European Journal of Cardiovascular Medicine

Dengue is the most rapidly spreading mosquito-borne viral disease in the world. It is caused by four antigenically distinct serotypes of dengue virus (DENV-1 to DENV-4), which belong to the genus Flavivirus and are transmitted mainly by Aedes aegypti and, less efficiently, by Aedes albopictus [1, 2]. Modelling studies estimate about 390 million dengue infections each year, of which roughly 96 million are clinically apparent [3]. The Global Burden of Disease Study 2013 likewise documented a steep rise in dengue incidence across the tropics over the preceding decades [4]. Urbanisation, population movement, inadequate water storage and climatic change have all contributed to the spread of the vector and the virus into new regions [5].

India is hyperendemic for dengue, and all four serotypes circulate in the country [6]. A national population-based serosurvey conducted in 2017 found evidence of past infection in nearly half of the people tested, with higher prevalence in urban than in rural areas [7]. A systematic review and meta-analysis of Indian studies reached a similar conclusion about the extent of exposure [8]. Transmission follows the monsoon, and temperature and rainfall strongly influence case numbers [9]. During the large Delhi epidemics of the 1990s, dengue haemorrhagic fever and shock were described predominantly in children [10, 11]. Subsequent reports from the same region documented a gradual shift of the disease towards older children, adolescents and adults [12, 13], and adults are now a major part of the inpatient burden in Indian hospitals [14, 15].

Infection ranges from an inapparent or mild febrile illness to life-threatening disease. The clinical course is conventionally divided into febrile, critical and recovery phases [16]. Severe dengue typically declares itself around defervescence, when increased capillary permeability can cause plasma leakage, shock, bleeding and organ dysfunction [17, 18]. The original WHO classification of 1997 separated dengue fever, dengue haemorrhagic fever and dengue shock syndrome [19]. Because many patients did not fit these categories, the 2009 revision introduced the groups of dengue without warning signs, dengue with warning signs and severe dengue. Several multicentre evaluations have found the revised scheme more useful for triage and clinical decisions [20–22], and it is the basis of the national guidelines followed in India [23].

Adults with dengue differ from children in their clinical behaviour. Bleeding manifestations and hepatic involvement appear to be more common in adults, whereas isolated plasma leakage is more typical of children [24–26]. Older age and coexisting illnesses such as diabetes mellitus, cardiac disease and asthma have been linked to severe disease and death [27–30]. Indian adult data are largely descriptive and come from single hospitals, and few of them report predictors of severity using multivariable methods.

Treatment of dengue is supportive, and outcome depends mainly on timely recognition of the critical phase and careful fluid management [31, 32]. During outbreaks, tertiary hospitals receive far more patients than they can monitor intensively, and clinicians must decide quickly who needs admission, closer observation or intensive care [33]. Simple predictors available at the first contact would therefore be of direct value. We conducted this study to describe the clinical profile of adults with dengue admitted to a tertiary care hospital, to estimate the proportion who developed severe dengue, and to identify independent clinical and laboratory predictors of severe disease.

Study design and setting

This was a prospective, hospital-based observational study conducted in the Department of General Medicine of a tertiary care teaching hospital in India over 12 months, from 1 January 2022 to 31 December 2022. The hospital receives patients from the city and from surrounding urban and rural districts, and it manages dengue in dedicated fever wards, a high-dependency unit and an intensive care unit (ICU).

Study population

All consecutive adults (≥18 years) admitted with an acute febrile illness compatible with dengue and a laboratory confirmation of infection were eligible. Confirmation required a positive NS1 antigen test (ELISA or rapid immunochromatographic assay) and/or a positive IgM antibody capture ELISA, as accepted in the national guidelines [23, 34]. Patients were excluded if they had a laboratory-proven co-infection (malaria, enteric fever, leptospirosis or scrub typhus), a pre-existing haematological disorder or treatment affecting platelet count or coagulation (immune thrombocytopenia, haematological malignancy, anticoagulant therapy), incomplete records, or left the hospital against medical advice before their severity could be classified. Written informed consent was obtained from every participant or a legally authorised relative.

