Indian Pediatrics 1998; 35:727-732

Anju Aggarwal, Jagdish Chandra, S. Aneja, A.K. Patwari and A.K. Dutta

From the Department of Pediatrics, Kalawati Saran Children's Hospital and Lady Hardinge Medical College,
 New Delhi 110 001, India.
Reprint requests: Dr. Anju Aggarwal, Flat No. 3C, Block C2B, Janak Puri, New Delhi 110 058, India.

Manuscript received: September 8, 1997; Initial review completed: October 1, 1997;
Revision accepted: December 15, 1998

Abstract:

Objective: To study clinical profiles and outcome of children of dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS) during 1996 Delhi epidemic. Design: Retrospective study. Setting: Hospital based study. Methods: Children hospitalized from September to November 1996 were studied. All patients were diagnosed, managed and monitored according to a standard protocol. Results: One hundred and thirty four children (80 (60%) males and 54 (40%) females) were studied. Sixty (45%) children were less than 6 years of age of which 12 presented during infancy. There were 92 (67%) cases of DHF and 42 (33%) cases of DSS. Common symptoms were fever (93%), abdominal pain (49%) and vomiting (68%). The commonest hemorrhagic manifestation was hematemesis (39%) followed by epistaxis (36%) and skin bleeds (33%). Hepatomegaly was observed in 97 (72%) cases and splenomegaly in 25 (19%). Serology was positive (IgM hemaglutination antibody titres >1: 160) for dengue type 2 in 31 (80%) of 39 patients in whom sera was tested during the acute phase of illness. Mortality was 6%. Hematocrit >40% was observed in only 25 (18%) patients and hence the management protocol was based on clinical signs and symptoms and not on hematocrit. Conclusions: A management protocol of DHF/DSS in which fluid therapy is not based on haematocrit values needs to be formulated.

Key words: Dengue hemorrhagic fever, Dengue shock syndrome, Epidemic.
 

DENGUE has worldwide distribution and remains a health problem or potential threat in areas infested with Aedes aegypti mosquito(1,2). In India, the first outbreak of dengue fever was recorded in 1812, virus was isolated in 1945 and many epidemics have been reported since then. The severe fatal forms of disease, i.e., dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS) though common in South East Asia since early 1940's were reported in India for the first time from Calcutta in 1963(3). Since then epidemics of DSS/DHF have been reported from various states of India(4-9). Similar epidemics have been reported from other South East Asian countries, namely, China, Thailand and Indonesia(10-13).

During all these epidemics certain common signs and symptoms such as fever, malaise, bodyache, anorexia and bleeding manifestations have been observed. There have been differences in certain abnormal presentations, age and sex distribution and frequency of various signs and symptoms. This study was undertaken to evaluate clinical profile and outcome of children admitted with DHF jDSS, in the DHF epidemic in Delhi in 1996, who were diagnosed and managed according to the WHO protocol(14).

Subjects and Methods

One hundred thirty four children of DHF /DSS who were admitted in the Pediatric Wards of Kalawati Saran Children's Hospital, New Delhi from September to December, 1996 were studied. An attempt was made to diagnose all patients as DHF / DSS according to the WHO criteria(14). Case definition of DHF consisted of patients who had fever, hemorrhagic manifestations including at least a positive tourniquet test, thrombocytopenia (platelet count < 100000/cu mm) and hemoconcentration (hematocrit increase by >20%) or objective evidence of increased capillary permeability. Definition of DSS consisted of all the above criteria with hypotension or narrow pulse pressure (20mm Hg or less). All the patients fulfilled all the above criteria except hemoconcentration which could be documented in only 54/134(40%) of the patients. This was due to non-availability of pre-illness hematocrit, modification of hematocrit by treatment (blood transfusion, colloid transfusion, etc.) and non availability of serial hematocrits in some patients due to retrospective nature of study.

Of all the patients diagnosed as DHF/DSS those who were admitted to the pediatric wards had one or more of the following: (i) bleeding manifestations; (ii) signs and symptoms of shock; and (iii) pulse pressure < 20mm Hg. No particular value of platelet count was taken as criteria for admission. All patients underwent a detailed clinical evaluation and relevant investigations. Severity of, DHF/DSS was graded according to the WHO criteria(14).

Hemoglobin, hematocrit, platelet count and total leukocyte count was done in all cases. Serial hematocrit and platelet count was done as and when required in certain cases. Liver function test, chest X-ray and ultrasound abdomen was done in some patients. Cerebrospinal fluid examination was done in cases presenting with convulsions and altered sensorium. Blood was collected from 39 patients during acute phase of illness' and serological diagnosis was carried out at National Institute of Communicable Diseases, Delhi. IgM hemagglutination antibody titres (>1:160) for dengue type 2 were interpreted as positive result.

An attempt was made to manage and monitor all patients according to the WHO protpcol(14). Children were discharged when they were afebrile, maintained vital signs without intravenous fluid therapy and did not have any bleeding manifestations. Data of clinical profile, laboratory finding and outcome of these 134 patients were analyzed.

Results

Of 134 patients admitted in pediatric wards 92 (67%) were diagnosed as DHF and 42 (33%) as DSS. Age distribution of study patients is shown in Table 1. There were 80 (60%) male and 54 (40%) females among the study patients. Male to female ratio was 1.4:1, in cases of DSS and 1.8:1 in cases of DHF. The youngest child was aged 3 months and 12 (9%) cases had presented with DHF during infancy. There was no case of DSS in infancy. One hundred and fourteen (85%) patients presented .between 3rd and 7th day of fever, 9 (7%) before 3 days and 12 (9%) after 7 days of fever. Mortality was 6%. All 8 of these were cases of DSS and expired within twenty four hours of admission. Four of these were males and four females aged 2-6 years and all had presented with Grade IV DHFlDSS. Clinical features of the study patients is shown in Table II.
 

