Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi, India S. Sharma*, S.K. Sharma *, A. Mohan*, J.Wadhwa*, L.Dar**, ″
By
*Departments of Medicine, **Microbiology and ***Radiodiagnosis, All India Institute of Medical Sciences New Delhi 110 029, India.
S. Thulkar*** and J.N. Pande*
Abstract An epidemic of dengue fever occurred in Delhi which started in the middle of August 1996. Virus serotype DEN-2 was found to be responsible for this outbreak. During this epidemic, 98 adult patients diagnosed to have dengue haemorrhagic fever (DHF) (n=75) and dengue shock syndrome (DS) (n=23) were admitted to one of the medical units of the Department of Medicine at the AIIMS hospital, New Delhi. Fever (100%), body aches (45.9%), abdominal pain (38.7%), purpura (33.6%), epistaxis (32.6%), melaena (26.5%), haematemesis (22.4%) and ecchymoses (20%) were commonly present symptoms. Mu capture ELISA IgM antibodies for serodiagnosis of dengue virus infection was positive in 23 of the 27 patients tested. At the time of admission, 94 patients had a platelet count below 100,000/mm3. Four patients with haemorrhagic manifestations had an initial platelet count of >100,000/mm3. Severe thrombocytopenia (platelet count <20,000/mm3 ) was present in 43.8% of the patients. The ultrasound tests showed pleural effusion in 10
_________________________ ″
Address for correspondence: Dr S.K.Sharma, Professor, Department of Medicine, All India Institute
of Medical Sciences, New Delhi 110 029, India, Fax: 91-11-686 2663, E-mail:surensk@hotmail.com
20
Dengue Bulletin – Vol 22, 1998
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
of the 12 patients and ascites in five patients tested when they were not clinically evident. Eight patients died. Logistic regression analysis revealed that patients who died had a lower haematocrit, higher respiratory rate, and extensive ecchymoses and were often in shock at the time of presentation. There was poor correlation between thrombocytopenia and ecchymoses. No correlation was found between platelet count and the outcome of multivariate analysis. Key words: Dengue haemorrhagic fever, Dengue shock syndrome, India.
Introduction Dengue fever has been known for more than a century in the tropical areas of the South-East Asia and Western recognized dengue reported 1963(3). Pacific in the regions(1). Philippines fever (DEN-2) The in was in haemorrhagic form was, however, first 1953(2). In India, the first epidemic of haemorrhagic in Calcutta Delhi
Material and methods During the 1996 dengue epidemic, 721 adults and children were hospitalized at the All India Institute of Medical Sciences (AIIMS) hospital, New Delhi, India. Of these, 98 adult patients diagnosed to have DHF were admitted to one of the medical units (unit II) of the Department of Medicine and form the basis of this presentation. Medical records of these 98 adult patients were analysed. The criteria used for diagnosing DHF included: an acute febrile illness of less than 10 days’ duration with purpuric spots or mucosal bleeding, hepatomegaly, and shock haematocrit (any (any two), or one). For plus raised the thrombocytopenia
witnessed
dengue 2)(2), An
epidemics in 1967 (DEN-2), 1970 (DEN-1 and 3), 1982 (DEN-1 and 1988 (DEN-2)(4,5) and 1991(6).
epidemic of dengue fever occurred in Delhi starting in the middle of August 1996. found Virus to be serotype DEN-2 for was this responsible
outbreak. In this paper, we describe clinical manifestations and laboratory investigations and outcome in 98 patients with DHF. Various prognostic factors affecting survival are also described.
diagnosis of dengue shock syndrome (DSS), in addition to the previously mentioned criteria, the presence of 21
Dengue Bulletin – Vol 22, 1998
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
hypotension or narrow pulse pressure was required. Of these 98 patients, 23 satisfied the criteria for the diagnosis of DSS. In all patients, a detailed history was taken and clinical examination was done In at all admission the patients, and the subsequently during the stay in the hospital. platelet count was estimated daily. Platelet function and platelet antibody tests were performed in six and three patients respectively. Haematocrit was estimated at admission in 71 patients and it was serially estimated in 49 patients on a daily basis until the time patients became afebrile for 1-2 days. Prothrombin time and activated partial thromboplastin time were estimated in seven patients. Serum biochemistry, chest radiograph, and ultrasound of chest and it abdomen was were feasible. done IgM whenever
Treatment Intravenous fluids, platelet-rich plasma (PRP), fresh whole blood and fresh frozen plasma were infused as and when considered necessary. In patients with isotonic dehydration, 5% dextrose, Ringer’s lactate, and normal saline were used. Bicarbonatecontaining solutions were given to patients having persistent fluid losses due to diarrhoea. We did not administer PRP unless the patient was actively bleeding or was in shock or had a platelet Fresh count below 20,000/mm3. whole-blood
transfusion was given when patients had either massive haemorrhage or a falling PCV. Nasogastric suction was avoided Paracetamol wherever was given possible. as an
antipyretic agent when temperature was higher than 390C, and no more than six doses were administered in 24 hours. Combinations of paracetamol, ibuprofen, salicylates and intramuscular injections were avoided.