Sample size

The sample size was calculated with the formula n = Z²pq/d², assuming a 25% proportion of severe dengue among admitted adults, 95% confidence and an absolute precision of 5%. This gave 288 patients, and the figure was rounded up to 320 to allow for incomplete data.

Data collection

A pre-tested structured proforma recorded age, sex, residence, duration of fever before admission, presenting symptoms, comorbidities, previous dengue illness and body mass index. At admission, a detailed examination documented temperature, pulse, blood pressure, pulse pressure, hepatomegaly, signs of pleural effusion or ascites, and any bleeding. Investigations included a complete blood count with haematocrit (automated analyser, performed at admission and at least daily until recovery), liver function tests (aspartate and alanine aminotransferases, bilirubin, serum albumin), serum creatinine and electrolytes, prothrombin time with INR, a chest radiograph, and ultrasonography of the chest and abdomen for pleural effusion, ascites and gallbladder wall thickening. Other tests were done when clinically indicated to exclude co-infections. Patients were followed daily until discharge or death.

Case definitions

Patients were classified according to the WHO 2009 criteria [1]. Dengue with warning signs was defined by abdominal pain or tenderness, persistent vomiting, clinical fluid accumulation, mucosal bleeding, lethargy or restlessness, liver enlargement >2 cm, or a rising haematocrit with a rapid fall in platelet count. Severe dengue was defined by any of the following: severe plasma leakage leading to shock or to fluid accumulation with respiratory distress; severe bleeding judged by the clinician to need intervention; or severe organ impairment (AST or ALT ≥1000 IU/L, impaired consciousness, or cardiac or other organ involvement). Patients with neither criterion were classified as dengue without warning signs. For analysis, the two non-severe groups were combined and compared with the severe group.

A haematocrit rise of ≥20% over the baseline or recovery value was taken as haemoconcentration. Thrombocytopenia was defined as a platelet count <100,000/µL, and leukopenia as a total leukocyte count <4,000/µL. Transaminase elevation was graded against an upper limit of normal of 40 IU/L. Obesity was defined as a body mass index ≥30 kg/m². Hypoalbuminaemia was defined as serum albumin <3.0 g/dL.

Management

All patients were managed according to the national and WHO guidelines [1, 23]. Patients with warning signs received intravenous isotonic crystalloids titrated to haematocrit, urine output and haemodynamic status, and those in shock received timely volume resuscitation with crystalloids, with colloids or blood products added in refractory cases. Platelet transfusion was reserved for significant bleeding or very low counts, in line with guidance against prophylactic transfusion [23, 35]. Antipyretic therapy was limited to paracetamol, and NSAIDs were avoided.

Outcome measures

The primary outcome was the development of severe dengue at any time during the hospital stay. Secondary outcomes were ICU admission, need for mechanical ventilation, vasopressor support, renal replacement therapy, blood component transfusion, length of hospital stay and in-hospital death.

Statistical analysis

Data were entered in Microsoft Excel and analysed with IBM SPSS Statistics version 25.0. Continuous variables were tested for normality with the Shapiro–Wilk test and are presented as mean ± standard deviation or median (interquartile range, IQR). Categorical variables are shown as numbers and percentages. The severe and non-severe groups were compared by the independent-samples t-test or Mann–Whitney U test for continuous data and by the chi-square test or Fisher’s exact test for categorical data. The chi-square test for trend was used for ordered age groups. Variables with p < 0.05 on univariate analysis, and those considered clinically important, were entered into a multivariable binary logistic regression model with backward conditional elimination, after checking for collinearity. Results are given as adjusted odds ratios (aOR) with 95% confidence intervals (CI). Model calibration was assessed with the Hosmer–Lemeshow test, explained variance with Nagelkerke R², and discrimination with the area under the receiver operating characteristic curve (AUC). Sensitivity, specificity and predictive values were calculated for selected predictors. A two-sided p value <0.05 was considered statistically significant.

