Original Studies
Clinical Characteristics and Risk Factors of Dengue Shock
Syndrome in Children
Dolores Lovera, MD,*† Celia Martinez de Cuellar, MD, MSc,*† Soraya Araya, MD,* Sara Amarilla, MD,*†
Nicolás Gonzalez, MD,*† Carlos Aguiar, MD,* Julia Acuña, MD,*† and Antonio Arbo, MD, MSc*†
Background: Dengue shock syndrome (DSS) represents one of the most
severe manifestations of dengue virus infection. The objective of the present
study was to analyze the clinical and laboratory characteristics, risk factors
and outcome of DSS in children.
Methods: Patients <15 years old admitted with DSS during the 2012 and
2013 outbreak of serotype 2 of dengue virus in Paraguay were included.
Demographic, clinical and laboratory data of patients with/without DSS
were analyzed.
Results: Of 471 children hospitalized with dengue, 354 patients (75%) pre-
sented with shock at admission or developed later. The mean age of patients
with DSS was 10.2 ± 4 years (no difference with patients without shock),
without gender preference. Rash (50% vs. 56%), myalgias (45% vs. 40%),
vomiting (66% vs. 68%) and bleeding manifestations (24% vs. 21.2%) were
similar for 2 groups. Similarly, there was no difference in the frequency of
DSS between primary versus secondary infection cases (76.2% vs. 71.6%,
P = 0.3). Age group >5 years [odds ratio (OR) 1.6, 95% confidence inter-
val (CI): 1–2.8, P < 0.05), presence of abdominal pain (OR 2.5, 95% CI:
1.3–4.9, P = 0.006), an activated partial thromboplastin time prolonged
(OR 4; 95% CI: 1.6–10, P < 0.001) and low fibrinogen level (OR 2.5; 95%
CI: 1–5.9, P = 0.02) were found significantly associated with DSS. About
12% of patients required intensive care unit admission, and 2 patients died
(lethality 0.35%).
Conclusions: This study validated most of the clinical variables present
in the current WHO guidelines as markers of severe disease and add
additional variables that can help to predict the risk of progression to
shock.
Key Words: dengue, shock, risk factors, children
(Pediatr Infect Dis J 2016;35:1294–1299)
D engue fever is a viral infection caused by a flavivirus (the den-
gue virus, DENV) with 4 different antigenically distinct sero-
types (DENV1-4).1 Currently, dengue is the most common arbovi-
ral disease globally.2 Dengue is endemic in more than 100 countries
worldwide, with an estimated 2.5 billion people living in areas at
risk, mainly in the countries of south and Southeast Asia, the Carib-
bean, Central and South America.3 Approximately 120 million peo-
ple travel to these regions each year. Thus, the number of dengue
cases has increased significantly in recent years among travelers
returning from trips to endemic countries.4
Accepted for publication May 15, 2016.
From the *Department of Pediatrics, Institute of Tropical Medicine, Asunción,
Paraguay; and †National University of Asuncion, San Lorenzo, Paraguay.
The authors have no funding or conflicts of interest to disclose.
Presented in part at the 54th Interscience Conference on Antimicrobial Agents
and Chemotherapy; September 5–9, 2014; Washington, DC. Abstract P-1285.
Address for correspondence: Antonio Arbo, MD, MSc, Department of Pediatric,
Instituto de Medicina Tropical, Avda, Venezuela y Florida, Asunción, Para-
guay. E-mail: antonioarbo@hotmail.com.
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
ISSN: 0891-3668/16/3512-1294
DOI: 10.1097/INF.0000000000001308
1294 | www.pidj.com The Pediatric Infectious Disease Journal  •  Volume 35, Number 12, December 2016
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Among the different clinical presentations of dengue, den-
gue shock syndrome (DSS) is the most severe form and affects
particularly children and young adults.4–6 Fatal cases of dengue
infection occur mostly in patients with DSS. Without proper treat-
ment, the mortality of DSS is reportedly 50 times higher than that
in dengue patients without DSS,7 and the fatality rates for DSS may
exceed 20%.7,8
In patients with DSS, early appropriate treatment can reduce
mortality.3,6,7,8 Therefore, tools that permit identification of which
patient will develop shock or other severe manifestations of dengue
can reduce the health care burden of dengue infection in endemic
countries and decrease the mortality. Therefore, determination of
risk factors of DSS is crucial for the early detection and proper
management of shock.
Although several studies have analyzed this subject, the
associations for some factors have not been observed consist-
ently across studies.8–16 An important outbreak of dengue serotype
2 occurred in Paraguay—country sited in the south of America
among Brazil, Argentina and Bolivia—during 2011 to 2013 that
resulted in more than 500 children hospitalized in the Tropical
Medicine Institute of Asuncion, the capital of the country.17 This
allowed us to reassess the risk factors and outcomes of DSS in the
pediatric population.
MATERIALS AND METHODS
Population and Study Design
Children admitted to the Institute of Tropical Medicine
(ITM) with the diagnosis of dengue fever during the 2011 to 2013
outbreak of dengue serotype 2 in Paraguay were included in this
study. ITM is a tertiary care teaching institution and is the main
referral center for infectious and tropical diseases in Paraguay. This
study was approved by the institution research ethics board.
The criteria for hospitalization of dengue cases were previ-
ously set in national guidelines for management of dengue fever,18
a local adaptation of the 2009 WHO guidelines for diagnosis, treat-
ment, prevention and control of dengue.4 In short, this guideline
distinguishes between severe and nonsevere dengue. Severe dengue
is defined by the occurrence of plasma leakage and/or fluid accumu-
lation leading to shock (DSS) or respiratory distress; and/or severe
bleeding; and/or severe organ impairment. The nonsevere dengue
group is divided into patients with and without warning signs. Cases
of dengue with warning signs included the presence of factors asso-
ciated with greater morbidity and mortality such as abdominal pain,
persistent vomiting, intolerance to oral, clinical accumulation of
fluid, bleeding of mucous membranes, lethargy/restlessness, precor-
dial pain, hepatomegaly or laboratory data, such as thrombocytope-
nia less than 100,000/mm3, hemoconcentration, documented by the
hematocrit increased up to 20% or more. Only cases of dengue with
warning signs or severe criteria were hospitalized.
Several laboratory methods were employed for the confirma-
tion of dengue including qualitative detection of IgG and IgM anti-
bodies specific to dengue (anti-DEN IgG and IgM; SD Dengue IgM
and IgG Capture enzyme-linked immunosorbent assay, Standard
The Pediatric Infectious Disease Journal  •  Volume 35, Number 12, December 2016 Risk Factors of DSS