antibodies to dengue specific virus were tested using Mu capture ELISA in the acute phase serum samples in 27 patients (reagents kindly gifted by Dr D.J. Gubler, Centers for Disease Control (CDC), Atlanta, USA).
Statistical analysis Modified students t-test applicable to two groups with unequal variances was applied to study the association
22
Dengue Bulletin – Vol 22, 1998
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
between a numerical variable and the outcome. were Univariate carried out and for multivariate analyses logistic regression
Variable Body aches Rash Purpura Ecchymoses
Number ( %) 45 (45.9) 36 (36.7) 33 (33.6) 19 (19.4)
predicting survival.
Results The patients’ mean age was 26.3 (SD 10) years. There were 74 males. The mean duration of symptoms was 5.6 (SD 2.6) days. The mean duration of fever was 4.88 days (range 3 to 8 days) in patients who died, whereas its mean duration was 5.53 days in survivors. Other clinical features at initial presentation are outlined in Tables 1 and macular extensive 2. skin The rash haemorrhagic (36.7%) and manifestations included erythematous ecchymoses (19.4%).
Bleeding from other sites* Abdominal pain Haematemesis Melaena Epistaxis Impaired consciousness Shock Lymphadenopathy Hepatomegaly Splenomegaly Jaundice Renal failure Pleural effusion ARDS rectal and vaginal bleeding
51 (52.0) 38 (38.7) 22 (22.4) 26 (26.5) 32 (32.6) 05 (05.1) 23 (23.5) 03 (03.1) 20 (20.4) 08 (08.2) 01 (01.0) 01 (01.0) 08 (08.2) 01 (01.0)
* Haemoptysis, conjunctival haemorrhage,haematuria, ARDS = Acute respiratory distress syndrome
Haematemesis occurred in 22.4% and melaena occurred in 26.5% of the patients. In 10.2% of the patients both haematemesis and melaena occurred. Epistaxis was noted in 32.6% of the patients. Haematuria, haemoptysis and bleeding from other sites such as gums, rectum, and vagina occurred in 51 patients (52%) (Table 1). Table 1. Clinical presentation in 98 patients with DHF/DSS Dengue Bulletin – Vol 22, 1998
Clinical examination revealed a mean pulse rate of 93.1 per minute (SD 14.2). Mean systolic blood pressure was 106.6 mm Hg (SD 17.0) and mean diastolic blood pressure was 70.2 mm Hg (SD 12.2). Shock was present in 23.5% of the patients. The mean respiratory rate at presentation was 16.3 per minute (SD 6.6). Tachypnoea (respiratory rate >18) was present in 23
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
17 of the 63 patients. All the patients who died had tachypnoea. Renal failure and acute respiratory distress syndrome (ARDS) were present in one patient each. Hepatomegaly (20.4%), splenomegaly (8.2%), and lymphadenopathy (3.1%) were also found. Dengue virus-specific IgM Mu capture ELISA was positive in 23 of the 27 patients (85.1%).All these patients had fever for five or more days. The haematological and biochemical parameters are shown in Table 2. Table 2. Laboratory parameters in 98 patients with DHF/DSS Variable Haemoglobin g/dl [n=88] Haematocrit % [n=71] Total leukocyte count /mm3 [n=73] Platelet count /mm3 [n=98] Serum bilirubin mg/dl [n=34] AST (IU/l) [n=43] ALT (IU/l) [n=43] Serum alkaline phosphatase (IU/l) [n=29] Mean (SD) 11.6 (2.9) 39.0 (8.6) 7163 (10365) 30000 (25000) 0.9 (0.6) 274 (634) 143 (242) 196 (155)
The
mean
haemoglobin
and
haematocrit were normal in a majority of patients. Raised haematocrit (>48 %) was observed in 6 of the 71 patients tested. Fourteen of the 49 patients demonstrated haematocrit >20% over the increase lowest in value
recorded, thus fulfilling the criteria for the diagnosis of DHF as per the WHO definition. Leukopenia (<4000/ mm3) was observed in 22 of the 73 patients for whom the total leukocyte count was available. At the time of initial presentation, 94 patients had a platelet count below 100,000/mm3. In four patients, the platelet count was >100,000/mm3. Severe thrombocytopenia (platelet count <20,000/mm3) was present in 43.8% of the patients. There was no correlation between the platelet counts and the presence of ecchymoses. Platelet function was studied in those six patients with haemorrhagic manifestations who had platelet count in the normal range. Of these six patients, a reduced platelet aggregation was found with adrenaline in five patients and with adenosine diphosphate (ADP) in three patients respectively. normal One patient showed with platelet aggregation
