it was certain that danger has passed. Frequent
recording of vital signs and hematocrit was important for
evaluating treatment results. Hematocrit was determined ev-
685
ery two to four hours and thereafter every four hours until
stable. A fluid balance sheet was used to record the type, rate,
and quantity of fluid administered (input), and to calculate
the amount of intravenous fluid per kilogram of body weight
given per 24 hours to determine whether there had been suf-
ficient volume replacement and to avoid fluid overload. The
frequency and volume of urine (output) were also recorded.
Intravenous fluid therapy was stopped when the patient’s
condition had been stable for more than 24 hours, or there
was any sign/symptom of fluid overload. In the convalescent
phase, usually between the seventh and the ninth days of
illness, some patients had signs of fluid overload because of
excessive intravenous therapy and reabsorption of extrava-
sated plasma from the interstitial compartment. For those
patients who developed symptomatic fluid overload, furose-
mide was given (0.5–1 mg/kg/dose).
Data and statistical analysis. The outcome measures were
the total volume of intravenous fluid infused; the requirement
for dextran; blood transfusion; duration of intravenous fluid
therapy; development of any complications related to fluid
therapy, such as recurrent shock, prolonged shock, and respi-
ratory failure; and death. Other complications of DHF/DSS
such as dengue encephalopathy, gastrointestinal (GI) bleed-
ing, and associated pneumonia were also noted. These out-
come measures were compared between two groups of the
patients (non-shock DHF versus DSS). An infant had devel-
oped recurrent shock if he or she had tachycardia, coolness of
the extremities, and a decrease in BP to ⱕ 20 mm of Hg while
BP had previously reached ⱖ 30 mm of Hg even though they
had received adequate fluid according to the guidelines for
volume replacement.7 Prolonged shock was when an infant in
shock did not improve after receiving ⱖ 60 mL/kg of intra-
venous fluid or the patient was still in shock after ⱖ 6 hours
of intravenous fluid therapy. Dengue encephalopathy was
when an infant with DHF/DSS had mental disturbances (leth-
argy, stupor, restlessness, coma); neurologic signs and symp-
toms (headache, vomiting, convulsions, hemiparalysis, tetra-
paralysis, hypereflexia); and cerebrospinal fluid examination
with normal levels of protein, glucose, negative gram stain
and bacterial culture. Coinfection with bacterial pneumonia
was when a patient had cough, tachypnea, retractions, the
presence of crackles, and suggestive radiologic findings (scat-
tered infiltrates to dense lobar pneumonia).
The statistical significance of differences in normally dis-
tributed data was compared using analysis of variance. The
Kruskal-Wallis test for two groups was used when the vari-
ances in the samples differed. Differences between propor-
tions were tested by the chi-square test or Fisher’s exact test.
Statistical analyses were performed with Epi-Info 2000, ver-
sion 1.1 (Centers for Disease Control and Prevention, At-
lanta, GA). Multivariate logistic regression analysis was per-
formed with SPSS statistical package version 12.0 for Win-
dows (SPSS Inc., Chicago, IL). A P value < 0.05 was
considered statistically significant.
RESULTS
Clinical findings of 208 DHF/DSS patients. Clinical and
laboratory findings on patients are shown in Table 1. The
mean age of the patients was 6.7 months (range ⳱ 1–11). The
mean age of infants with DSS was significantly higher than
that of infants with non-shock DHF (7.3 versus 6.5 months;
686 HUNG AND OTHERS

TABLE 1
Demographic characteristics and clinical and laboratory findings for 145 infants with nonshock dengue hemorrhagic fever (DHF) and 63 infants
with dengue shock syndrome (DSS)*