Ethical considerations

The study was approved by the Institutional Ethics Committee and conducted in accordance with the Declaration of Helsinki. Patient confidentiality was maintained throughout, and the participation of patients had no effect on their treatment.

Study population and seasonality

During the study period, 368 adult patients with suspected dengue were screened. Forty-eight were excluded (laboratory-proven co-infection in 19 [malaria 8, scrub typhus 6, enteric fever 5]; incomplete records or discharge against medical advice in 17; pre-existing haematological disorder or anticoagulant therapy in 7; and referral after more than 48 hours of treatment elsewhere with incomplete records in 5), leaving 320 patients for analysis. Cases occurred throughout the year but clustered after the monsoon. The monthly numbers were: January 3, February 2, March 2, April 3, May 5, June 9, July 18, August 34, September 78, October 94, November 56 and December 16. The three months from September to November accounted for 228 patients (71.3%), with a peak in October (29.4%). Diagnosis was based on NS1 antigen alone in 168 patients (52.5%), IgM antibody alone in 71 (22.2%) and both tests in 81 (25.3%).

Baseline characteristics

The mean age of the patients was 34.6 ± 13.2 years (range 18–78 years), and 43.1% were aged 18–30 years. Men accounted for 61.9% of the cohort. About two-thirds of the patients lived in urban areas. One or more comorbidities were present in 104 patients (32.5%), most often hypertension (19.1%) and diabetes mellitus (16.3%). Fourteen of the 122 women (11.5%) were pregnant. The median duration of fever before admission was 4 days (IQR 3–5). Twenty-one patients (6.6%) reported a previous episode of dengue (Table 1).

Table 1. Demographic and baseline characteristics of the study population (N = 320)

Age: mean ± SD 34.6 ± 13.2 years. Duration of fever before admission: median 4 days (IQR 3–5). *Percentage of female patients.

Clinical presentation

Fever was present in all patients. The commonest accompanying symptoms were myalgia or arthralgia (74.4%), headache (66.9%), nausea or vomiting (58.4%) and retro-orbital pain (37.8%). Abdominal pain was reported by 35.0%, rash by 30.0% and persistent vomiting by 29.4%. Mucosal bleeding was seen in 63 patients (19.7%). On examination, pleural effusion or ascites was detected clinically or by ultrasonography in 98 patients (30.6%), and hepatomegaly in 58 (18.1%) (Table 2).

Table 2. Presenting symptoms and signs (N = 320)

USG: ultrasonography.

Severity classification

According to the WHO 2009 classification, 146 patients (45.6%) had dengue without warning signs, 100 (31.3%) had dengue with warning signs and 74 (23.1%) had severe dengue. Among the 74 patients with severe dengue, severe plasma leakage with shock occurred in 41 (55.4%), severe bleeding in 29 (39.2%) and severe organ impairment in 38 (51.4%). Many patients met more than one criterion. Organ involvement comprised hepatic injury with AST or ALT ≥1000 IU/L in 18, acute kidney injury in 14, encephalopathy in 6 and cardiac involvement (myocarditis or arrhythmia) in 5 patients. Severe manifestations appeared at a median of day 5 of illness (IQR 4–6), usually at or soon after defervescence (Table 3).

Table 3. Classification of patients by WHO 2009 criteria and manifestations of severe dengue

*Percentage of the 74 patients with severe dengue for manifestations. Patients could have more than one manifestation.

Age and severity

The risk of severe dengue rose steeply with age. It was 17.4% in patients aged 18–30 years and 19.3% in those aged 31–45 years, but 33.3% in the 46–60 group and 54.5% in patients above 60 years (chi-square for trend, p < 0.001) (Table 4).

Table 4. Distribution of severe dengue by age group

Chi-square for trend, p < 0.001.