adrenaline and ADP. In another three patients, antiplatelet antibodies were demonstrable. Coagulation profile
24
Dengue Bulletin – Vol 22, 1998
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
(n=7) was normal in all the patients tested. Serum bilirubin was elevated in three of the 34 patients in whom it was tested. Serum aspartate aminotransferase (AST) (n= 43) and alanine aminotransferase (ALT) (n= 43) were elevated in 88.4% and 76.7% of the patients respectively. Serum alkaline phosphatase was elevated in three of the 29 patients in whom it was tested. One patient had a raised level of blood urea and serum creatinine at initial presentation. After haemodialysis the values became normal. Twelve patients were studied for the presence of pleural effusion using chest radiograph pleural and effusion ultrasound radiograph in three examination. revealed Chest
of these patients had any clinical evidence of pleural effusion or ascites. Collection of fluid in the pelvic cavity was not observed in any of the patients studied ultrasonographically. Univariate revealed platelet that count, analysis the (Table 3) rate, and
respiratory haematocrit,
haemoglobin were significant predictors of outcome. Logistic regression analysis for the predictors of survival showed that tachpnoea, ecchymoses and shock were of significance in predicting (Table 4). Table 3. Predictors of survival in 98 patients with DHF/DSS [Univariate analysis] Variable Respiratory rate/ minute* Platelet count (× 100 000 /mm3)* Haemoglobin g/dl* Haematocrit %* Ecchymoses Purpura Shock 12.9 (2.8) 36.5 (8.6) 14/76 28/62 18/72 10.0 (3.5) 25.6 5/8 5/8 5/8 p=0.03 RR=6.93 RR=3.28 [0.84-12.9] RR=5.43 p=0.05 Alive (n=90) 15.5 (4.5) 1.12 (0.9) Dead (n=8) 27.5 (4.5) 0.56 (0.48) p=0.01 Significance p=0.05
the
survival
of
patients
patients and consolidation in one. Pleural effusion was detected by ultrasound in 10 of these 12 patients (83%). The pleural effusion was rightsided and was bilateral in two patients. Unilateral left-sided pleural effusion was not found in any of the patients. Ultrasound examination of the abdomen revealed ascites (n=5), perihepatic collection of fluid (n=1), gall bladder wall thickening (n=5) and sludge in the gall bladder (n=1). None
(1.3)
[1.81-26.49]
Dengue Bulletin – Vol 22, 1998
25
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
[1.4-21.03]
Malaysia also, dengue fever occurs in young adults. Further, increased mortality from this disease has also occurred in young adults in both countries(7,9). Studies from Thailand indicate that the mortality from DHF occurs as a result of hypovolaemic shock which, in turn, results from (1,10)
* Expressed as mean (SD); RR=relative risk
Table 4. Logistic regression analysis for predictors of survival in 98 patients with DHF/DSS Coefficient 7.20 -0.17 0.84 0.730.96 -2.14 -2.00 0.12 0.14 0.020.85 0.021.04 0.05 0.03 0.01
Parameter
OR
95 % ConfiP value dence interval
increased leading to
Constant Respirator y rate Shock Ecchymose s
capillary permeability
raised haematocrit. Therefore, WHO has included raised haematocrit as one of the important criteria for the diagnosis of DHF. Ideally, serial estimation of haematocrit should be done in patients with suspected DHF. However, in the present epidemic, haematocrit was not estimated in all the patients In because workload the of on the the study, tremendous
No. of observations = 98; Pseudo r2 = 0.43; AU ROC = 0.92 OR = Odds ratio
Discussion This clinical presentation profile of describes DHF in the adult
laboratory.