All infants Infants with nonshock DHF Infants with DSS

Variable (n ⳱ 208) (n ⳱ 145) (n ⳱ 63) P†

Clinical findings
Boys/girls 119/89 79/66 40/23
Age, mean months ± SD (range) 6.7 ± 2.4 (1–11) 6.5 ± 2.5 (1–11) 7.3 ± 2.2 (3–11) 0.03
Date of admission from the onset of 4.1 ± 1.1 (1–7) 4.0 ± 1.1 (1–7) 4.3 ± 1.0 (2–6) 0.07
fever, mean days ± SD (range)
Peak temperature, mean °C ± SD 38.9 ± 0.7 (38–41) 39.0 ± 0.7 (38–41) 38.8 ± 0.8 (38–41) 0.07
(range)
Duration of fever, mean days ± SD 5.2 ± 1.8 (2–13) 5.2 ± 1.7 (2–13) 5.3 ± 2.0 (2–12) 0.7
(range)
Bleeding manifestations, no. (%) of
infants
Petechiae 208 (100) 145 (100) 63 (100)
Gastrointestinal bleeding 14 (16.7) 8 (5.5) 6 (9.5) 0.4‡
Liver size,§ mean cm ± SD (range) 3.1 ± 1.0 (1–6) 2.8 ± 0.9 (1–6) 3.6 ± 1.0 (2–6) 0.001
Dengue encephalopathy, no. (%) of 16 (7.7) 8 (5.5) 8 (12.7) 0.1‡
infants
Laboratory findings
Peak hematocrit, mean % ± SD 40.5 ± 4.5 (30–60) 39.2 ± 3.6 (30–60) 43.6 ± 4.8 (35–57) 0.001
(range)
Date of peak hematocrit,¶ mean 4.7 ± 1.0 (3–7) 4.8 ± 1.0 (3–7) 4.7 ± 1.0 (3–7) 0.7
days ± SD (range)
Increase in hematocrit, mean % ± 36.0 ± 14.0 (10–68) 32.4 ± 12.7 (10–68) 44.1 ± 13.6 (21–68) 0.001
SD (range)
Lowest platelet count, mean cells × 61.5 ± 32.3 (14–190) 64.3 ± 33.3 (14–190) 54.9 ± 29.2 (20–129) 0.03
103/mm3 ± SD (range)
Date of lowest platelet count,¶ 4.9 ± 0.9 (3–8) 4.9 ± 0.9 (3–8) 4.9 ± 1.0 (3–7) 0.8
mean days ± SD (range)
WBC count, mean cells/mm3 ± SD 6,787.6 ± 3,047.2 7,211.2 ± 3,173.2 5,719.5 ± 2,451.7 0.03
(range) (2,600–15,300) (2,800–15,300) (2,600–12,900)
* WBC ⳱ white blood cell.
† Nonshock DHF group vs. DSS group by Kruskal-Wallis test.
‡ By chi-square ␹2 test.
§ Below right costal margin.
¶ Days from the onset of fever.

P ⳱ 0.003). There were no significant differences in the date
of admission from the onset of fever, duration of fever, and
peak temperature between non-shock DHF and DSS groups
(mean ⳱ 4.0 days versus 4.3 days [P ⳱ 0.07]; 5.2 days versus
5.3 days [P ⳱ 0.7]; and 39.0°C versus 38.8°C [P ⳱ 0.7], re-
spectively, Table 1). Compared with infants with non-shock
DHF, infants with DSS had a larger liver size below right
costal margin (mean ⳱ 3.6 cm versus 2.8 cm [P ⳱ 0.001]).
Gastrointestinal bleeding and dengue encephalopathy were
noted in 14 (5.7%) and 16 (7.7%) patients, respectively. The
proportions of GI bleeding and dengue encephalopathy were
similar between infants in the non-shock DHF and DSS
groups (P ⳱ 0.4, and P ⳱ 0.1; Table 1).
Pulse rate in the critical phase in patients with DSS was
faster than that in patients with non-shock DHF (mean ⳱ 142
versus 136 beat/minute [P ⳱ 0.006]; Table 2). Systolic BP in
the critical phase of patients with DSS was significantly lower
than that of patients with non-shock DHF, while diastolic BP
was similar in the two groups (mean ⳱ 83.7 versus 92.4 mm of
Hg [P ⳱ 0.001], and 62.4 versus 59.3 mm of Hg [P ⳱ 0.06],
respectively, Table 2). Pulse rate, systolic BP, and diastolic BP
in the convalescent phase in both two groups returned to the
normal ranges (Table 2). Compared with infants with non-
shock DHF, infants with DSS had a significantly higher re-
spiratory rate in the acute and convalescent phases of illness
(mean ⳱ 46.0 versus 39.7 breaths/minute [P ⳱ 0.001] and
48.9 versus 42.5 breaths/minutes [P ⳱ 0.001]; Table 2).
Laboratory findings. The mean peak hematocrit and in-
crease in hematocrit in all patients was 40.5% (range ⳱ 30–
60%) and 36% (range ⳱ 10–68%), respectively. Peak hemat-
ocrit level and increase in hematocrit were significantly higher
in DSS group than that in non-shock DHF group (mean ⳱
43.6% versus 39.2% [P ⳱ 0.001] and 44.1% versus 32.4%
[P ⳱ 0.001], respectively, Table 1). The mean time to reach
the peak hematocrit level was 4.7 days (range ⳱ 3–7). The
mean lowest platelet count of all patients was 61.5 × 103 cells/
mm3 (range ⳱ 14–190 × 103 cells/mm3), and infants with DSS
had a significantly lower mean platelet count compared with
those with non-shock DHF (mean ⳱ 54.9 × 103 versus 64.3 ×
103 cells/mm3 [P ⳱ 0.03]). The mean time for platelets to
reduce to the lowest level in all patients was 4.9 days (range
⳱ 3–8).
Intravenous fluid replacement. The intravenous fluid
therapy and clinical outcome for 208 infants with DHF/DSS
in the study are summarized in Figure 1. The mean volume of
intravenous fluid used was 110.4 mL/kg (range ⳱ 27.5–243
mL/kg) administered over a mean period of 25.8 hours (range
⳱ 6–72). The mean volume of intravenous fluid replacement
in infants with DSS was significantly higher than that in those
with non-shock DHF (129.8 versus 102.1 mL/kg [P ⳱ 0.001]).
The mean duration of intravenous fluid administration and
the time to normalize the hematocrit level from the start of
intravenous fluid therapy for both the non-shock DHF and
DSS groups did not differ significantly (mean ⳱ 25.9 versus
25.7 hours [P ⳱ 0.5] and 29.9 versus 30 hours [P ⳱ 0.9]; Ta-
ble 3).
 VOLUME REPLACEMENT IN INFANTS WITH DENGUE 687