Diagnostics INC, South Korea), demonstration of nonspecific pro- factor-specific relationships with DSS was performed. Platelet
tein of DENV (dengue NS1) by immunochromatography (Bioeasy, count, besides median count, was dichotomized at 100,000/mm3 and
Standard Diagnostics INC, Republic of Korea) and detection of 50,000/mm3, as these are the levels at which complications occur
DENV genomic sequences by polymerase chain reaction (PCR). more frequently. Univariate analyses were performed to investigate
A laboratory-confirmed case was defined by the presence of posi- the association of the relevant independent variables with the out-
tive dengue NS1 test or by the detection of DENV RNA in plasma. comes; χ2 tests, Fisher exact tests and odds ratios (OR) were calcu-
Patients in whom the dengue-specific IgM was already high at the lated where appropriate. Factors associated with shock on univariate
onset of shock were also considered as confirmed dengue, provided analysis with a P value <0.10 were further assessed by multivariate
that the overall clinical picture was consistent with DSS, and no logistic regression with correction for age and gender, with results
alternative diagnosis was established. In patients whose PCR, NS1 given as OR and 95% confidence intervals (CIs). A P value <0.05
antigen test, anti-DEN IgM/IgG-negative testing were negative was accepted as statistically significant for all analyses. All analyses
on the sample obtained at the time of admission, a second serum were performed with the statistical software R, version 2.15.0.
sample was obtained (convalescent sera) for determination besides
anti-DEN IgM, anti-DEN IgG neutralization titers. A ≥4-fold rise
RESULTS
in DENV HAI IgG titers between 2 different points in time was
also confirmatory of dengue. Cases with a clear epidemiological In the study period, 560 children less than 15 years of age
link were also included (classical clinical pictures plus one or more were admitted to the ITM with the clinical diagnosis of dengue
cases of dengue in people living in the same household). viral infection. Only patients with laboratory confirmation of the
Patients who exhibited at admission PCR or NS1 antigen diagnosis or with an unquestionable epidemiological link were
capture test positive with anti-DEN IgM positive and IgG nega- included in the analysis.
tive were considered to have a primary Dengue infection. In the The diagnosis of dengue was confirmed by a positive PCR
absence of PCR or NS1 Ag test positive, the new appearance of or NS1 antigen capture test in 242 patients (139 of this group
anti-DEN IgM with anti-DEN IgG negative was also considered to showed also the presence of anti-DEN IgM and 74 of both anti-
be a primary infection. Patients with PCR or NS1 antigen test posi- DEN IgM and IgG antibodies), by new appearance of anti-DEN
tive at admission with IgG positive (with or without anti-DEN IgM IgM in 170 children (125 of this group showed also anti-DEN IgG).
positive) were considered to have a secondary dengue infection. In addition, in 23 patients, the diagnosis was based in ≥4-fold rise
The population of hospitalized patients was divided into 2 in DENV HAI IgG titers between 2 points in time, and in 36 chil-
groups, depending on the presence of shock. Patients were consid- dren, dengue diagnosis was made by epidemiological link. In sum-
ered to have shock if the pulse pressure was 20 mm Hg or lower, or mary, 471 patients were finally included in the analysis.
Relevant demographic, clinical and laboratory characteristics
signs of poor capillary perfusion (cold extremities, delayed capil-
of the study population are shown in Table 1. Briefly, the mean age
lary refill or rapid pulse rate) were present. Patients with hema-
of dengue patients admitted was 10 ± 4 years, and the main age group
tologic disorders, congenital heart diseases (known or suspected),
affected has been children beyond 5 years of age (84%), with a simi-
congenital or acquired immunodeficiency, cancer and chronic lung
lar gender distribution. Symptoms and clinical signs at admission
or renal diseases were excluded.
were those commonly observed in a series of cases of dengue.
All subjects were admitted to the hospital and monitored
until 24 hours after defervescence. Fluid management was car-
ried out following the recommendations of the 2009 WHO Dengue TABLE 1.  Demographic, Clinical and Laboratory
Guideline,4 which is characterized by early and titrated intravenous Characteristics of Patients
fluid therapy. Our management of shock has been previously pub-
lished18 and included a bolus of 20 mL/kg of isotonic crystalloid No of Patients = 471
solution (such as 0.9% normal saline or lactated Ringer solution)
Age (mean ± SD; years) 10 ± 4
infused as rapidly as possible. The 20 mL/kg fluid boluses were
 <5 years 77 (16%)
repeated until blood pressure, tissue perfusion and oxygen deliv-  >5 years 394 (84%)
ery were adequate, or signs of fluid overload (rales, gallop rhythm, Sex male 241 (51%)
enlarged liver) develop. Once effective circulating volume has been Obesity/overweight 124 (26%)
restored, the rate of infusion of saline solution was decreased and Malnutrition 57 (12%)
further followed by the maintenance volume of hydration.4  Severe 7 (1.5%)
Symptoms/signs
 Headache 303 (64%)
Data Collection  Myalgias 226 (48%)
Signs and symptoms, findings on physical examination and  Vomiting/nausea 321 (68%)
routine laboratory test data were obtained at admission and during  Bleeding 122 (26%)
the hospital stay with a standardized case report form. Laboratory    Severe hemorrhage 35 (7.4%)
parameters, including hemoglobin, total and differential leukocyte  Rash 262 (56%)
 Abdominal pain 360 (76%)
count, platelet count, serum C-reactive protein, blood urea, serum Laboratory
creatinine, prothrombin time (PT), activated partial thromboplastin  WBC/mm3 (mean ± SD) 5,522 ± 3,617
time (APTT) with international normalized ratio, electrolytes, blood  Mean hemoglobin ± SD (g/dL) 14 ± 3
pH, serum bicarbonate, arterial oxygen (PaO2) and carbon dioxide  Platelet/mm3 (mean ± SD) at admission 121,697 ± 85,502
content (PaCO2), were obtained. A complete blood count was done  Creatinine (mg/dL; n = 391) 0.6 ± 0.5
 Urea (mg/dL; n = 391) 23 ± 14
daily. Blood chemistry, radiographic imaging and abdominal ultra-
 Sodium (meq/L; n = 229) 139 ± 4.4
sound were done to monitor the course of the disease, especially  Potassium (meq/L; n = 229) 4.2 ± 0.8
during defervescence.  Calcium (meq/L; n = 265) 1.1 ± 0.1
 ALT level (UI/mL, mean ± SD) 89 ± 205
Statistical Analysis  AST level (UI/mL, mean ± SD) 149 ± 380
Continuous variables were expressed as mean ± SD and ALT indicates alanine aminotransferase; AST, aspartate aminotransferase; WBC,
categorical variables as numbers and percentages. Analysis of white blood cells.