present
haematocrit was >48% in six of the 71 patients in whom it was tested; and only 14 cases demonstrated >20% increase in haematocrit over the lowest value recorded. This may be due to the fact that haemoglobin in most of the patients was low (possibly due to severe haemorrhage or chronic anaemia). This could also have been due to the fact that most of the patients received intravenous fluids before they sought treatment at the
patients. Most of these patients were young adults. A similar trend was reported in studies from Singapore and Malaysia (9). (7,8)
This may be due to
the fact that adults are not immune to all strains of the dengue virus. In Singapore where vector control measures are being carried out since 1973, the mean age of occurrence of DHF had increased from 14 years in 1973 to 28 in 1994. Similarly, in 26
Dengue Bulletin – Vol 22, 1998
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
AIIMS Jakarta,
hospital.Other Indonesia(10),
studies
from
also seen in some of our patients, and the possible explanation for this could be that the platelets align themselves onto the ‘leaky endothelial cells’ which are lining the blood vessels damaged by the virus, and with the recovery, these platelets return to circulation in the coagulation profile in (12).
had also shown
that one-third of the confirmed fatal cases of dengue with massive gastrointestinal from the (11,12)
haemorrhage Philippines
did and
not Kuala
manifest haemoconcentration. Reports Lumpur also showed signifycant
haemoconcentration in only 39.5% and 22% of the patients respectively. Bleeding from various sites was found in 70 of the 98 patients. The cause of bleeding could have been thrombocytopenia, platelet dysfunction consumption (13).
We did not encounter any abnormality our patients. This finding is in agreement with the observation that coagulation abnormalities are uncommon in DHF. Hepatomegaly was observed in 20.4% of the patients in the present study. Hepatomegaly in (3)
coagulopathy, capillary fragility(12) or Although constant and and thrombocytopenia between found the was a
has
been in
described in Calcutta the
16.8% (14),
patients
finding, there was poor correlation thrombocytopenia platelet platelet count bleeding diathesis. No correlation was between ecchymoses, indicating thereby that abnormal than aggregation in absolute rather reduction
Vishakapatnam Philippines
in 22.2% patients Most of the
and in 13.5% patients in (11).
patients in the present study in whom liver functions were tested had raised ALT and AST levels. In the study in Malaysia, five of the 9 cases where liver (12).
numbers was the cause of bleeding diathesis. However, one patient with normal platelet bleeding. indicates platelet count and of normal with antifunction presented
profile
was
done
showed
significant liver enzyme abnormalities Dengue virus-induced damage to the hepatocytes, hypoxia, shock or associated liver disease have all been postulated to be the pathogenetic mechanisms for the occurrence of transaminits in patients with DHF(12). Splenomegaly was found in 8.2% of
Demonstration that platelet
platelet antibodies in three patients destruction could also be immunogenic. Rapid fluctuation in daily platelet count was Dengue Bulletin – Vol 22, 1998
27
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
our cases whereas in Calcutta, it was observed in 9.3% of the cases (3).
Gram-negative sepsis could have been the causes of ARDS in this patient. One patient developed acute renal failure, and after haemodialysis this patient made complete recovery. Prolonged shock would be the possible cause of renal failure in this patient. Pleural effusion or ascites demonstrable by ultrasonography in DHF patients with no clinical evidence of these findings has been reported earlier in paediatric patients with DHF (16).
Erythematous, morbilliform macular or maculo-papular rash was found in 36.7% of the patients, whereas Krishnamurty et al. 27.1% of the In fever Visakhapatnam. haemorrhagic Calcutta diffuse (3), (14)
found it in in of from flush. series
patients the cases
40% of the patients had erythematous
Lymphadanopathy was found in only 3% of our patients compared to 10.6% reported in the study from the Philippines(11). Impaired consciousness was found in 5 of the 98 (5.1%) patients in the present study. In a study from Thailand (15),
In the present study, ultrasound
picked up pleural effusion in 10 of the 12 patients tested and has been found to be more in sensitive than chest this radiograph demonstrating
finding. Gall bladder wall thickening and perihepatic fluid collection have been found to predict the development of shock because of (17).