Comparison of severe and non-severe dengue: clinical features

Patients with severe dengue were older (41.8 ± 15.1 vs 32.4 ± 11.9 years, p < 0.001), and were more often over 45 years (39.2% vs 17.9%). Sex did not differ between the groups. Diabetes mellitus (32.4% vs 11.4%), hypertension, obesity, chronic kidney disease, chronic liver disease and any comorbidity were all more frequent in severe disease, as was a previous dengue illness (12.2% vs 4.9%). Patients with severe dengue more often presented late, after more than five days of fever (41.9% vs 19.9%). Persistent vomiting, abdominal pain, mucosal bleeding, lethargy or restlessness, clinical fluid accumulation and hepatomegaly were significantly more common in the severe group, while rash was not (Table 5).

Table 5. Univariate comparison of clinical features between severe and non-severe dengue

Values are n (%) unless stated. Chi-square or Fisher’s exact test for categorical, and independent t-test for continuous variables.

Laboratory and imaging findings

At admission, patients who went on to develop severe dengue had lower platelet counts (median 48,000 vs 92,000/µL), a higher peak haematocrit (44.6% vs 39.8%) and a lower leukocyte count (Table 6). A haematocrit rise of ≥20% occurred in 70.3% of the severe group and 24.8% of the non-severe group. Transaminases were markedly higher in severe disease, with a median AST of 312 IU/L against 86 IU/L, and AST above three times the upper limit of normal was seen in 78.4% versus 31.7%. Hypoalbuminaemia, raised creatinine, prolonged INR, hyponatraemia and gallbladder wall thickening on ultrasonography were all significantly more frequent in the severe group.

Table 6. Laboratory and ultrasonographic findings in severe and non-severe dengue

Values are n (%) unless stated. ULN: upper limit of normal; IQR: interquartile range; USG: ultrasonography.

Independent predictors of severe dengue

Thirteen variables were significant on univariate analysis. After adjustment for confounding and collinearity (fluid accumulation was closely related to hypoalbuminaemia, and abdominal pain, late presentation and raised creatinine dropped out on backward elimination), nine variables remained as independent predictors of severe dengue (Table 7). The strongest was clinical fluid accumulation (aOR 9.47, 95% CI 4.31–20.82), followed by mucosal bleeding (aOR 4.12), AST >3 × ULN (aOR 3.71), lethargy or restlessness (aOR 3.58), haematocrit rise ≥20% (aOR 3.26), persistent vomiting (aOR 2.96), diabetes mellitus (aOR 2.84), admission platelet count <50,000/µL (aOR 2.78) and age >45 years (aOR 2.31). The model fitted well (Hosmer–Lemeshow p = 0.58; Nagelkerke R² = 0.61) and discriminated strongly between severe and non-severe disease, with an AUC of 0.91 (95% CI 0.87–0.94); the bootstrap-corrected AUC was 0.89.

Table 7. Multivariable logistic regression for independent predictors of severe dengue

OR: odds ratio; CI: confidence interval. Hosmer–Lemeshow p = 0.58; Nagelkerke R² = 0.61; AUC 0.91 (95% CI 0.87–0.94).

Diagnostic performance of individual predictors

No single marker was both highly sensitive and specific, but clinical fluid accumulation performed best, with a sensitivity of 83.8%, a specificity of 85.4% and a negative predictive value of 94.6%. AST >3 × ULN and thrombocytopenia below 50,000/µL were sensitive but non-specific, whereas mucosal bleeding and lethargy were specific but less sensitive (Table 8).

Table 8. Diagnostic performance of selected admission predictors for severe dengue

PPV: positive predictive value; NPV: negative predictive value.