© 2016 Wolters Kluwer Health, Inc. All rights reserved. www.pidj.com | 1295

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Lovera et al The Pediatric Infectious Disease Journal  •  Volume 35, Number 12, December 2016

TABLE 2.  Demographic and Anthropometric Characteristics of Patients With DSS

No of patients Shock Without Shock

(n = 471) (N=354) (N = 117) P OR

Age (mean ± SD; years) 10 ± 4 10 ± 4 9.5 ± 4.5 0.07

 <24 months 37 (8%) 26 (7.3%) 11 (9.4%) 0.04 1.6 (1–2.8)
 2–5 years 40 (8%) 25 (7%) 15 (13%)
 >5 years 394 (84%) 303 (86%) 91 (78%)*
Sex male 241 (51%) 183 (52%) 58 (50%) 0.7 1 (0.7–1.4)
Sex female 230 (49%) 171 (48%) 59 (50%)
Urban 409 (87%) 310 (88%) 98 (84%) 0.29 1.3 (0.7–2.4)
Rural 62 (13%) 44 (12%) 19 (16%)
Obesity/overweight 124 (26%) 97 (27%) 27 (23%) 0.35 1.2 (0.7–2)
 Obesity 26 (6%) 21 (5.9%) 5 (4%) 0.49 1.4 (0.5–4)
Malnutrition 57 (12%) 41 (11.6%) 16 (14%) 0.54 0.8 (0.4–1.5)
Severe malnutrition 7 (1.5%) 3 (0.8%) 4 (3.4%) 0.06 0.2 (0.0–1)
*patients >5 years of age were more likely to develop shock compared with other groups (see in the text), P < 0.05.