TABLE 2
Vital signs for 145 infants with nonshock dengue hemorrhagic fever (DHF) and 63 infants with dengue shock syndrome (DSS)*

Infants with
All infants nonshock DHF Infants with DSS
Variable (n ⳱ 208) (n ⳱ 145) (n ⳱ 63) P

In critical phase
Pulse rate,† mean beats/min ± SD 137.6 ± 14.1 136.0 ± 13.4 142.0 ± 15.1 0.006,‡ 0.001,§
(range) (120–210) (120–210) (120–180) 0.0001¶
Systolic BP,† mean mm of Hg ± SD 90.1 ± 9.8 92.4 ± 8.2 83.7 ± 10.8 0.001,‡ 0.3,§
(range) (60–120) (80–120) (60–110) 0.0001¶
Diastolic BP,† mean mm of Hg ± SD 60.1 ± 9.8 59.3 ± 8.2 62.4 ± 12.9 0.06,‡, 0.8,§
(range) (40–95) (40–90) (40–95) 0.09¶
Respiratory rate, mean breaths/min ± SD 40.9 ± 10.2 39.7 ± 9.7 46.0 ± 11.0 0.001,‡ 0.01,§
(range) (12–68) (12–68) (28–64) 0.2¶
In convalescent phase
Pulse rate,# mean beat/min ± SD 117.7 ± 7.7 118.4 ± 6.8 116.1 ± 9.4 0.1‡
(range) (90–140) (90–140) (90–140)
Systolic BP,# mean mm of Hg ± SD 92.1 ± 5.9 91.7 ± 5.7 93.0 ± 7.5 0.1‡
range (80–110) (80–110) (80–110)
Diastolic BP,# mean mm of Hg ± SD 59.2 ± 5.2 59.2 ± 4.2 59.1 ± 7.2 0.6‡
(range) (40–80) (50–80) (40–80)
Respiratory rate,# mean breaths/min ± SD 45.1 ± 12.7 42.5 ± 9.6 48.9 ± 15.7 0.001‡
(range) (30–98) (30–65) (36–98)
* BP ⳱ blood pressure.
† In 9 infants with DSS grade IV the pulse and blood pressure were not detectable; thus, n ⳱ 199 for all infants, n ⳱ 145 for nonshock DHF group, and n ⳱ 54 for DSS group.
‡ Nonshock DHF group vs. DSS group by Kruskal-Wallis test.
§ Mean value in critical phase vs. mean value in convalescent phase in nonshock DHF group by Kruskal-Wallis test.
¶ Mean value in critical phase vs. the mean value in convalescent phase in DSS group by Kruskal-Wallis test.
# Four infants with DSS grade III had fatal outcomes; thus, n ⳱ 204 for all infants, n ⳱ 145 for nonshock DHF group, and n ⳱ 59 for DSS group.

Administration of colloidal solution (dextran 40 or dextran
70) was indicated in 48 infants (23%). The proportion of in-
fants with DSS requiring dextran infusion (35 infants, 55.5%)
was significantly higher than that of infants with non-shock
DHF (13 infants, 8.9%) (P ⳱ 0.001). The mean volume of
dextran administered was 55.1 mL/kg (range ⳱ 13–119 mL/
kg); this was correlated with severity of the disease (P ⳱ 0.01;
Table 3). Twenty-eight infants (13.4%), five of whom had
severe GI bleeding, received a transfusion of fresh whole
blood (FWB; mean ⳱ 38.7 mL/kg, range ⳱ 12–140 mL/kg).
Compared with patients with non-shock DHF, patients with
DSS had a higher requirement for FWB (26.9% versus 7.5%
[P ⳱ 0.001]). However, the mean volume of FWB adminis-
FIGURE 1. Schematic diagram showing the intravenous fluid
therapy and clinical outcome for 208 infants with dengue hemor-
rhagic fever/dengue shock syndrome (DHF/DSS). RL ⳱ Ringer’s
lactate; FWB ⳱ fresh whole blood.
tered did not different between the groups (P ⳱ 0.1). Seven-
teen (8.1%) patients needed furosemide for symptoms of
fluid overload. The requirement for furosemide was related
to severity of the disease (4.8% in the non-shock DHF group
versus 15.8% in DSS group [P ⳱ 0.007]). In most of these
cases furosemide was used for fluid overload because of re-
absorption of extravasated plasma from the interstitial com-
partment in the convalescent phase of the illness.
For the subgroup of patients with DSS grade IV (nine
cases) admitted in the late stage of DSS, all patients required
infusion of dextran and intensive care with close monitoring.
However, there were no differences in the mean volume of
intravenous fluid administered; the volume of dextran, FWB,
or the duration of fluid replacement between the subgroups
of DSS grade IV and grade III (P ⳱ 0.5, P ⳱ 0.4, P ⳱ 0.5,
and P ⳱ 0.7, respectively, Table 4).
Complications related to intravenous fluid therapy. Recur-
rent shock was recorded in eight patients and prolonged
shock was noted in six patients. All eight patients with recur-
rent shock had a second episode of shock at a mean (SD) time
of 6.2 (2.9) hours (range ⳱ 3–11) from beginning of intrave-
nous fluid therapy. Each of these patients responded well to
dextran solution. Of six patients with prolonged shock, four
had DSS grade III and two had DSS grade IV. Two patients
with prolonged shock died. Patients with prolonged shock
required careful fluid management with catheters inserted to
monitor central venous pressure and correction of severe
metabolic acidosis. Nevertheless, there were no significant
differences in the total volume of intravenous fluid adminis-
tered, or of dextran, FWB, or the duration of fluid replace-
ment in the patients with prolonged shock compared with
other DSS patients (P ⳱ 0.5, P ⳱ 0.09, P ⳱ 0.6, and P ⳱ 0.6,
respectively, Table 4).
Thirteen patients (3 patients with non-shock DHF and 10
with DSS, 6.2%) developed acute respiratory failure between
day 4 and day 7 after onset of illness (day 4–5 ⳱ 4 patients;
688 HUNG AND OTHERS