Clinical outcomes

ICU admission was needed for 58 patients (18.1%), of whom 52 had severe dengue. Seventeen patients (5.3%) required mechanical ventilation, 24 (7.5%) vasopressors and 5 (1.6%) renal replacement therapy. Platelet transfusion was given to 64 patients (20.0%) and packed red cells to 20 (6.3%); in both cases the majority were in the severe group. The mean hospital stay was 6.2 ± 2.8 days overall and was significantly longer in severe disease (9.4 vs 5.2 days). Nine patients died, giving an overall in-hospital mortality of 2.8% and a case-fatality of 12.2% among those with severe dengue. There were no deaths among the 246 non-severe patients. The causes of death were refractory shock with multiorgan failure (5), massive gastrointestinal haemorrhage (2), acute liver failure with encephalopathy (1) and acute respiratory distress syndrome (1) (Table 9).

Table 9. Clinical outcomes of the study population

Values are n (%) unless stated.

In this prospective study of 320 adults hospitalised with laboratory-confirmed dengue, almost one in four (23.1%) developed severe dengue under the WHO 2009 criteria, one in five needed intensive care, and 2.8% died. Nine independent predictors of severe disease were identified at or soon after admission: older age, diabetes mellitus, persistent vomiting, lethargy or restlessness, mucosal bleeding, clinical fluid accumulation, haemoconcentration, thrombocytopenia and a raised AST. Together they discriminated severe from non-severe disease with high accuracy. The proportion of severe dengue is higher than in community or outpatient series, which is expected for a tertiary referral hospital that receives the sickest patients.

The cohort was young, with a mean age of 34.6 years and more than three-quarters of the patients below 45 years. This is in line with Indian hospital series of adult dengue and with the long-recognised shift of the disease from children to adolescents and young adults in northern and southern India [34-35]. The male predominance (61.9%) has also been noted in other Indian reports and probably reflects greater outdoor and occupational exposure to Aedes and possibly differences in care-seeking [36]. Cases clustered from September to November, which agrees with the post-monsoon peak of transmission described for India and the influence of rainfall and temperature on vector density. Two-thirds of our patients lived in urban areas, consistent with the higher exposure found in urban populations by the national serosurvey.

Fever, myalgia, headache, vomiting and rash dominated the presentation, as in the classical descriptions of the disease. The relatively high frequency of abdominal pain (35.0%), persistent vomiting (29.4%) and mucosal bleeding (19.7%) in a hospitalised adult population is similar to that of other adult series. Previous studies comparing age groups suggest that adults have more bleeding and organ involvement, whereas children are more prone to isolated plasma leakage and shock. The paediatric experience in India, including the Delhi epidemics and later hospital series from eastern and southern India, also emphasised shock and effusion, with a different pattern of laboratory abnormalities from that seen here [37, 38]. Atypical presentations such as encephalopathy, myocarditis and acute kidney injury were uncommon but contributed to severity and mortality, as documented earlier from India [39–41].

Severe dengue was about two times more likely in patients above 45 years, and the proportion with severe disease climbed from 17.4% in the youngest group to 54.5% in those above 60 years. Elderly patients are known to have a higher risk of haemorrhage and death from dengue  and age was among the risk factors for fatality in a Singapore case-control study of adult inpatients . Immune senescence, reduced physiological reserve and a higher burden of comorbidity probably all contribute.

Diabetes mellitus was the only comorbidity that remained significant after adjustment (aOR 2.84). Earlier case-control studies reported diabetes as a risk factor for dengue haemorrhagic fever and for severe organ involvement in adults. Chronic low-grade inflammation, endothelial dysfunction and altered cytokine responses in diabetes may amplify the vascular leak that characterises severe dengue [42]. Hypertension, obesity, chronic kidney disease and chronic liver disease were associated with severity on univariate analysis only, probably because of their overlap with age and diabetes and the limited number of events. These findings argue for a lower threshold for admission and closer monitoring of older adults and people with diabetes.