Three hundred and fifty-four patients (75%) were admitted
with shock (cases of DSS) or developed it during hospitalization.
The others were cases of dengue with warning signs or severe den-
gue without shock (n = 117, 25%). In 310 patients (87, 6%), the
shock was during the compensatory stage, whereas 44 (12%) of
the remainder exhibited hypotension. Tables 2 and 3 show the main
characteristics of patients with shock in comparison to those patients
who required hospitalization but did not develop it. The mean age
of patients with shock was no different from patients who did not
develop shock during the hospitalization (10, 2 ± 4 years vs. 9, 5 ± 4.5
years). However, in age group comparisons, patients >5 years of age
were more likely to develop shock compared with other groups (77%
vs. 62.5% in the group from 2 to 5 years and 62% in children <24
months, P < 0.05; OR 1, 95% CI: 1–2.8). There were no significant
differences based on gender (male/female ratio 1:1) and location of
the residence between both groups. Further, neither malnutrition,
obesity nor overweight was associated with DSS (Table 2).
There were no differences in the frequency of rash (49.4%
vs. 58.9%), myalgias/arthralgias (49.4% vs. 42.7%), headache
(65.2% vs. 60.6%), vomiting (68.6% vs. 65.8%) and bleeding
manifestations (26% vs. 26%) among those who developed shock
compared with those who never presented it. When only severe
cases of bleeding were analyzed, this finding was significantly more
frequent in patients with shock (9% vs. 2.5%; OR 3.7, 95% CI:
1.1–12, P = 0.02). Abdominal pain was the other clinical sign found
with higher incidence in patients with shock (78.5% vs. 69%; OR
1.6, 95% CI: 1–2.6, P < 0.05; Table 3).
Patients with shock demonstrated an almost similar degree
of hemoconcentration compared with patients without shock, both
at admission (39.5% ± 5.8% vs. 39.1% ± 5.7%) and during the
hospitalization (41.5% ± 6.8% vs. 40.5% ± 6.8%). When a hema-
tocrit cutoff value >40% was explored as risk factor of shock, it
was not significantly associated with the development of shock (OR
1.4, 95% CI: 0.9–2.1, P = 0.1). When signs of plasma leakage was
analyzed, both groups exhibited the same frequency. However, the
presence of effusion in more than 1 cavity (eg, pleural effusion plus
ascites or thickened gallbladder) compared with its absence or its
presence in only 1 cavity was significantly related to major fluid
requirements, indicating that it is a marker of severity (Table 4).
The median platelet count at admission in patients who
develop shock was 120,277/mm3 (range 6300–510,000/mm3), and
the lowest value observed during the hospitalization was 68,594/mm3
(range 3300–360,000/mm3). Both figures were no different in patients
who were hospitalized but did not develop shock [platelet count at
admission 126,438/mm3 (range 7300–480,000/mm3) and lowest value
of 79,115/mm3 (range 5000–376,000/mm3)]. When a platelet count
cutoff value of <100,000/mm3 was explored as risk factors of shock,
it was found that children who did not develop shock exhibit in simi-
lar percentage; this level of thrombocytopenia compared with patients
who eventually developed it (47% vs. 49%, respectively; OR 1.7, 95%
CI: 0.7–1.8, P = 0.59; Table 3). The results were similar when the
comparison included patients only with platelets counts <50,000/mm3
(28% vs. 25.6%, respectively; OR 1, 95% CI: 0.6–1.8, P = 0.6).
Among the other laboratory data, only the presence of a
prolonged activated partial thromboplastin time and low fibrino-
gen level at admission was associated with DSS. None of the other
laboratory variables in patients were different when they were com-
pared between patients who developed shock versus patients who

TABLE 3.  Clinical Characteristics of Patients With DSS

Shock Without Shock

(N = 354) (N = 117) P OR

Headache 231 (65%) 71 (61%) 0.37 1.2 (0.7–1.8)

Myalgias 175 (49%) 50 (43%) 0.29 1.3 (0.8–2)
Vomiting/nausea 243 (66%) 77 (66%) 0.56 1 (0.7–1.7)
Bleeding 92 (26%) 30 (26%) 0.94 1 (0.6–1.6)
 Severe hemorrhage 32 (9%) 3 (2.5%) 0.02 3.7 (1.1–12)
Rash 192 (54%) 70 (60%) 0.29 0.7 (0.5–1)
Abdominal pain 279 (79%) 81 (69%) 0.03 1.6 (1–2.6)
Plasma leakage signs 206 (58%) 63 (54%) 0.4 1 (0.7–1.8)
 Hemoconcentration 214 (60%) 63 (54%) 0.2 1 (0.8–1.8)
 Pleural effusion 152 (43%) 40 (34%) 0.09 1.4 (0.9–2)
 Ascites 187 (53%) 54 (46%) 0.21 1.3 (0.8–1.9)
 Hypoalbuminemia (<3.5  g/dL) 110 (77%) 33 (23%) 0.55 1 (0.7–1.8)
 Hypoproteinemia (<5.7  g/dL) 66/172 (19%) 18/51 (15%) 0.42 1 (0.5–2)

1296 | www.pidj.com © 2016 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
The Pediatric Infectious Disease Journal  •  Volume 35, Number 12, December 2016 Risk Factors of DSS

TABLE 4.  Plasma Leakage and Severity of Dengue

No of Patients ≥3 Fluid Infusion <3 Fluid Infusion

(N = 418) Boluses (N = 170) Boluses (N = 248)

Absence of pleural/peritoneal effusion 195 53 (27%) 142 (73%)

