Predictors of Spontaneous Bleeding in Dengue

So. Shivbalan, K. Anandnathan, S. Balasubramanian, Manjula Datta 1 and Edwin Amalraj 1

Kanchi Kamakoti ChiIds Trust Hospital, Nungambakkam, Chennai and Department of Clinical Epidemiology, "Dr.
MGR Medical University" Chennai, India.

Abstract. Objective : To identify the predictive factors for spontaneous bleeding manifestations in Dengue illness in infants
and children. Methods : 60 cases of Dengue viral infection with spontaneous skin and or mucosal bleed were compared with
72 cases without spontaneous bleed. The protean bleeding manifestations in dengue infection were recorded. Various clinical
and laboratory parameters were analyzed using univariate and logistic regression analysis. Results and Conclusion :
Prothombin time was abnormal only in cases with spontancous bleed. A combination of (a) biphasic pattern of fever, (b)
hemoconcentration, (c) platelet count less than 50,000/mm 3 and (d) elevated ALT had a sensitivity of 79.2%, specificity of
64.7% with a positive predictive value of 70% and a negative predictive value of 75% in predicting spontaneous bleeding in
dengue. [Indian J Pediatr 2004; 71 (1) : 33-36] E-mail : [email protected]

Key words : Spontaneous bleeding; Dengue illness

Dengue fever is usually a benign syndrome caused by an DHF grade 2 with bleeding other than skin bleeds, DHF/
arthropod borne virus. Bleeding in dengue is one of the DSS grade 3 and 4, anxious parents referred for admission
dreaded complications. Clinical manifestations of bleed and patients from long distance who had difficulty in
are highly variable from simple skin bleeds like petechiae availing immediate medical attention. All patients
or purpura to severe bleeds like bleed from puncture site, belonged to a single unit under a senior consultant.
GI bleeds, fatal intracranial bleed and bleeds from viscera. Case definition for Probable DF, DHF, DSS and
The exact mechanism of bleeding is not known. Factors grading of DHF were based on WHO criteria. 3 These
like mild degree of DIVC, hepatic derangement and patients were divided into 2 groups i.e. the study group
thrombocytopenia are said to act synergistically. 1 and control group. Study group consisted of cases who
Bleeding manifestations are associated with higher had a skin a n d / o r mucosal spontaneous bleeding
mortality in Dengue Haemorrhagic Fever (DHF) / Dengue manifestation either on admission or any time during the
Shock Syndrome (DSS).2These manifestations are highly period of hospital stay. Control group consisted of cases
variable and do not always correlate with the laboratory without spontaneous bleeding. Control group also
abnormalities in the coagulation profile 1. Hence it would included cases with induced bleeds like positive tourniqet
be of value to identify factors that predict spontaneous test with or without easy bruisability.
bleed in dengue illness. The aim of the study was to Clinical data were collected using a structured
identify the predictive factors for spontaneous bleeding proforma, which included details about the case history,
manifestations in Dengue illne~,sin infants and children. clinical findings, laboratory investigations and
management. The clinical and laboratory data of dengue
MATERIALS AND METHODS cases (which included probable DF, DHF1 and 2, DHF/
DSS 3 and 4) were analyzed and compared between the
This case control study was conducted in Kanchi study and control group. Various clinical and laboratory
Kamakoti Childs Trust Hospital, Chennai, between parameters were analyzed to assess their value as
August 2001 - February 2002. During this period, Chennai predictors for bleeding in dengue infection. Univariate
City witnessed an unprecedented epidemic of Dengue (95% confidence interval (CI) with significant unadjusted
viral infection. All infants and children hospitalised with odds ratio (OR) and logistic regression (95% CI with
a diagnosis of Dengue fever (DF), DHF and DSS were significant adjusted OR) were used for the statistical
enrolled for the study. "I~e criteria for hospitalisation were analysis using SPSS package version 10.
as follows: Restlessness/lethargy, hematocrit >40%,
increasing hematocrit (>10% from baseline) with RESULTS
decreasing platelets or platelet count < 1,00,000 / m m 3,
respiratory distress, profuse vomiting, oliguria, seizure, One hundred and thirty two cases of dengue infection
were included in the study. Of which 60 (Probable DF=4,
Reprint requests : Dr. So. Shivbalan, F-49, Plrst Main Road,
AnnanagarEast,Chennai-600102. DHF-I=O, DHF-II=21, DHF/DSS-III=16, DHF/DSS-IV=19)