Persistent vomiting, lethargy or restlessness and mucosal bleeding were independent predictors, which supports their position among the warning signs of the WHO classification. Clinical fluid accumulation had the largest effect in our model (aOR 9.47) and the best combination of sensitivity (83.8%) and negative predictive value (94.6%). This is biologically plausible, since plasma leakage is the central lesion of severe dengue and arises from endothelial dysfunction driven by viral NS1 protein, cytokines and other mediators [43]. Evaluations of the revised classification likewise support plasma leakage and bleeding as the features that mark patients needing intensive treatment. A haematocrit rise of ≥20% was also an independent predictor, in keeping with its use as a marker of haemoconcentration in the guidelines [44]. Bedside ultrasonography can therefore add useful information in the early critical phase.

Thrombocytopenia is almost universal in dengue, and its degree is only a partial guide to severity. An admission platelet count below 50,000/µL was independently associated with severe disease, but its specificity was only 60.2%, so that many patients with marked thrombocytopenia remained clinically stable. Indian studies have likewise found that platelet count is a useful but imperfect indicator of disease course [45]. This matters for practice because platelet transfusion is not recommended as prophylaxis, and bleeding in dengue is driven chiefly by vascular and coagulation abnormalities rather than by the platelet count alone . In our cohort, 20% of patients received platelets, mostly because of bleeding or shock. Hospital-based reports from North India have described substantial use of platelet components in dengue [46], and a clear algorithm based on bleeding and haemodynamics may reduce unnecessary use.

A transaminase elevation with AST exceeding ALT was typical, and AST above three times the upper limit of normal was an independent predictor of severe dengue (aOR 3.71). Hepatic dysfunction is a well-recognised feature of dengue in adults, with a higher degree of derangement in severe disease [47–49]. The mechanisms include direct viral injury to hepatocytes and Kupffer cells, immune-mediated damage and hypoperfusion during shock. Indian data on liver involvement in dengue describe the pattern of aminotransferase rise and its clinical significance [47]. As AST is a cheap and universally available test, it is a pragmatic marker for risk stratification in resource-limited settings. Gallbladder wall thickening was also commoner in severe dengue in our series, which agrees with its role as a sonographic sign of plasma leakage.

Overall mortality of 2.8% was within the range reported by Indian tertiary care centres, which vary widely according to referral patterns and the proportion of severe disease. The deaths were concentrated in patients with refractory shock, massive bleeding and multiorgan failure, which are also the main causes of death reported in other studies. The absence of deaths among patients without severe disease emphasises that mortality in dengue is largely preventable by early recognition and correct fluid management. The trial by Wills and colleagues showed that careful, protocol-driven resuscitation with isotonic crystalloids can restore haemodynamics in most children with shock and similar principles apply to adults. Our findings suggest that a patient with fluid accumulation, a rising haematocrit and falling platelets should be monitored in a high-dependency area rather than in a general ward.

Decision rules for adults with dengue have been explored previously, for example in Singapore, to guide hospital admission. Our data suggest a simple bedside approach for Indian hospitals. A patient above 45 years or with diabetes who presents with persistent vomiting, lethargy, mucosal bleeding or signs of fluid accumulation should be treated as high risk. Haematocrit, platelet count and AST should be repeated at least daily during the critical phase, and abdominal and chest ultrasonography should be considered early. Patients without any of these features have a low probability of severe disease (negative predictive values of 86–95% for single markers) and, in the absence of other indications, may be managed with careful ambulatory follow-up during periods of high admission load, as permitted by the national guidelines.

Strengths and limitations

The strengths of this study include its prospective design, consecutive enrolment, uniform use of the WHO 2009 classification and a multivariable analysis with adequate model fit. Several limitations should be acknowledged. It was conducted at a single tertiary care centre, and referral bias may have inflated the proportion of severe disease and limited generalisability to primary care. Infecting serotype and primary versus secondary infection were not determined, although both influence severity. The number of severe cases (n = 74) was modest, so some confidence intervals are wide. Ultrasonography was not performed on every patient at the same time point, and some predictors were measured at admission and others at their nadir. The model has not been validated in an independent cohort, and a clinical scoring system derived from it should be tested prospectively before routine use. Finally, patients with documented co-infections were excluded, whereas in practice co-infection with malaria or scrub typhus is a diagnostic challenge in endemic regions.