Pleural effusion 15 6 (40%) 9 (60%)
Ascites 33 10 (30%) 23 (70%)
Pleural effusion + ascites 38 18 (47%)* 20 (53%)
Pleural effusion + ascites + thickened 137 83 (61%)† 54 (39%)
gallbladder wall
*P = 0.018 versus absence of pleural/peritoneal effusion.
†P < 0.001 versus absence of pleural/peritoneal effusion.

did not; hence they were not considered to be specific risk factors
for the development of shock (Table 5).
In 388 patients (82.5%), we could determine whether
the cases were primary (n = 141) or secondary dengue infection
(n = 248). In 82 patients (17.4%), the distinction could not be made.
Patients with primary infection developed shock (71.6%) in the
same proportion as those with secondary dengue (75%, P = 0.2).
When the comparisons were done for the different age groups, the
results were similar: 70.5% versus 67% in the <2 years of age group
(P = 0.8), 65% versus 76% in the 2–5 years old group (P = 0.8)
and 73% versus 84% in the group of children with >5 years of age.
Fifty-seven patients (12%) developed prolonged shock
and needed intensive care unit admission. Only 6 children
(1.3%) were given blood products (packed red blood cells in 2,
platelet concentrates in 2 and both in 1 patient). The overall case
fatality rate in the study population was 1.3% (6/471 patients).
DISCUSSION
One of the most important challenges faced by the staff phy-
sicians who care for patients with dengue is the early identification
of hospitalized patients at higher risk for developing shock. A pre-
vious multicenter study involving cases of dengue from different
countries including Paraguay analyzed the association of dengue
serotypes with clinical manifestations, however, the study did not
include inpatients.19 This study done in hospitalized patients with
dengue focused in particular on the correlation of clinical symp-
toms/signs and laboratory data with the development of DSS.
Although the clinical presentations of dengue patients at
admission were similar to previous descriptions of dengue with
warning signs, we found differences regarding symptoms/signs
that were reported to be associated with DSS. Series from Latin
America and Southeast Asia showed higher prevalence of severe
dengue in women, including a recently published meta-analysis,
which clearly demonstrated the association of female gender with
the risk of DSS.12,20,21 However, we did not observe any differences
in gender when correlated with the development of shock. This
same finding was reported recently with dengue in El Salvador.22
Different reports have indicated that young children have
an increased risk of DSS15,16,23,24 probably because of an increased
microvascular fragility in younger children.25 In the present study,
however, shock was not more common in the children under
2 years of age. Among the different pediatric age groups, DSS was
more frequent in children >5 years of age. Other authors have also
reported this finding.13,14,26 Because the group of children older than
5 years of age in our study was significantly larger (n = 394, 84%)
than those younger than 5 (n = 77, 16%), a bias of the sample dis-
tribution since the enrollment cannot be excluded.

TABLE 5.  Laboratory Characteristics of Patients With DSS

Shock Without Shock

(N = 354) (N = 117) P OR

WBC/mm3 (mean ± SD) 5468 ± 3596 5683 ± 3690 0.54

Mean hemoglobin ± SD (g/dL) 13 ± 2 13 ± 2 0.2
Mean hematocrit ± SD (%) 39.5 ± 6 39 ± 5 0.2
Hematocrit >38 at admission 229 (65%) 75 (64%) 0.9 1 (0.6–1.5)
Platelet/mm3 (mean ± SD) 120,277 ± 86,431 126,029 ± 9036 0.5
Platelet <100,000/mm3 173 (49%) 55 (47%) 0.7 1 (0.7–1.6)
Platelet <50,000/mm3 99 (28%) 30 (25.6%) 0.6 1 (0.6–1.8)
Prothrombin time <70% 34/167 (20%) 6/54 (11%) 0.1 2 (0.8–5)
Prolonged APTT 56/162(35%) 6/52 (12%) 0.001 4 (1.6–10)
Fibrinogen level <200 mg/dL 50/137 (36%) 8/44 (18%) 0.02 2.5 (1–5.9)
Creatinine (mg/dL; n = 391) 0.6 ± 0.47 0.6 ± 0.48 0.11
Urea (mg/dL; n = 391) 22 ± 14.6 22 ± 14.6 0.38
Sodium (meq/L; n = 229) 139 ± 4.4 138 ± 4.4 0.6
Potassium (meq/L; n = 229) 4.2 ± 0.8 4.2 ± 0.8 0.3
Calcium (meq/L; n = 265) 1.09 ± 0.1 1.09 ± 0.1 0.6
ALT level (UI/mL, mean ± SD) 89 ± 205 90 ± 207
 ALT level >5 times 23/333 (7%) 6/109 (5.5%) 0.6 1 (0.5–3)
 ALT level >1000 UI 4/333 (1.2%) 0/109 0.5 Undefined
AST level (UI/mL, mean ± SD) 148 ± 380 150 ± 383
 AST level >5 times 48/333 (14%) 12/109 (11%) 0.36 1 (0.6–2.6)
 AST level >1000 UI 7/333 (2%) 1/109 (0.9%) 0.42 2 (0.2–19)
CK level (U/L; n = 63) 372 ± 492 330 ± 333 0.3
Increased CK-MB level >100 U/L 9/53 (17%) 1/10 (10%) 0.49 2 (0.2–16)
ALT indicates alanine aminotransferase; APTT, activated partial thromboplastin time; AST, aspartate aminotrans-
ferase; WBC, white blood cells.