Indian Journal of Pediatrics, Volume 71~anuary, 2004 33

 So. Shivbalan et al

patients fulfilled the criteria for study group with a mean TABLE2. Predictive Factors for Spontaneous Bleeding Derived
age of 5.26 _+4.27years and 72 (ProbableDF=40,DHF-I=12, from Logistic Regression Analysis
DHF-II=O, DHF/DSS-III=14, DHF/DSS-IV=6) patients Parameter OR 95% CI
fulfilled the criteria for control group with a mean age of
4.45 + 3.65 years. Platelet count 50,000/mm3 4.129 1.830-9.314
Serological studies for d e n g u e (IgG and IgM) were ALT* 2.952 1.306-6.674
Biphasic pattern of fever 2.811 0.948-8.338
carried out in 104 cases (46 in s t u d y g r o u p a n d 58 in Abdominal pain 2.401 0.928-6.208
control group). In 22 cases serological tests were done
within 5 days of illness. IgM was positive in 49 (47.1%), *ALT >3 times the normal value
IgG was positive in 9 (8.65%), both of these were positive
in 35 (33.65%) while both were negative in 11 (10.57%). Of
these s e r o l o g y w a s p o s i t i v e in all cases of p r o b a b l e approach. There were 4 factors of which 2 w e r e m o r e
dengue fever. To judge the ability to predict spontaneous significant, as g i v e n in Table 2. The c o m b i n a t i o n of
bleed in dengue infection the s t u d y and control group biphasic pattern of fever with hemoconcentration, platelet
were c o m p a r e d and the odds ratio (OR) was calculated count less than 50,000 and elevated ALT has a sensitivity
using univariate analysis for various clinical and labora- of 79.2 and specificity of 64.7, with a positive predictive
tory parameters. These were later confirmed with logistic v a l u e of 70 a n d a n e g a t i v e p r e d i c t i v e v a l u e of 75 in
regression analysis. The parameters assessed were : predicting spontaneous bleeding in dengue.

Clinical data : Age, sex, duration of fever, biphasic fever, DISCUSSION

rash, myalgia, vomiting, abdominal pain, cough, edema,
a b d o m i n a l distension, r e s p i r a t o r y distress, low pulse D e n g u e infection like a n y other viral illness is a self-
v o l u m e , low b l o o d p r e s s u r e , n a r r o w pulse pressure, limiting disease. Shock and bleeding are two d r e a d e d
ascites, pleural effusion, h e p a t o m e g a l y w i t h / w i t h o u t manifestations of concern complicating the outcome of
tenderness, splenomegaly. the disease. Shock is a clinical entity, which is recognized
Laboratory data : Hemoconcentration, hypoprotenemia, b y m o n i t o r i n g the vital p a r a m e t e r s . This r e s p o n d s to
Serum Aspartate aminotransferase (AST), Serum Alanine appropriate and timely fluid resuscitation. Bleeding m a y
aminotransferase (ALT), Prothrombin Time (PT), Partial be variable and occurs at various sites (skin, mucosal,
thromboplastin time (PTT), Platelet count, chest X-ray external or internal). Dengue fever also m a y be associated
(pleural effusion), ultrasonographic findings. with unusual bleeding manifestations. Internal bleeding
Among the above cited data 13 parameters were found m a y be difficult to recognize in the presence of hemocon-
to be significantly associated with spontaneous bleeding centration. A drop in hematocrit with no clinical improve-
manifestations in dengue infection b y univariate analysis ment despite adequate fluid administration m a y indicate
as given in Table 1. Abnormal PT of >189times the control significant internal hemorrhage? Bleeding manifests at or
value was found only in study group (8 out of 54 cases in after the time of its occurrence. This is associated with
which it was done). Of the 13 factors, only those factors increased morbidity, and can be fatal if it goes unnoticed.
available in m o r e t h a n 80% cases w e r e t a k e n u p for Multiple bleeding manifestations a n d DSS h a v e been
analysis. Hence the corrected total n u m b e r of cases was associated with higher mortality. 2
104 as compared to the initial 132 cases. These cases were B l e e d i n g in d e n g u e i n f e c t i o n is m u l t i - f a c t o r i a l . 4
subjected to logistic regression using likelihood ratio Reduced platelets and fibrinogen were two most
prominent hemostatic defects responsible for bleeding
TABLE1. Predictive factors for spontaneous bleeding derived from d i s o r d e r in D H F . s S t u d i e s a c r o s s the w o r l d h a v e
Univariate analysis documented various bleeding manifestations in dengue
infection (Table 3). Menorrhagia has been reported as an
Parameter n OR 95% CI
unusual manifestation of DFY G u m bleeds h a v e b e e n
Biphasic pattern of fever 132 3.095 1.168 - 8.201 reported to be the commonest. 12 Our s t u d y g r o u p h a d
Abdominal pain 132 2.317 1.016 - 5.285 various unusual bleeds in addition to the ones documen-
Low pulse volume 132 3 1.472 - 6.114 ted earlier (Table 3).
Narrow pulse pressure 132 4 1.61 - 9.938 Transfusion requirement has been significantly
Respiratory distress 132 2.333 1.093 - 4.982
Abdominal distension 132 3.455 1.685 - 7.084 correlated with the occurrence of bleed, especially GI
Liver tenderness 132 3.242 1.471 - 7.143 bleed, whereas platelet count was not related to trans-
Hemoconcentration 130 3.776 1.822 - 7.826 fusion requirements. 13Thus, predictors for the occurrence
CXR - pleural effusion 98 2.485 1.068 - 5.782 of b l e e d helps the clinician a n t i c i p a t e the p r o b l e m ,
USG - pleural effusion 63 3.033 1.034 - 8.899 prognosticate, decide about referral to a higher center for
Platelet < 50,0.00/mm 3 132 4.057 1.957-8.409
ALT 124 2.649 1.279 - 5.48 monitoring and blood product transfusion when
FIT 106 2.72 1.24 - 5.967 required.
Age, sex, duration of fever, rash, hypotension, edema,