TABLE 3
Intravenous fluid (iv) therapy for 145 infants with nonshock dengue hemorrhagic fever (DHF) and 63 infants with dengue shock syndrome (DSS)*

Infants with
All infants nonshock DHF Infants with DSS
Variable (n ⳱ 208) (n ⳱ 145) (n ⳱ 63) P

Total volume of iv fluid infused, mean mL/kg ± SD 110.4 ± 33.6 102.1 ± 28.4 129.8 ± 36.9 0.001†
(range) (27.5–243) (27.5–211.7) (50–243)
Requirement for dextran, no. (%) of infants 48 (23) 13 (8.9) 35 (55.5) 0.001‡
Volume of dextran, mean mL/kg ± SD 55.1 ± 25.9 39.4 ± 16.2 60.9 ± 26.5 0.01†
(range) (13–119) (13–65) (20–119)
Requirement for FWB, no. (%) of infants 28 (13.4) 11 (7.5) 17 (26.9) 0.001‡
Volume of FWB, mean mL/kg ± SD 38.7 ± 32.5 27.3 ± 18.2 44.7 ± 38.1 0.1†
(range) (12–140) (13–68) (12–140)
Duration of iv fluid administration, mean hours ± SD 25.8 ± 8.8 25.9 ± 8.1 25.7 ± 10.2 0.5†
(range) (6–72) (8–53) (6–72)
Time to normalize hematocrit, mean hours ± SD 29.9 ± 12.9 29.9 ± 13.2 30.0 ± 12.7 0.9†
(range) (12–72) (12–72) (18–72)
Requirement for furosemide, no. (%) of infants 17 (8.1) 7 (4.8) 10 (15.8) 0.007‡
Length of hospital stay, mean days ± SD 5.7 ± 2,6 5.7 ± 2.5 5.5 ± 2.7 0.3†
(range) (1–19) (2–16) (1–19)
* FWB ⳱ fresh whole blood.
† Nonshock DHF group vs. DSS group by Kruskal-Wallis test.
‡ Nonshock DHF group vs. DSS group by chi-square test.

day 6–7 ⳱ 9 patients). Among patients with acute respiratory
failure, fluid overload was diagnosed in two patients on day 5
and day 6 of illness. Fluid overload resulting from reabsorp-
tion of extravasated plasma from the interstitial compartment
in the convalescent phase (day 6–7) was considered the most
likely cause of respiratory failure in another five patients.
Respiratory failure was associated with prolonged shock, den-
gue encephalopathy, GI bleeding, and coinfection with pneu-
monia in five, three, three, and six patients, respectively. All
of these patients were treated with oxygen delivered by nasal
cannula (five patients) or nasal continuous positive airway
pressure (NCPAP, eight patients). Furosemide was used in
seven patients. Cefotaxim was prescribed for pneumonia.
Univariate analysis showed that factors related to acute re-
spiratory failure included a total volume of intravenous fluid
infused > 140 mL/kg, a requirement for dextran, prolonged
shock, GI bleeding, and coinfection with pneumonia (Table
5). Multivariate logistic analysis identified prolonged shock
and coinfection with pneumonia as significantly associated
with acute respiratory failure (P ⳱ 0.01 and P ⳱ 0.001, re-
spectively, Table 5).
Results of treatment. All but four patients recovered after
receiving fluid replacement and good nursing care. Of the
patients who died, two had prolonged shock, three had en-
cephalopathy, two had respiratory failure, and three had mas-
sive GI bleeding. Clinical characteristics and laboratory find-
ings for these four patients have been reported.5 Compared
with survivors, fatal cases had higher rates of severe compli-
cations such as prolonged shock, encephalopathy, respiratory
failure, and massive GI bleeding (P ⳱ 0.04, P ⳱ 0.001, P ⳱
0.02, and P ⳱ 0.001, respectively, Table 6). Thrombocytope-
nia in fatal patients was greater than that in surviving patients
(P ⳱ 0.007; Table 6).
DISCUSSION
We describe attributes of the clinical course and manage-
ment of DHF/DSS that occur in a unique immunologic group,
infants less than one year of age who experience severe dis-
ease during their initial dengue infection. Their only known
risk factor for severe dengue is the pre-natal acquisition of
maternal antibodies to dengue virus.15 As in older children,
the onset of DHF/DSS coincides with defervescence, a time
when T cells are actively terminating cellular infection and
generating cytokines.16 Virus infection, viremia, or cytokines