Nearly one-quarter of adults admitted to this tertiary care hospital with dengue developed severe disease, with an ICU requirement of 18.1% and a mortality of 2.8%. Age above 45 years, diabetes mellitus, persistent vomiting, lethargy or restlessness, mucosal bleeding, clinical fluid accumulation, haemoconcentration, admission platelet count below 50,000/µL and AST above three times the upper limit of normal were independent predictors of severe dengue. These variables are simple, inexpensive and available at the point of care, and can be used to prioritise admission, closer monitoring and early fluid management. Multicentre studies with serotype data and external validation are needed to confirm these predictors and to develop a practical risk score for adults.

1. World Health Organization. Dengue: Guidelines for Diagnosis, Treatment, Prevention and Control. New edition. Geneva: WHO; 2009.
2. Gubler DJ. Dengue and dengue hemorrhagic fever. Clin Microbiol Rev. 1998;11(3):480-96.
3. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW, Moyes CL, et al. The global distribution and burden of dengue. Nature. 2013;496(7446):504-7.
4. Stanaway JD, Shepard DS, Undurraga EA, Halasa YA, Coffeng LE, Brady OJ, et al. The global burden of dengue: an analysis from the Global Burden of Disease Study 2013. Lancet Infect Dis. 2016;16(6):712-23.
5. Guzman MG, Harris E. Dengue. Lancet. 2015;385(9966):453-65.
6. Chaturvedi UC, Nagar R. Dengue and dengue haemorrhagic fever: Indian perspective. J Biosci. 2008;33(4):429-41.
7. Murhekar MV, Kamaraj P, Kumar MS, Khan SA, Allam RR, Barde P, et al. Burden of dengue infection in India, 2017: a cross-sectional population based serosurvey. Lancet Glob Health. 2019;7(8):e1065-73.
8. Ganeshkumar P, Murhekar MV, Poornima V, Saravanakumar V, Sukumaran K, Anandaselvasankar A, et al. Dengue infection in India: a systematic review and meta-analysis. PLoS Negl Trop Dis. 2018;12(7):e0006618.
9. Mutheneni SR, Morse AP, Caminade C, Upadhyayula SM. Dengue burden in India: recent trends and importance of climatic parameters. Emerg Microbes Infect. 2017;6(8):e70.
10. Kabra SK, Jain Y, Pandey RM, Madhulika, Singhal T, Tripathi P, et al. Dengue haemorrhagic fever in children in the 1996 Delhi epidemic. Trans R Soc Trop Med Hyg. 1999;93(3):294-8.
11. Aggarwal A, Chandra J, Aneja S, Patwari AK, Dutta AK. An epidemic of dengue hemorrhagic fever and dengue shock syndrome in children in Delhi. Indian Pediatr. 1998;35(8):727-32.
12. Gupta E, Dar L, Kapoor G, Broor S. The changing epidemiology of dengue in Delhi, India. Virol J. 2006;3:92.
13. Chakravarti A, Kumaria R. Eco-epidemiological analysis of dengue infection during an outbreak of dengue fever, India. Virol J. 2005;2:32.
14. Karoli R, Fatima J, Siddiqi Z, Kazmi KI, Sultania AR. Clinical profile of dengue infection at a teaching hospital in North India. J Infect Dev Ctries. 2012;6(7):551-4.
15. Kumar A, Rao CR, Pandit V, Shetty S, Bammigatti C, Samarasinghe CM. Clinical manifestations and trend of dengue cases admitted in a tertiary care hospital, Udupi district, Karnataka. Indian J Community Med. 2010;35(3):386-90.
16. Simmons CP, Farrar JJ, Nguyen VV, Wills B. Dengue. N Engl J Med. 2012;366(15):1423-32.
17. Whitehorn J, Farrar J. Dengue. Br Med Bull. 2010;95:161-73.
18. Kularatne SA. Dengue fever. BMJ. 2015;351:h4661.