34 Indian Journal of Pediatrics, Volume 71~anuary, 2004

 Predictors of Spontaneous Bleeding in Dengue

TABLE3. Bleeding Profile in Various Studies

Present study* Delhi 6 Delhi 7 Delhi2 Indonesia 8 Chennai 9 Thailand ~~ Cuban "~
% % % % % % % %

Petechiae 46.66 10.66 33 23 29.7 24 46 50

Purpura 2O
Epistaxis 8.33 24.34 36 38.4 39.1 19 9
Gum bleeds 8.33 4.74 9 28.06 ] - ] -
Hemetemesis 26 23.74 39 22.86 76 I12 /30
Melena 21.66 4.4 10
Hemoptysis -
Subconjunctivat 6.66
-hemorrhage
Hematuria
Retroperitoneal-bleed
Hemoperitoneum 1.66
Mennorhagia
*Only 9 cases (1-purpura, 3- petechiae & purpura, 5-petechiae) had isolated skin bleed, others had more than one bleeding manifestation.

ascites, hepatosplenomegaly and h y p o a b l u m i n e m i a were REFERENCES

not f o u n d to be significant risk factors in predicting bleed
1. Halstead SB. Dengue fever/dengue hemorrhagic fever. In
in the present study. In a study conducted in Philippines Behrman RE, Kliegman RM, Jenson HB, eds. Nelson Textbook of
to d e t e r m i n e predictors of bleeding, PT a n d PTT w e r e Pediatrics, 16~ edn. WB Saunders Company, 2000; 1005-1007.
f o u n d to p r e d i c t bleeds. Patients with platelet c o u n t s 2. Anuradha S, Singh NP, Rizvi SN, Agarwal SK, Gur R, Mathur
< 5 0 , 0 0 0 / m m 3 h a v e b e e n r e p o r t e d to h a v e a six-fold MD. The 1996 outbreak of dengue hemorrhagic fever in Delhi,
India. South East Asian J Tropical Medicine Public Health 1998;29
increase in mortalilty. 14 Risk of s p o n t a n e o u s bleed has
(3): 503-506.
been associated with a platelet count <20,000. ~5 A s t u d y 3. World Health Organization. Dengue Hemorrhagic Fever:
c o n d u c t e d is D e l h i , c o n c l u d e d t h a t t h e r e w a s no Diagnosis, Treatment and Control, Geneva, 2 "d edn. World
significant difference in bleeding manifestations a m o n g Health Organization, 1998.
thrombocytopenic and non-thrombocytopenic patients in 4. Huang YH, Liu CC, Wang ST, Lei HY, Liu HL, Lin YS et al.
D e n g u e ? The present study validates the predictive value Activation of coagulation and fibrinolysis during dengue virus
infection. ] Medical Viroloy,y 2001; 63 (3) : 247-251.
of abnormal coagulation profile (PT, PTT) and thrombo- 5. Mitrakul C, Poshyachinda M, Futrakul P, Sangkawibha N,
cytopenia (<50,000) in d e n g u e cases with bleeds. Ahandrik S. Hemostatic and platelet kinetic studies in Dengue
Increased levels of aminotransferase e n z y m e occur in Hemorrhagic fever. American l Tropical Medicine Hyyfiene 1977;
D H F (61% AST, 56% ALT). 7'~6Increased levels of AST, 26 : 975-984.
ALT, G-GT have been observed in patients with episodes 6. Gomber S, Ramachandran VG, Kumar S, Agarwal KN, Gupta
P, Dewan DK et al. Hematological observations as diagnostic
of bleeding. ~7,~sGI bleed in DHF has been associated with
markers in dengue hemorrhagic fever - A reappraisal. Indian
e l e v a t e d a m i n o t r a n s f e r a s e s levels, AST b e i n g m o r e Pediatr 2001; 38: 477-481.
e l e v a t e d t h a n ALT. 16,17,19 H e p a t o m e g a l y a n d liver 7. Aggarwal A, Chandra J, Aneja S, Patwari AK, Dutta AK. An
dysfunction (elevated hepatic enzymes) were more epidemic of Dengue Hemorrhagic fever and Dengue shock