TABLE 4
Intravenous (iv) fluid therapy for subgroups of infants with dengue shock syndrome (DSS)*

Infants with Infants with DSS infants without DSS infants with
DSS grade III DSS grade IV prolonged shock prolonged shock
Variable (n ⳱ 54) (n ⳱ 9) (n ⳱ 57) (n ⳱ 6) P

Total volume of iv fluid infused, mean mL/kg ± SD 129.8 ± 35.1 128.7 ± 46.9 128.1 ± 35.5 144.2 ± 47.2 0.5,† 0.5‡
(range) (50–215) (73–243) (50–243) (94–215)
Requirement for dextran, no. (%) of infants 26 (48.1) 9 (100) 29 (50.8) 6 (100) 0.01,† 0.06†
Volume of dextran, mean 57.8 ± 25.6 66.6 ± 32.5 55.6 ± 24.1 80.5 ± 33.6 0.4,† 0.09‡
mL/kg ± SD (range) (20–105) (20–119) (20–100) (33–119)
Requirement for FWB, no. (%) of infants 14 (25.9) 3 (33.3) 13 (22.8) 4 (66.6) 0.9,† 0.3‡
Volume of FWB, mean 39.8 ± 31.7 67.3 ± 64.6 49.9 ± 42.3 27.7 ± 10.3 0.5,† 0.6‡
mL/kg ± SD (range) (12–131) (16–140) (12–140) (17–37)
Duration of iv fluid administration, 25.5 ± 8.5 27.1 ± 17.7 25.8 ± 10.4 24.1 ± 7.8 0.7,† 0.6‡
mean hours ± SD (range) (6–48) (8–72) (6–72) (14–34)
* FWB ⳱ fresh whole blood.
† DSS grade III vs. DSS grade IV by Kruskal-Wallis test for comparison of the means and chi-square test for comparison of the proportions.
‡ DSS infants with prolonged shock vs. DSS infants without prolonged shock by Kruskal-Wallis test for comparison of means and chi-square test for comparison of the proportions.
 VOLUME REPLACEMENT IN INFANTS WITH DENGUE 689

TABLE 5
Factors related to respiratory failure using univariate and multivariate analysis*

Univariate analysis Multivariate analysis

Odds Odds
Variable ratio 95% CI P† ratio 95% CI P

Total volume of iv fluid infused > 140 mL/kg 3.88 1.09–13.78 0.04 4.40 0.70–27.47 0.11
Requirement for dextran 9.00 2.63–30.75 0.001 1.61 0.19–13.57 0.66
Recurrent shock 0.73 0.03–7.59 1
Prolonged shock 52.00 5.03–537.11 0.001 25.91 1.92–348.19 0.01
Gastrointestinal bleeding 5.01 1.20–20.89 0.04 0.21 0.03–1.46 0.11
Dengue encephalopathy 4.20 1.02–17.16 0.06 1.57 0.10–24.78 0.74
Coinfected pneumonia 166.28 17.57–1573.96 0.001 315.36 15.78–6,301.15 0.001
* CI ⳱ confidence interval; iv ⳱ intravenous.
† By Fisher exact two-tailed test.

