Dengue fever is an acute arboviral febrile illness transmitted by Aedes mosquitoes and has four serotypes: DENV1, DENV2, DENV3, and DENV4 ¹ . It has been reported in over 70 countries and is endemic in the tropics, where ≥2.5 billion people are at risk ^{2– 4} . India is a category A endemic country ² ; in such countries dengue strains the already over-stretched health care system, and increases the economic burden due to hospitalization, absence from school, loss of productivity, loss of tourism revenue, and out-of-pocket expenses ^{2, 4} . Even this burden is a gross underestimate due to underreporting of cases and deaths, poor networking of healthcare facilities, and interstate/inter-regional differences in the availability of optimal healthcare ⁵ . The number of reported dengue cases in Pondicherry, South India had fallen from >3500 in 2012 to <500 in the year 2016, with a reported total mortality of just six cases during those five years ⁶ .

Guidelines by the World Health Organization (WHO) have been used to manage patients with dengue for ~45 years ⁷ . Until 2009, dengue cases were classified as dengue fever, dengue haemorrhagic fever and dengue shock syndrome, and appropriate management protocols were prescribed by WHO and national guidelines ⁸ . The revised classification by severity categorised patients into dengue fever without warning signs (Class A), with warning signs (Class B), and those with severe dengue (Class C) ⁸ . The warning signs include abdominal pain or tenderness, hepatomegaly, persistent vomiting, clinical fluid accumulation, mucosal bleed, lethargy and restlessness, and increase in haematocrit associated with fall in platelet counts. Patients in Group B must be admitted for the administration of fluids and observation during the critical phase, while patients in Group C should receive emergency care and/or urgent referral. Volume replacement guidelines by the WHO have been based on studies in children ⁹ . The quantum of fluids to be administered to patients in Group B has not been studied so far. Previously, only two studies had described the trends in fluid requirements over the patients’ entire hospital stay but using the earlier dengue severity classification ^{10, 11} . Two other studies, one each from Malaysia and Nicaragua, had described the effectiveness of drinking one litre of fluid before hospital admission ^{12, 13} .

Even though only patients in class B and C require admission, most patients with dengue fever are often admitted due to inability to differentiate it from other acute febrile illnesses or from non-severe dengue itself ⁹ . Guideline-based administration of intravenous fluids (IVF) are based on patients’ weight and are generally a step-down regimen as the patient’s clinical and laboratory parameters improve. This is often not practical in primary care due to the shortage of doctors and nurses to monitor as prescribed in the guidelines, spacing constraints, and availability of 24-hour laboratory services. Surveillance for severe dengue is generally poor in countries with epidemics. This often leads to overcrowding at hospitals and clinics with dengue fever/mild illness. In association with overworked medical personnel and confusion among the lay public, this overcrowding often predisposes to less-than-optimal care for patients with severe dengue and increased mortality ⁴ .

In view of such a scenario, we decided to study patients hospitalized with dengue warning signs, wherein, they could be monitored and administered fluids, even with a low doctor/nurse-to-patient ratio, by informing patients in advance what their daily fluid intake was expected to be and using the assistance of their relatives in measuring fluid intake and output. This study aimed to find out the quantity of fluids adult patients with dengue fever and warning signs received to maintain hemodynamic stability and to discern which anthropometric measurement-based fluid requirements correlated best with manifestations of severe dengue.

There are 50–100 million new cases of dengue annually, with the vast majority of them in individuals >15 years of age ^{4, 17} . Dengue began being reported in Pondicherry from 2007 onwards ¹ . A study from a hospital in Pondicherry during 2017 described the peak incidence of dengue between September and October; the dengue seroprevalence among antenatal mothers and children with fever was found to be 25.8% ¹⁸ . Severe dengue is more common in children ¹⁹ . The mean age of patients with severe dengue in our study was 30±12 years, compared to 26±11 years in those without severe dengue. Reports from India show that confirmed cases are being described mostly in the 21–30 years age group ¹ . Even though dengue shock requires admittance to an ICU, we managed all patients in the wards due to a shortage of beds in the ICU ³ . None of our patients developed other indications of ICU care such as ARDS, cardiogenic shock, encephalitis, or secondary pneumonia.

Fluid calculation during the critical phase is calculates as follows: 100 ml/kg for the first 10 kg (1000 ml), 50 mL/kg for the second 10 kg(500 ml) and 20 ml/kg for >20 kg up to 50 kg (600 ml), and 5% deficit, calculated as 50 ml/kg up to 50 kg (2500 ml) which yields a total of 4600 ml for a 50 kg person for about 48 hours ²⁰ . Since 33/37 patients had severe dengue, the fluids administered were much higher than this amount during each day of the two-day study. Some studies have shown that the new WHO guidelines increased the sensitivity of identifying patients with severe illness, with a risk of over-admission and increased workload, which would not be desirable in endemic regions that were resource-limited ⁷ . In a study from a tertiary hospital at Singapore, some cases of dengue with warning signs were safely managed in outpatient settings, but this practice cannot be extrapolated to other settings of primary/secondary care ⁷ . Most patients with severe illness progressed within a day of developing warning signs ⁷ . Thus, our study attempted to find out whether a range of fluid requirement would satisfy most patients (at least in a hospital) with dengue and warning signs without the minute assistance of a doctor or a nurse.