c o m m o n in D H F w h e n c o m p a r e d to DF. ~8 T e n d e r syndrome in children in Delhi. Indian Pediatr 1998; 35 : 727-732.
8. Chairulfatah A, Setiabudi D, Ridad A, Colebunders R. Clinical
hepatomegaly had a 57% correlation with the diagnosis of
manifestations of Dengue hemorrhagic fever in children in
D H F Y We f o u n d ALT, tender hepatomegaly, abdominal Bandung, Indonesia. Annales de la Societe Belge de Medecine
pain, a b d o m i n a l distension and respiratory distress to be Tropicale 1995; 75 (4) : 291-295.
significant predictors for bleeding in dengue. 9. Sarala R, Mukundan D. Neurological manifestations in
Dengue Hemorrhagic fever. Indian Pediatr 1994; 31 : 688-690.
10. Kabra SK, Jain Y, Singhal T, Ratageri VH. Dengue
CONCLUSION
Hemorrhagic fever: Clinical manifestations and management.
Indian ] Pediatr 1999; 66 : 93-101.
The present s t u d y puts forth the following predictors to 11. McGready R, Paw E, Nosten F. Menorrhagia caused by
suggest the possible occurrence of spontaneous bleed in Dengue fever. Australian New Zealand J Obstetrics Gynaecology
dengue. PT was abnormal only in cases with spontaneous 2000; 40 (3): 345-355.
12. Agarwal R, Kapoor S, Nagar R, Misra A, Tandon R, Mathur A
bleed. A combination of (a) biphasic pattern of fever, (b)
et al. A clinical study of the patients with dengue hemorrhagic
h e m o c o n c e n t r a t i o n , (c) platelet count less than 50,000/ fever during the epidemic of 1996 at Lucknow, India. South
m m 3 a n d (d) elevated ALT, h a d a sensitivity of 79.2%, East Asian t Tropical Medicine Public Health 1999; 30 (4) : 735-740.
specificity of 64.7% with a positive predictive value of 13. Chuansumrit A, Phimolthares V, Tardtong P, Tapaneya-Olarn
70% and a negative predictive value of 75% in predicting C, Tapaneya-Olarn W, Kowsathit Pet al. Transfusion require-
ments in patients with dengue hemorrhagic fever. South East
spontaneous bleeding in dengue.
Asian J Tropical Medicine Public Health 2000; 31 (1): 10-14.

Indian Journal of Pediatrics, Volume 71---January, 2004 35

 So. Shivbalan et al

14. Chua MN, Molanida R, de Guzman M, Laberiza F. 17. Kuo CH, Tai Di, Chang-Chien CS, Lan CK, Chiou SS, Liaw YF.
Prothrombin time and partial thromboplastin time as a Liver biochemical tests and dengue fever. Am ] Tropical
predictor of bleeding in patients with dengue Hemorrhagic Medicine Hygiene 1992; 47 (3) : 265-270.
fever. South East Asian J TropicalMedicine Public Health 1993; 24 18. Wahid SF, Sanusi S, Zawawi MM, Ali RA. A comparison of the
(1): 141-143. pattern of liver involvementin dengue hemorrhagic fever with
15. Kabra SK, Jain Y, Madhulika, Tripathi P, Singhal T, Broor Set classical dengue fever. South East Asian ] Tropical Medicine
al. Role of Platelet Transfusion in dengue hemorrhagic fever. Public Health 2000; 31(2) : 259-263.
Indian Pediatr 1998; 35 : 452455. 19. Nguyen TL, Nguyen TH, Tieu NT. Res Viro11997; 148 (4): 273-
16. Kalayanarooj S, Vaughn DW, Nimmannitya S, Green S, 277.
Suntayakorn S, Kunentrasai N e t al. Early clinical and 20. Lucas GN, Amerasinghe A, Sriranganathan S. Dengue
laboratory indicators of acute dengue illness. J Infectious Dis hemorrhagic fever in Sri Lanka. Indian ] Pediatr 200; 67 (7): 503-
1997; 176 (2) : 313-321. 504.

36 Indian Journal of Pediatrics, Volume 71~January, 2004