and chemokines are thought to damage endothelial cells and
platelets, which results in the opening of endothelial pores
and leads to the extravasation of fluid, hypovolemia, and in
some cases, shock. As is evident in this study, clinically sig-
nificant hypovolemia that required administration of intrave-
nous fluids is essential to the management of patients who
may not be in frank shock. We detail treatment regimens in
infants with evidence of vascular leakage with and without
shock and compare the physiologic status of infant DHF/DSS
with that of children greater than one year of age, a group
who acquire severe disease during a second dengue infection,
which has been extensively documented.
The percentage of non-shock DHF infants requiring intra-
venous fluid therapy in the present study was 79.6%. This is
higher than that of non-shock DHF in children ⱖ 1 year of
age who were admitted at the same time to the Department of
Dengue Hemorrhagic Fever (25–33%).9,17 According to the
WHO guidelines (1997), the amount of intravenous fluid is
usually equal to daily fluid maintenance plus a 5% deficit over
a 24–48-hour period. Daily fluid maintenance is calculated as
100 mL/kg for the first 10 kg of body weight, 50 mL/kg for the
second 10 kg of body weight, and 20 mL/kg for body weight
greater than 20 kg.2 Thus, an infant with a body weight of 10
kg will require 150 mL/kg of intravenous fluid. The estimate
for infants was higher than the amount actually administered
to non-shock DHF cases (mean ⳱ 102.1 mL/kg) over the
mean time of 25.9 hours. All non-shock DHF infants recov-
ered; only 13 patients (7.1%) required dextran with the mean
volume of 39.4 mL/kg, and 11 patients (6%) required FWB
transfusions with the mean volume of 27.3 mL/kg.
Volume replacement is the mainstay of treatment of DSS.
Early and effective replacement of lost plasma with electro-
lyte solution, plasma, or plasma expanders results in a favor-
able outcome. Nimmannitya18 reported 487 cases of DSS
treated at Bangkok Children’s Hospital. In these patients,
61% were successfully treated with crystalloid solution (Ring-
er’s lactate/acetate solution) and in 22% there was a need for
colloidal solution (dextran 40). Approximately 15% of the
shock cases had significant bleeding that required blood
transfusion and some in this group received other blood com-
ponents, e.g., concentrated platelets, fresh frozen plasma, and
cryoprecipitate.18 In another study, 44.6% of 240 DSS pa-
tients at the Department of Dengue Hemorrhagic Fever,
Children’s Hospital No.1 in Ho Chi Minh City needed dext-
ran 40 (Chi PB and Huyen NTD, unpublished data). In the
present study, all 63 DSS patients received intravenous infu-
sion, 27 (42.8%) patients received only RL, 35 (55.5%) pa-
tients received dextran plus RL, and 17 (26.9%) patients
needed FWB because of GI bleeding or internal bleeding.
Special care must be taken to manage intravenous fluids
administered to infants. Fluids account for a greater propor-
tion of body weight in infants than in children and minimum
daily requirements are correspondingly higher. Infants have
lower intracellular fluid reserves than older children and
adults. Moreover, capillary beds are intrinsically more per-
meable in infants than those in older children or adults. Both
early cardiovascular compromise and significant fluid over-
load are more likely to occur if capillary leaks occur in these
circumstances. All infants must be treated as high-risk pa-
tients who require early intervention with colloids, as in older
children with grade IV disease. In the present study, the mean
volume of intravenous fluid replacement and dextran in in-
fants with DSS was significantly higher than that given to
infants with non-shock DHF (129.8 versus 102.1 mL/kg [P ⳱

TABLE 6
Comparison of clinical and laboratory data between four infants with fatal dengue shock syndrome (DSS) and 204 surviving infants with dengue
hemorrhagic fever (DHF)/DSS