All 134 children had a platelet count less than 100,000/cu mm. Majority of patients, i.e., 47'0 had platelet count between 25000-50000/cu mm (Table III). X-ray chest was done in 25 patients of which 9 (36%) revealed evidence of pleural effusion. Ultrasound abdomen was' done in 8 patients of which two revealed evidence of ascites. Clinical evidence of pleural effusion or ascites was not found in any patient.

Mean duration of stay in cases of DSS was 6.17 days and DHF 3.63 days. Peripheral smear for malarial parasite and blood culture done in cases with splenomegaly did not reveal any abnormality. Cerebrospinal fluid examination done in cases of convulsions and altered sensorium did not reveal any abnormality.

Discussion

In the present study 601134 (45%) cases presented with DHF lDSS at the age of 6 years or less.. In two earlier studies from Delhi describing 1988 epidemic all cases of DHF lDSS were aged 6 years or more(4,5). In the present study we also had 12 (9%) cases who presented with DHF during in- fancy with the youngest child being 3 months of age. It is generally believed that 90% of DHF lDSS cases have circulating antibodies against dengue either by previous infection or passive transfer from the mother(15). Since in the present epidemic we had cases belonging to younger age groups it shows endemic nature of dengue virus so that children acquire antibodies at an early age. Presence of shock and hemorrhagic manifestations seen during infancy can be attributed to passively transferred circulating antibodies from the mother. Some authors have suggested that viral virulence is a risk factor for DHFlDSS independent of pre-infection antibody status(16). Therefore, cases of dengue in earlier age groups could also be due to increased virulence of virus during the present epidemic.

 

IgM antibody titres of >1:160 against Dengue type 2 were seen in 31 (80%) cases out of 39 in whom serology was sent during 'acute phase of illness. Usually samples for IgM antibody detection should be collected after 5 days and not later than 6 weeks of onset of fever(17). In eight serologically negative cases samples were collected before 5 days of fever hence the negative report. Serological diagnosis in rest of the cases and demonstration of rising or falling antibody titres after 14 days was not done due to lack of resources and poor follow up of patients. Nearly 100% positivity of samples tested after 5 days of fever confirms the presence of epidemic due to Dengue type 2.

Common symptoms were fever, vomit- ing and abdominal pain' as reported in previous studies(4-6). Hepatomegaly was observed in 72% of our cases, reported between 71-79% by others(4,5). Variable incidence of splenomegaly has been reported between 6-27%(4,6,11) compared to 19% in the present study. No case was proved to have malaria or enteric fever in the present study. Convulsions and altered sensorium were seen in 8% cases, though a higher incidence of 12.5 to 20% has been observed in two previous studies from Delhi(4,5). Incidence of convulsions was similar to that reported from China and Thailand i.e. 3-7%(10,11). No other abnormal clinical manifestation was seen during the present epidemic. Patients with Grade I severity were usually not admitted and hence most of patients belonged to Grades II and III as seen in previous studies(4-6).

All patients had a platelet count < 100,000/cu mm and hence fulfilled the WHO criteria of thrombocytopenia for diagnosis of DHF/DSS. The WHO protocol for management of DHF/DSS requires 1-2 hourly determination of hematocrit values and adjustment of fluid therapy accord- ingly(14). In the present study, only 25 (18%) cases had hematocrit >40% and since preillness hematocrit was not known it was difficult to document hemoconcentration. Fluid therapy was monitored according to clinical criteria and vital signs because if preillness hematocrit is not known it is difficult to document the percentage of hemoconcentration and monitor fluid therapy by hematocrit. Though serial hematocrit values may help in increasing or decreasing fluids, the amount to be in- creased or decreased depends on percentage of hemoconcentration. Our observations suggest that hemoconcentration may not be a good indicator for diagnosis and monitoring of fluid if pre-illness hematocrit is not known, particularly because there is a high prevalence of anemia in the population.

The commonest hemorrhagic manifestation in the present study was hematemesis (29%), followed by epistaxis (27%). Positive tourniquet test was observed in 32% patients. Positivity of tourniquet test is very variable in previous studies ranging from 13-66%(5,10). Positivity of tourniquet test was reported to be 100%,80% and 62% in patients of Grades I, II and III, respectively(18). This maybe due to various techniques and timing of the test. Tourniquet test may be negative during the stage of shock and is positive only for a few days during the course of illness(4).

Mortality in the present study was 6% compared to 12-13% in previous shidies(4,5). This could be due to delay in recognition of epidemic in previous years or delay in seeking medical attention. Due to increased awareness, endemic nature of disease, better transport facilities and case management according to the WHO guide- lines, mortality in the present epidemics was low. Since majority of patients presented with hemorrhagic manifestations they were recognized and treated at an early stage. All patients who expired belonged to Grade IV DHF/DSS according to the WHO classification.

During the present epidemic, a higher proportion of Young children < 6 years of age were affected suggesting either a increase in endemicity or virulence of dengue virus. Also, a management protocol of DHF/DSS in which fluid therapy is not based on hematocrit values needs to be formulated.