19. World Health Organization. Dengue Haemorrhagic Fever: Diagnosis, Treatment, Prevention and Control. 2nd ed. Geneva: WHO; 1997.
20. Narvaez F, Gutierrez G, Perez MA, Elizondo D, Nunez A, Balmaseda A, et al. Evaluation of the traditional and revised WHO classifications of dengue disease severity. PLoS Negl Trop Dis. 2011;5(11):e1397.
21. Barniol J, Gaczkowski R, Barbato EV, da Cunha RV, Salgado D, Martinez E, et al. Usefulness and applicability of the revised dengue case classification by disease: multi-centre study in 18 countries. BMC Infect Dis. 2011;11:106.
22. Srikiatkhachorn A, Rothman AL, Gibbons RV, Sittisombut N, Malasit P, Ennis FA, et al. Dengue - how best to classify it. Clin Infect Dis. 2011;53(6):563-7.
23. National Vector Borne Disease Control Programme. National Guidelines for Clinical Management of Dengue Fever. New Delhi: Directorate General of Health Services, Ministry of Health and Family Welfare, Government of India; 2014.
24. Kittigul L, Pitakarnjanakul P, Sujirarat D, Siripanichgon K. The differences of clinical manifestations and laboratory findings in children and adults with dengue virus infection. J Clin Virol. 2007;39(2):76-81.
25. Hammond SN, Balmaseda A, Perez L, Tellez Y, Saborio SI, Mercado JC, 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(6):1063-70.
26. Tantawichien T. Dengue fever and dengue haemorrhagic fever in adolescents and adults. Paediatr Int Child Health. 2012;32(Suppl 1):22-7.
27. Lee IK, Liu JW, Yang KD. Clinical and laboratory characteristics and risk factors for fatality in elderly patients with dengue hemorrhagic fever. Am J Trop Med Hyg. 2008;79(2):149-53.
28. Figueiredo MA, Rodrigues LC, Barreto ML, Lima JW, Costa MC, Morato V, et al. Allergies and diabetes as risk factors for dengue hemorrhagic fever: results of a case control study. PLoS Negl Trop Dis. 2010;4(6):e699.
29. Pang J, Hsu JP, Yeo TW, Leo YS, Lye DC. Diabetes, cardiac disorders and asthma as risk factors for severe organ involvement among adult dengue patients: a matched case-control study. Sci Rep. 2017;7:39872.
30. Thein TL, Leo YS, Fisher DA, Low JG, Oh HM, Gan VC, et al. Risk factors for fatality among confirmed adult dengue inpatients in Singapore: a matched case-control study. PLoS One. 2013;8(11):e81060.
31. Wills BA, Nguyen MD, Ha TL, Dong TH, Tran TN, Le TT, et al. Comparison of three fluid solutions for resuscitation in dengue shock syndrome. N Engl J Med. 2005;353(10):877-89.
32. Rajapakse S. Dengue shock. J Emerg Trauma Shock. 2011;4(1):120-7.
33. Lee VJ, Lye DC, Sun Y, Leo YS. Decision tree algorithm in deciding hospitalization for adult patients with dengue haemorrhagic fever in Singapore. Trop Med Int Health. 2009;14(9):1154-9.
34. Kumarasamy V, Wahab AHA, Chua SK, Hassan Z, Chem YK, Mohamad M, et al. Evaluation of a commercial dengue NS1 antigen-capture ELISA for laboratory diagnosis of acute dengue virus infection. J Virol Methods. 2007;140(1-2):75-9.
35. Kaur P, Kaur G. Transfusion support in patients with dengue fever. Int J Appl Basic Med Res. 2014;4(Suppl 1):S8-12.
36. Mallhi TH, Khan AH, Adnan AS, Sarriff A, Khan YH, Jummaat F. Clinico-laboratory spectrum of dengue viral infection and risk factors associated with dengue hemorrhagic fever: a retrospective study. BMC Infect Dis. 2015;15:399.