Infants with fatal DSS Survived infants with DHF/DSS

Variable (n ⳱ 4) (n ⳱ 204) P

Gastrointestinal bleeding, no. (%) of infants 3 (75) 11 (5.3) 0.001*

Liver size,† mean cm ± SD (range) 4.7 ± 0.9 (4–6) 3.0 ± 0.9 (1–6) 0.004‡
Respiratory failure, no. (%) of infants 2 (50) 11 (5.3) 0.02*
Dengue encephalopathy, no. (%) of infants 3 (75) 13 (6.2) 0.001*
Prolonged shock, no. (%) of infants 2 (50) 4 (6.7) 0.04*
Lowest platelet count, mean cells × 103/mm3 ± SD 25.5 ± 6.4 62.2 ± 32.2 0.007‡
(range) (20–32) (14–190)
* Infants with fatal DSS vs. surviving infants with DHF/DSS by Fisher exact two-tailed test.
† Below right costal margin.
‡ Infants with fatal DSS vs. surviving infants with DHF/DSS by Kruskal-Wallis test.
690 HUNG AND OTHERS
0.001] and 60.9 mL/kg versus 39.4 mL/kg [P ⳱ 0.01]). It is
important to detect early warning signs of shock in infants so
that intravenous fluid therapy can be started promptly.18
Prolonged shock was documented in six patients. Pro-
longed shock led to severe complications such as respiratory
failure, massive GI bleeding, metabolic acidosis, multiple or-
gan dysfunctions, and death.2 In the present study, patients
with prolonged shock required catheterization to monitor
central venous pressure and correction of the metabolic aci-
dosis.
The cause of acute respiratory failure in DHF patients is
usually caused by the administration of intravenous fluids too
rapidly or for too long a period. Pulmonary edema may occur
with the sudden healing of the capillary leak if administration
of intravenous fluid continues. With normal capillaries, the
body begins the process of readsorbing fluids from the extra-
cellular compartment. Lum and others described acute respi-
ratory distress syndrome in three patients with DHF with
prolonged shock and tissue hypoxia when crystalloids were
administered too rapidly.19 These patients became hypoxemic
and required positive pressure ventilation. Large pleural ef-
fusions and ascites may compress the lungs and limit the
movement of the diaphragm, resulting in respiratory distress.
The central nervous system may be depressed in patients with
dengue encephalopathy, which leads to dysfunction of the
respiratory center. Severe pneumonia caused by bacterial su-
perinfection may result in respiratory failure. In the present
study, pneumonia in two patients and fluid overload in one
patient resulted in respiratory failure in infants with non-
shock DHF. Multivariate logistic analysis showed that pro-
longed shock and coinfection with pneumonia had a strong
association with acute respiratory failure (P ⳱ 0.01 and P ⳱
0.001, respectively).
Acute respiratory failure in infants was treated with oxygen
by nasal cannula (in five patients), or NCPAP (in eight pa-
tients). Cam and others reported that NCPAP improved out-
come of respiratory failure in DHF/DSS patients.20 Severe
respiratory failure that failed to respond to respiratory sup-
port was observed in two out of four fatal cases.
Although DHF/DSS in infants comprises less than 5% of
all DHF/DSS cases, mortality rates are higher in infants than
in older children.5 In our study, four infants died on day 5 to
day 7 after the onset of fever, with the mean time of 29.7
hours after onset of shock. DHF/DSS in infants can quickly
result in death. In a hospital-based study in Jakarta, Indone-
sia, there were 188 DSS patients with overall mortality of
19.7%.10 Those in this study less than one year of age had the
highest mortality compared with other groups, but no one
more than 10 years of age died. After implementing an im-
proved case management system at the Department of Den-
gue Hemorrhagic Fever, Children’s Hospital No. 1 in Ho Chi
Minh City, we have treated 10,248 DHF/DSS patients of all
ages, of whom only 20 have died, resulting in a case fatality
rate of 0.19% during the period 1997–2002. However, the case
fatality rate remains higher in infants compared with older
children (0.76% versus 0.15% [P ⳱ 0.001]) (Hung NT and
others, unpublished data].
The present study emphasizes that fluid replacement in in-
fants with DHF/DSS is a challenge to good clinical manage-
ment. Intravenous fluids must be administered with special
care to avoid fluid overload. This involves following estab-
lished procedures for use of colloidal solutions and blood
transfusions. To further reduce case fatalities, special empha-
sis needs to be given to infants with DHF/DSS who have or
develop severe complications.
Received August 3, 2005. Accepted for publication November 25,
2005.
Acknowledgments: We thank Tran Tan Tram (former Director of
the Children’s Hospital No. 1, Ho Chi Minh City) and Chung-Ming
Chang (National Health Research Institutes, Taiwan) for help and
support during this study. We also thank the doctors and nurses of the
Department of Dengue Hemorrhagic Fever, Children’s Hospital
No.1 for providing excellent patient care.
Financial support: This study was supported in part grant NHRI-CN-
CL9303P from the National Health Research Institutes, Taiwan.
Disclosure: None of the authors have any commercial or other asso-
ciations that might pose a conflict of interest.