© 2016 Wolters Kluwer Health, Inc. All rights reserved. www.pidj.com | 1297

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Lovera et al The Pediatric Infectious Disease Journal  •  Volume 35, Number 12, December 2016
Several reports suggest that normal nutrition is a risk factor
of DSS, while malnutrition could be a protective factor, probably
related to suppressed immune activation in malnourished chil-
dren.12–14,23 In our series of patients, however, we did not find any
differences in the nutritional status among patients who develop
shock compared with those with dengue without shock.
Various symptoms and signs were included as warning signs in
the last WHO dengue guideline.4 Although they have been shown to be
very useful for early intervention (mainly fluid therapy), in the present
series only few signs of dengue severity were identified to be associ-
ated with the development of shock, such as abdominal pain as well as
effusion in more than 1 cavity.27 These findings are manifestations of
the severe increase in capillary leakage, the main pathological cause of
development of shock27–29 and are consistent with findings in the pedi-
atric population in Puerto Rico, Cuba and Colombia.26,30,31
Although other series identified bleeding manifestations
as risk factors of shock,16,26,32,33 in our study only the presence of
severe bleeding was associated with shock. It has been shown that
the severity of bleeding is closely related, among others factors,
with increased microvascular fragility and poor capillary perfu-
sion.34,35 Thus, severe bleeding during the course of the disease
should alarm clinicians that the dengue patients are at risk of shock.
In our study, the presence of hemoconcentration in patients
hospitalized with warning signs was a frequent finding in the age
group of children >5 years of age but was only found in 17% in
children <2 years of age. Compared with studies from Southeast
Asia, where hemoconcentration was found in 90% of patients with
DSS,36–38 our patients demonstrated a much smaller degree of hemo-
concentration (10%–20%) than expected by the current definitions.
The mean of hematocrit was similar when patients with shock were
compared with patients without shock, and hemoconcentration ≥20%
was observed only in 55% of the patients with shock. However, a sim-
ilar percentage was observed also in dengue patients without shock
(50.5%). Hemoconcentration below the WHO threshold of 20% has
been noted previously in DSS patients.22,39
Four hundred and seven hospitalized patients (86%) exhib-
ited thrombocytopenia, which was <100,000/mm3 [1 of the 4 com-
ponents of the definition of dengue hemorrhagic fever (DHF)] in 351
of them (74.5%). This figure shows the severity of hospitalized cases
of dengue in our institution, where only patients with warning signs
or with shock are hospitalized. It was demonstrated that the severity
of the thrombocytopenia is correlated with plasma viral load, which
has been shown to correlate also with the extent of plasma leakage.35
Contrary to other studies, however, where severe thrombocytopenia
was associated with higher risk of shock,4,8,12,40 in our study both the
mean platelet count and a low platelet count (<20,000/mm3) were
found in similar proportion in patients with and without shock.
A reduction in fibrinogen concentration and an increase
in the PT and APTT were consistent coagulation abnormalities
detected between 16% and 25% of the patients. Other studies have
demonstrated the temporal association of the APTT and fibrinogen
changes with the period of maximal vascular leakage. The decrease
in the fibrinogen concentration could be explained by the increase
in transcapillary filtration in a similar manner to that of albumin.29
Changes in the capillary structure can also explain the prolonga-
tion of the APTT. Damage of the glycocalix of endothelial cells can
result in the release of heparan sulfate into the circulation, which
would act as an anticoagulant similar to heparin. Reduction of
fibrinogen level and prolongation of APTT showed a positive asso-
ciation with DSS. These observations are not surprising, because
both coagulation abnormalities are strongly associated with vascu-
lar leakage both temporally and in terms of severity.41,42
Prospective studies in Thailand and other countries from the
Southeast Asia,43–45 as well as studies of sequential epidemics of
1298 | www.pidj.com
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
dengue in Cuba,46 have shown the association of secondary infection
with more severe disease, probably related to the role of antibody-
dependent enhancement in DSS pathogenesis.1 In our study, however,
cases of DSS were observed in similar percentage in patients with pri-
mary or secondary infection. It is important to notice that the majority
of patients were hospitalized during the outbreak of infection of DEN
2 serotype. The genotype of DEN 2 virus has been associated with