altered
sensorium no
had
been of these be
reported possible
and
evidence for could
encephalitis was found on autopsy. The explanation symptoms neurological
These findings have leakage and
been postulated to have developed plasma haemmorhage. In the present study, one of the five patients in whom gall bladder these wall thickening might occurred the developed shock. The presence of findings suggest diagnosis of DHF in the appropriate clinical setting. Dengue virus specific IgM Mu capture ELISA was positive in 23 of the
metabolic acidosis, severe disseminated intravascular coagulation, gross haemorrhage or oedema in brain, or hepatic dysfunction leading to encephalopathy(15). One patient in the present study developed ARDS and succumbed to the illness. Diffuse pulmonary involvement due to the virus, aspiration of gastric contents, or
28
Dengue Bulletin – Vol 22, 1998
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
27 patients in whom it was tested. All samples, with a duration of fever of five or more days, were tested by ELISA. The four negative samples had a duration of fever of 5-6 days and may not have seroconverted by this time. A study using Mu capture ELISA in patients with confirmed dengue fever (18)
2. World Health Organization. Monograph on dengue and dengue haemorrhagic fever. Geneva; 1993. 3. Aikat BK, Konar NR, Banerjee G. Haemorrhagic fever in Calcutta area. Indian J Med Res 1964;52:660-75. 4. Kabra SK, Verma IC, Arora NK, Jain Y, Kalra V. Dengue haemorrhagic fever in Delhi. Bull World Health Organization. 1992;70:105-8. 5. Acharya SK, Buch P, Irshad M, Gandhi BM, Joshi YK, Tandon BN. Outbreak of dengue fever in Delhi [letter]. Lancet 1988;2:14856. 6. Mohan Rao CVR. Epidemics of dengue fever in India. Indian J Paediatr 1987;54:11-14. 7. Goh KT. Changing epidemiology of dengue in Singapore [letter]. Lancet 1995;346:1098. 8. Chan KL, Ng SK, Chew LM. The 1973 dengue haemorrhagic fever outbreak in Singapore and its control. Singapore Med J 1977;18:81-93. 9. Dengue and dengue haemmorahagic fever. WHO Wkly Epidemiol Rec 1995;34:243. 10. Sumarmo, Wulur H, Jahja E, Gubler DJ, Suharyono W, Sorensen K. Clinical observations on virologically confirmed fatal dengue infections in Jakarta, Indonesia. Bulletin World Health Organization. 1983;61:693-701. 11. Hayes CG, Manaloto CR, Gonzales A, Ranoa CP. Dengue infection in the Philippines: clinical and virological findings in 517 hospitalized patients. Am J Trop Med Hyg 1988;39:110-16. 12. Ibrahim NM, Cehong fever at I. Ault. Kuala Dengue Lumpur
showed that 96% of the 76
blood samples drawn between 7 and 20 days after the onset of the illness were positive. In a study in Thailand, the sensitivity of this test was shown to be 97% in convalescent samples while none of the samples from the two control groups was positive(19). Eight of the 98 patients died. These patients had lower haemoglobin and haematocrit anaemia rate, levels. and/or The low haematocrit could have been due to chronic ratory and bleeding. Further, they also had faster respiextensive with the to ecchymoses, shock as survivors. presented
compared
Tachypnoea could have been due to bleeding, fluid overload, infection or shock.
References 1. World Health Organization. Dengue haemorrhagic fever: diagnosis, treatment and control. Geneva; 1986.
haemorrhagic
hospital: retrospective study of 102 cases. Br J Clin Pract 1995;49;189-91.
Dengue Bulletin – Vol 22, 1998
29
Clinical Profile of Dengue Haemorrhagic Fever in Adults during 1996 – Outbreak in Delhi
13. Haalstead
SB.
Dengue:
haematological
17. Setiawan MW, Samsi TK, Pool TN, Sugianto D, Wulur H. Gall-bladder wall thickening in dengue haemorrhagic fever: an ultrasonographic study. J Clin Ultrasound 1995;23:357-62. 18. Hayes EB, Gubler DJ. Dengue and dengue haemorrhagic 1992;11:311-17. 19. Innis BL, Nisalak S, assay A, Nimmanitya Chongswadi to S, V, Kusalerdchariya immunosorbent dengue fever. Pediatr Inf J
aspects. Semin Haematol 1982;19:116-31. 14. Krishnamurthy K, Kasturi TE, Chittipantulu G. Clinical and pathological studies of an outbreak of dengue-like Indian J illness Med in Res Vishakapatnam. 1965;53:800-12. 15. Nimmanitya S, Thisyakorn U, Hemsrichart V. Dengue haemorrhagic fever with unusual manifestations. South-East Asian J Trop Med Pub Health 1987;19:398-406. 16. Pramuljo HS, Harun SR. Ultrasound findings in dengue haemorrhagic fever. Pediatr Radiol 1991;21:100-102.
Suntayakorn S, et al. An enzyme-linked characterise dengue and infections, where
Japanese encephalitis cocirculate. Am J Trop Med 1989;40:418-27.
30
Dengue Bulletin – Vol 22, 1998