Authors’ addresses: Nguyen Thanh Hung, Nguyen Trong Lan, and
Le Bich Lien, Department of Dengue Hemorrhagic Fever, Children’s
Hospital No. 1, Ho Chi Minh City, Vietnam, Telephone: 84-8-927-
1119, Fax: 84-8-927-0053, E-mail: hungdhf@hcm.fpt.vn. Huan-Yao
Lei, Yee-Shin Lin, Kao-Jean Huang, and Chiou-Feng Lin, Depart-
ment of Microbiology and Immunology, College of Medicine, Na-
tional Cheng Kung University, Tainan, Taiwan, Republic of China.
Do Quang Ha and Vu Thi Que Huong, Arbovirus Laboratory, Pas-
teur Institute- Ho Chi Minh City, Vietnam. Lam Thi My, Department
of Pediatrics, University of Medicine and Pharmacy, Ho Chi Minh
City, Vietnam. Trai-Ming Yeh, Department of Medical Technology,
Department of Pediatrics, College of Medicine, National Cheng
Kung University, Tainan, Taiwan, Republic of China. Jyh-Hsiung
Huang, Division Of Research and Diagnosis, Department of Health,
Center For Disease Control, Taipei, Taiwan, Republic of China.
Ching-Chuan Liu, Department of Pediatrics, College of Medicine,
National Cheng Kung University, Tainan, Taiwan, Republic of
China. Scott B. Halstead, Uniformed Services University of The
Health Sciences, 4301 Jones Bridge Road, Bethesda, MD 20814-4799.
REFERENCES
1. Gubler DJ, 1998. Dengue and dengue hemorrhagic fever. Clin
Microbiol Rev 11: 480–496.
2. World Health Organization, 1997. Dengue Hemorrhagic Fever:
Diagnosis, Treatment, Prevention and Control. Second edition.
Geneva: World Health Organization.
3. World Health Organization (WHO), 2002. Dengue and dengue
hemorrhagic fever. Fact sheet No.117. Geneva: WHO.
4. Halstead SB, Lan NT, Myint TT, Shwe TN, Nisalak A, Kalyana-
rooj S, Nimmannitya S, Soegijanto S, Vaughn DW, Endy TP,
2002. Dengue hemorrhagic fever in infants: research opportu-
nities ignored. Emerg Infect Dis 8: 1474–1479.
5. Hung NT, Lei HY, Lan NT, Lin YS, Huang KJ, Lien LB, Lin CF,
Yeh TM, Ha DQ, Huong VTQ, Chen LC, Huang JH, My LT,
Liu CC, Halstead SB, 2004. Dengue hemorrhagic fever in in-
fants: a study of clinical and cytokine profiles. J Infect Dis 189:
221–232.
6. Nimmannitya S, 1997. Dengue hemorrhagic fever: diagnosis and
management. Gubler DJ, Kuno G, eds. Dengue and Dengue
Hemorrhagic Fever. Wallingford, United Kingdom: CAB In-
ternational, 133–146.
7. Ngo NT, Cao XT, Kneen R, Wills B, Nguyen VM, Nguyen TQ,
Chu VT, Nguyen TT, Simpson JA, Solomon T, White NJ,
Farrar J, 2001. Acute management of dengue shock syndrome:
a randomized double-blind comparison of 4 intravenous fluid
regimens in the first hour. Clin Infect Dis 32: 204–213.
8. Lan NT, Hung NT, Ha DQ, Phuong BT, Lien LB, Tuan LA,
Huong VT, Hieu LTM, Tran TN, Cam LT, Tuan NA, 1998.
Treatment of dengue haemorrhagic fever at Children’s Hospi-
tal No. 1, Ho Chi Minh City, 1991–1996. Dengue Bull 22: 150–
161.
9. Hung NT, Lan NT, 2003. Improvement of case management: a
key factor to reduce case-fatality rate of dengue hemorrhagic
fever in southern Vietnam. Dengue Bull 22: 144–148.
 VOLUME REPLACEMENT IN INFANTS WITH DENGUE
10. Hadinegoro SR, Purwanto SH, Chatab F, 1999. Dengue shock
syndrome: clinical manifestations, management and out-
come—A hospital-based study in Jakarta, Indonesia. Dengue
Bull 23: 105–106.
11. Hongsiriwon S, 2002. Dengue hemorrhagic fever in infants.
Southeast Asian J Trop Med Public Health 33: 49–55.
12. Hung NT, Lan NT, Lei HY, Lin YS, Lien LB, Huang KJ, Lin CF,
Ha DQ, Huong VTQ, My LT, Yeh TM, Huang JH, Liu CC,
Halstead SB, 2005. Association between sex, nutritional status,
severity of dengue hemorrhagic fever, and immune status in
infants with dengue hemorrhagic fever. Am J Trop Med Hyg
72: 370–374.
13. Shu PY, Chen LK, Chang SF, Yueh YY, Chow L, Chien LJ, Chin
C, Yang HH, Lin TH, Huang JH, 2003. Comparison of capture
immunoglobulin M (IgM) and IgG enzyme-linked immunosor-
bent assay (ELISA) and nonstructural protein NS1 serotype-
specific IgG ELISA for differentiation of primary and second-
ary dengue virus infections. Clin Diagn Lab Immunol 10: 622–
630.
14. Kuno G, Gomez I, Gubler DJ, 1987. Detecting artificial anti-
dengue IgM immune complexes using an enzyme-linked im-
munosorbent assay. Am J Trop Med Hyg 36: 153–159.
691
15. Kliks SC, Nimmannitya S, Nisalak A, Burke DS, 1988. Evidence
that maternal antibodies are important in the development of
dengue hemorrhagic fever in infants. Am J Trop Med Hyg 38:
411–419.
16. Lei HY, Yeh TM, Liu HS, Lin YS, Chen SH, Liu CC, 2001.
Immunopathogenesis of dengue virus infection. J Biomed Sci
8: 377–388.
17. Hung NT, Lan NT, 2001. Treatment of dengue hemorrhagic fever
at Children’s Hospital No. 1, Ho Chi Minh City, 1991–2000.
Med J Ho Chi Minh City 3: 149–152.
18. Nimmannitya S, 1987. Clinical spectrum and management of den-
gue hemorrhagic fever. Southeast Asian J Trop Med Public
Health 18: 392–397.
19. Lum LCS, Thong MK, Cheak YK, Lam SK, 1995. Dengue asso-
ciated with adult respiratory distress syndrome. Ann Trop Pe-
diatr 15: 335–339.
20. Cam BV, Tuan DT, Tuan HM, Tien NM, Fonsmark L, Poulsen
A, 2002. Randomized comparison of oxygen mask treatment
vs. nasal continuous positive airway pressure in Dengue shock
syndrome with acute respiratory failure. J Trop Pediatr 48:
335–339.