increased disease severity, whether in the context of secondary infec-
tion or otherwise.47,48 Preliminary observations indicate that the Den-
gue 2 virus responsible for the last outbreak that occurred in Paraguay
are phylogenetically associated with the strains originated in Southeast
Asia (Cinthia Vazquez, BS, personal communication, December 1,
2015). The major success of the Southeast Asian dengue 2 strain is
probably because of more efficient replication in human target cells as
well increased transmission by vector mosquitoes.47
The mortality rate of the present series was low (1.2%),
considering that the denominator has included only severe cases
of dengue. Same aspect of the management of the cases are worthy
of mention. The use of blood products (packed red blood cells or
platelet concentrates) was infrequent, even in patients with a very
low platelet count. Most patients recovered well with the standard
titrated intravenous fluid therapy, and only a minimal proportion of
patients required inotropic support.
Our study has a number of limitations because of its retro-
spective design and the fact that data were collected from a referral
center. Thus, the results may not be applicable to community health
care settings. In addition, it should be considered that the vast
majority of patients were cases of infection by the dengue serotype
2. The epidemiology of dengue is different in different geographi-
cal areas and depends upon circulating serotypes, and thus it can
vary from year to year. For this reason, the variables found to be
associated with the development of shock may not be applicable
to other areas.
Despite the known limitations of this study, our findings
suggest specific risk factors that may predict the development of
shock. Prospective studies to confirm our findings are needed.
REFERENCES
1. Simmons CP, Farrar JJ, Nguyen vV, et al. Dengue. N Engl J Med.
2012;366:1423–1432.
2. Gubler DJ. The economic burden of dengue. Am J Trop Med Hyg.
2012;86:743–744.
3. Suaya JA, Shepard DS, Siqueira JB, et al. Cost of dengue cases in eight
countries in the Americas and Asia: a prospective study. Am J Trop Med
Hyg. 2009;80:846–855.
4. Wilder-Smith A, Schwartz E. Dengue in travelers. N Engl J Med.
2005;353:924–932.
5. World Health Organization. Dengue: Guidelines for Diagnosis, Treatment,
Prevention and Control. New ed. Geneva, Switzerland: World Health
Organization; 2009.
6. Wills BA, Nguyen MD, Ha TL, et al. Comparison of three fluid solutions for
resuscitation in dengue shock syndrome. N Engl J Med. 2005;353:877–889.
7. Nimmannitya S. Clinical spectrum and management of dengue haemor-
rhagic fever. Southeast Asian J Trop Med Public Health. 1987;18:392–397.
8. Anders KL, Nguyet NM, Chau NV, et al. Epidemiological factors associated
with dengue shock syndrome and mortality in hospitalized dengue patients
in Ho Chi Minh City, Vietnam. Am J Trop Med Hyg. 2011;84:127–134.
9. Mena Lora AJ, Fernandez J, Morales A, et al. Disease severity and mortality
caused by dengue in a Dominican pediatric population. Am J Trop Med Hyg.
2014;90:169–172.
10. Srikiatkhachorn A, Green S. Markers of dengue disease severity. Curr Top
Microbiol Immunol. 2010;338:67–82.
11. Ranjit S, Kissoon N. Dengue hemorrhagic fever and shock syndromes.
Pediatr Crit Care Med. 2011;12:90–100.
12. Kalayanarooj S, Nimmannitya S. Is dengue severity related to nutritional
status? Southeast Asian J Trop Med Public Health. 2005;36:378–384.
© 2016 Wolters Kluwer Health, Inc. All rights reserved.
The Pediatric Infectious Disease Journal  •  Volume 35, Number 12, December 2016 Risk Factors of DSS
13. Huy NT, Van Giang T, Thuy DH, et al. Factors associated with dengue shock syn-
drome: a systematic review and meta-analysis. PLoS Negl Trop Dis. 2013;7:e2412.
14. Wichmann O, Hongsiriwon S, Bowonwatanuwong C, et al. Risk factors and
clinical features associated with severe dengue infection in adults and chil-
dren during the 2001 epidemic in Chonburi, Thailand. Trop Med Int Health.
2004;9:1022–1029.
15. Pham TB, Nguyen TH, Vu TQ, et al. [Predictive factors of dengue shock
syndrome at the children Hospital No. 1, Ho-chi-Minh City, Vietnam]. Bull
Soc Pathol Exot. 2007;100:43–47.
16. Shah I, Deshpande GC, Tardeja PN. Outbreak of dengue in Mumbai and pre-
dictive markers for dengue shock syndrome. J Trop Pediatr. 2004;50:301–305.
17. Hammond SN, Balmaseda A, Pérez L, et al. Differences in dengue severity
in infants, children, and adults in a 3-year hospital-based study in Nicaragua.
Am J Trop Med Hyg. 2005;73:1063–1070.
18. Lovera D, Araya S, Mesquita MJ, et al. Prospective applicability study of
the new dengue classification system for clinical management in children.
Pediatr Infect Dis J. 2014;33:933–935.
19. Ministerio de Salud Pública y Bienestar Social del Paraguay. Dengue: Guia
de manejo clínico. 2012. Available at: http://www.mspbs.gov.py/documen-
tacion/Dengue_guia_2012.pdf. Accessed November 15, 2015.