Only four studies have quantified the fluid requirements of patients with dengue, among which two have been performed in hospitalised patients ^{10– 13} . Three had used the earlier WHO classification for severity of dengue: dengue fever, dengue haemorrhagic fever and dengue shock syndrome. The actual requirements of fluids in these three subtypes have not been quantified and vary with individuals; it depends on the availability of resources, staff, and place of treatment. Two studies had reported fluid requirements from admission to discharge ^{10, 11} . A large prospective study of 1501 patients in Asia and Latin America with children constituting 65% described fluid accumulation, shock, and respiratory distress in relation to IV fluids, especially boluses. Median IVF volumes for maintenance, rehydration, and resuscitation were 50, 50 and 64.7 l/kg, respectively. The quantity of oral fluids consumed was not reported ¹¹ . Another Sri Lankan study had shown that the quantity of fluid consumed had no impact on the development of third space fluid accumulation ⁹ . We are the first to assess fluid (oral and IVF) requirements based on the severity classification of A, B, and C. We had practical difficulties with both oral and IVF (as listed under Results), but persisted in explaining the need for fluids, especially by mouth.

Fluids losses into the third spaces may occur even prior to the fourth day of fever and it is imperative that fluid replacement begins early, especially prior to admission, beginning at home ¹⁰ . Our study showed that the maximum total fluids (13,700 and 14,100 ml on Day 1 and Day 2, respectively) had been administered on the fourth day of fever ( Figure 2). During the first two days of fever, it is difficult to predict whether the disease may progress to shock (compensated or hypotensive) and hence fluid supplementation is important. In the Sri Lankan study auditing the trends in fluid requirements, 6267 ml was the mean requirement on the fourth day of fever similar to that of our study (6141 ml on Day 1 and 6433 ml on Day 2) ¹⁰ .

The Malaysian study showed that reading out advice from a dengue home care card to patients attending primary care clinics improved fluid intake by at least one litre and reduced hospitalization rates ¹² . They also found that a fluid chart maintained by the patient improved awareness of fluid intake and enhanced compliance. In Nicaragua, where the amount of fluid was also quantified, drinking five glasses of fluids prior to being seen by a doctor was protective against hospitalization ¹³ . The various liquids ingested in that study included juices, lemonade, milk, coffee, tea, and oral rehydration salts. We did not advise tea or coffee because of their weak diuretic action. We also suggested tender coconut water, dilute millet or rice porridge. Promoting high fluid intake was suggested to make an economic impact in countries with dengue fever ¹³ . We suggested an empirical 4–5 litre count based on the 100 ml/kg for an average 50 kg person, to account for both maintenance and replacement fluids arising from losses due to vomiting, diarrhoea, high fever, anorexia and poor intake in hot and humid weather conditions.

Volume replacement guidelines are based on studies in children and hence caution was advised in order to avoid fluid overload ¹⁷ . Intravenous fluids were given at a mean volume of 110 ml/kg over a mean of 25.3 hours in infants in a Vietnamese study which studied both patients with dengue hemorrhagic fever and dengue shock syndrome ¹⁷ . This contrasted with a mean IVF volume of 40.44 ml/kg (2318.6 ml/57.33 kg) on Day 1 and 29.69 ml/kg (1702.33 ml/57.33 kg) on Day 2 in our study. Hence the requirement of administered fluids is much lower when considered for their weights. A different anthropometric measurement may perhaps be useful in adults to gauge actual requirements, since the BSA and BMI correlated better with laboratory parameters of severe dengue in our study. Fluids/BSA correlated best with hypotensive shock on both days, while fluids/BMI correlated significantly with all the lab predictors of severe dengue.

This study had very small numbers due to the fact it was an audit of treatment practices in a single treating unit in Medicine (total of five units) during an epidemic that lasted 2 months. The dengue species was not known. The blood pressure was measured using a mercury sphygmomanometer, where errors in a few millimetres could probably misclassify patients with compensated and hypotensive shock. Also, some patients with severe postural symptoms had their BP checked in the sitting position, which could have led to a higher pulse pressure being recorded. The ECG heart rate was documented only at admission and patients could have been misclassified as having or not having a warning sign during Day 1 of the study. For logistic reasons, the study began only at 0800 AM of the following day after admission and hence the fluid administration for patients with shock or warning signs during the previous hours was not quantified. This amount may have been higher since all patients had been admitted initially with warning signs.

Most adults hospitalised with dengue fever and warning signs seem to require >100 mL/kg/day of fluids during their hospital stay without developing symptoms of fluid overload. Even this amount was not sufficient to help two patients (2/12) recover with compensated shock and prevent one patient from worsening to hypotensive shock. The quantum of fluids given in this study is over and above what has been described in the national guidelines, which are weight-based calculations. Abdominal pain, overweight individuals, tachycardia on ECG, hypotensive shock, increasing haematocrit and elevated CK levels were significantly associated with having received >5 l/day. Patients with warning signs may in addition have compensated shock (dizziness and postural hypotension) and severe organ impairment (arrhythmia) which may be missed at admission during epidemic-like situations. Since some patients developed warning signs on the third day of fever, it would be prudent to sometimes lower the threshold for admission in patients who come to the emergency department. Advising a larger amount of fluids (as in our study) at home during dengue epidemics could probably further reduce the need for admissions and/or development of severe dengue- a fraction of the fluid quantity that we administered our patients had already shown to reduce hospitalization rates in two other studies. In resource-poor settings, letting patients and their attendants participate in the acute bedside management of dengue with warning signs may go a long way in preventing morbidity in dengue fever.

Harvard Dataverse: Quantum of fluids in hospitalised patients with dengue and warning signs- a pilot cross-sectional study. https://doi.org/10.7910/DVN/SKEJAM ¹⁶ .

This project contains information on the fluids consumed and patient blood values.

Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).