20. Halsey ES, Marks MA, Gotuzzo E, et al. Correlation of serotype-specific dengue
virus infection with clinical manifestations. PLoS Negl Trop Dis. 2012;6:e1638.
21. Hammond SN, Balmaseda A, Pérez L, et al. Differences in dengue severity
in infants, children, and adults in a 3-year hospital-based study in Nicaragua.
Am J Trop Med Hyg. 2005;73:1063–1070.
22. Souza LJ, Pessanha LB, Mansur LC, et al. Comparison of clinical and labo-
ratory characteristics between children and adults with dengue. Braz J Infect
Dis. 2013;17:27–31.
23. Marón GM, Escobar GA, Hidalgo EM, et al. Characterization of dengue
shock syndrome in pediatric patients in El Salvador. Pediatr Infect Dis J.
2011;30:449–450.
24. Hung NT, Lan NT, Lei H-Y, et al. Association between sex, nutritional
status, severity of dengue hemorrhagic fever, and immune status in
infants with dengue hemorrhagic fever. Am J Trop Med Hyg. 2005;
72:370–374.
25. Guzmán MG, Kouri G, Bravo J, et al. Effect of age on outcome of secondary
dengue 2 infections. Int J Infect Dis. 2002;6:118–124.
26. Gamble J, Bethell D, Day NP, et al. Age-related changes in microvascular
permeability: a significant factor in the susceptibility of children to shock?
Clin Sci (Lond). 2000;98:211–216.
27. Gupta V, Yadav TP, Pandey RM, et al. Risk factors of dengue shock syn-
drome in children. J Trop Pediatr. 2011;57:451–456.
28. Balasubramanian S, Janakiraman L, Kumar SS, et al. A reappraisal of the
criteria to diagnose plasma leakage in dengue hemorrhagic fever. Indian
Pediatr. 2006;43:334–339.
29. Leong AS, Wong KT, Leong TY, et al. The pathology of dengue hemor-
rhagic fever. Semin Diagn Pathol. 2007;24:227–236.
30. Trung DT, Wills B. Systemic vascular leakage associated with dengue infec-
tions—the clinical perspective. Curr Top Microbiol Immunol. 2010;338:57–66.
© 2016 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
31. Rigau-Pérez JG, Laufer MK. Dengue-related deaths in Puerto Rico, 1992-
1996: diagnosis and clinical alarm signals. Clin Infect Dis. 2006;42:1241–
1246.
32. Guzmán MG, Alvarez M, Rodríguez R, et al. Fatal dengue hemorrhagic
fever in Cuba, 1997. Int J Infect Dis. 1999;3:130–135.
33. Méndez A, González G. [Dengue haemorrhagic fever in children: ten years
of clinical experience]. Biomedica. 2003;23:180–193.
34. Alexander N, Balmaseda A, Coelho IC, et al; European Union, World
Health Organization (WHO‐TDR) supported DENCO Study Group.
Multicentre prospective study on dengue classification in four South-
east Asian and three Latin American countries. Trop Med Int Health.
2011;16:936–948.
35. Chen RF, Yang KD, Wang L, et al. Different clinical and laboratory man-
ifestations between dengue haemorrhagic fever and dengue fever with
bleeding tendency. Trans R Soc Trop Med Hyg. 2007;101:1106–1113.
36. Martina BE, Koraka P, Osterhaus AD. Dengue virus pathogenesis: an inte-
grated view. Clin Microbiol Rev 2009; 22:564–581.
37. Lam PK, Tam DT, Diet TV, et al. Clinical characteristics of Dengue shock
syndrome in Vietnamese children: a 10-year prospective study in a single
hospital. Clin Infect Dis. 2013;57:1577–1586.
38. The TD, Thu TLT, Minh DN, et al. Clinical features of dengue in a large
vietnamese cohort: intrinsically lower platelet counts and greater risk for
bleeding in adults than children. PLoS Negl Trop Dis. 2012;6:e1679.
39. Balasubramanian S, Anandnathan K, Shivbalan S, et al. Cut-off hematocrit
value for hemoconcentration in dengue hemorrhagic fever. J Trop Pediatr.
2004;50:123–124.
40. Kalayanarooj S, Vaughn DW, Nimmannitya S, et al. Early clinical and labo-
ratory indicators of acute dengue illness. J Infect Dis. 1997;176:313–321.
41. Balmaseda A, Hammond SN, Pérez MA, et al. Short report: assessment of
the World Health Organization scheme for classification of dengue severity
in Nicaragua. Am J Trop Med Hyg. 2005;73:1059–1062.
42. Wills BA, Oragui EE, Stephens AC, et al. Coagulation abnormalities in
dengue hemorrhagic Fever: serial investigations in 167 Vietnamese children
with Dengue shock syndrome. Clin Infect Dis. 2002;35:277–285.
43. Wills B, Tran VN, Nguyen TH, et al. Hemostatic changes in Vietnamese
children with mild dengue correlate with the severity of vascular leakage
rather than bleeding. Am J Trop Med Hyg. 2009;81:638–644.
44. Burke DS, Nisalak A, Johnson DE, et al. A prospective study of dengue
infections in Bangkok. Am J Trop Med Hyg. 1988;38:172–180.
45. Sangkawibha N, Rojanasuphot S, Ahandrik S, et al. Risk factors in dengue
shock syndrome: a prospective epidemiologic study in Rayong, Thailand.
I. The 1980 outbreak. Am J Epidemiol. 1984;120:653–669.
46. Thein S, Aung MM, Shwe TN, et al. Risk factors in dengue shock syndrome.
Am J Trop Med Hyg. 1997;56:566–572.
47. Guzmán MG, Kourí G, Valdés L, et al. Enhanced severity of secondary den-
gue-2 infections: death rates in 1981 and 1997 Cuban outbreaks. Rev Panam
Salud Publica. 2002;11:223–227.
48. Kyle JL, Harris E. Global spread and persistence of dengue. Annu Rev
Microbiol. 2008;62:71–92.
www.pidj.com | 1299