https://orcid.org/0000-0002-7175-258X
Abstract
One of the World Health Organization criteria for dengue haemorrhagic fever (DHF) is a 20% increase in haematocrit. This study on dengue hospitalised patients compares haematocrit levels at baseline, at the onset of the critical phase and the maximum during the critical phase in rural Sri Lanka.
This observational study included patients with dengue in the febrile phase who progressed to the critical phase. Haematocrit was recorded and ultrasound scans were performed thrice daily when the platelet count dropped to <100 000/μl. The onset of the critical phase was confirmed by ultrasound-detected plasma leakage.
Forty-three patients were included in the final analysis. The mean haematocrit at baseline and at the onset of the critical phase was 40.6% (standard deviation [SD] 4.7) and 41.3% (SD 5.1), respectively, with no difference (p=0.14, paired t-test). None of the participants showed a 20% increase in haematocrit at the onset of the critical phase, with the maximum observed increase being 17.0%.
A 20% increase in haematocrit was not observed during the critical phase, suggesting that haemoconcentration is unreliable for diagnosing DHF in hospitalised patients. This study recommends revisiting the WHO criterion of a 20% haematocrit increase for diagnosing plasma leakage in dengue patients. It suggests that frequent ultrasound scans may be a more reliable method for early detection of plasma leakage.
Introduction
Dengue is the deadliest mosquito-borne viral infection in the world. It is transmitted primarily by Aedes aegypti and to a lesser extent by Aedes albopictus.1,2 Before 1970, it was a minor, geographically restricted disease. Now it is endemic in >100 countries, with >2.5 billion people at risk. The World Health Organization (WHO) estimates that 50–100 million infections occur yearly, including 500 000 dengue haemorrhagic fever (DHF) cases and 22 000 deaths, mostly among children.2
Dengue is a significant public health concern in Sri Lanka, with frequent outbreaks and a high disease burden.3 The country has seen a substantial increase in dengue cases over the past decade, even affecting rural areas like Anuradhapura. In 2023, nearly 90 000 dengue cases were reported in Sri Lanka, with 62 deaths.4 During this period, 845 cases were reported in the Anuradhapura district.5 Current diagnostic practices in Sri Lanka rely on dengue antigen and immunoglobulin M (IgM) antibody during the febrile phase and symptoms such as fever, haemorrhagic manifestations and laboratory features of thrombocytopenia (<100 000/μl) and a 20% increase in haematocrit (HCT) as criteria to diagnose DHF. This is in addition to the widespread use of point-of-care ultrasound scanning in major hospitals.
The clinical disease spectrum is broad, from mild, non-specific flu-like illness to severe DHF and dengue shock syndrome (DSS).2,6 DHF and DSS typically occur during the second episode of dengue when patients are infected with a different viral subtype than the first. Plasma leakage is the characteristic feature of DHF. With the onset of plasma leakage, the patient enters the critical phase, usually lasting 24–48 h.6
There is no specific antiviral agent for dengue, DHF or DSS. Early identification of the critical stage and treatment is essential to prevent complications. Therefore, diagnosing plasma leakage at its initial stages is crucial to avoid death and dengue complications.
According to the 2011 WHO Southeast Asia guidelines, fever, haemorrhagic manifestations, thrombocytopenia (<100 000/μl) and an increase in HCT by 20% from the baseline are used to diagnose DHF.7 However, the current WHO classification system focuses on severe dengue rather than DHF as a separate entity.
The interpretation of haemoconcentration is not always reliable, especially when the patient is bleeding, dehydrated or on intravenous fluids.8,9 An individual may have significant haemoconcentration in the absence of plasma leakage in dehydration, while another may not show haemoconcentration if plasma leakage is complicated with bleeding. Further, there may not be significant haemoconcentration in the early leakage phase.
The key clinical feature of DHF is plasma leakage.10 The most objective way of diagnosing this is to detect pleural effusions and ascites by ultrasound imaging.7 Pooled sensitivity and specificity of ultrasonography in diagnosing a pleural effusion were 94% and 98%, respectively, while sensitivity and specificity of chest radiography were 51% and 91%, respectively.11 This meta-analysis concluded that ultrasonography was more accurate than radiography in detecting pleural effusions.11 In current clinical practice, ultrasonography has become the standard investigation to diagnose pleural effusions due to its high accuracy and freedom from radiation.
This study compared HCT at baseline, at onset and throughout the critical phase in patients with dengue.
Methods
This observational study was conducted in the university medical unit of the Anuradhapura Teaching Hospital, Anuradhapura, Sri Lanka, between February 2018 and June 2023. Inclusion criteria were patients ≥12 y of age with acute febrile illness, not dehydrated, a positive dengue non-structural protein 1 antigen result measured by lateral flow (immunochromatographic) assay and developed plasma leakage during the hospital stay. The diagnosis was further confirmed by dengue IgM antibody testing using enzyme-linked immunosorbent assay. Exclusion criteria were patients with leaking detected by ultrasound on admission and patients with a platelet count <100 000/μl on admission.
All patients who presented in the early febrile phase and deteriorated to the critical phase with DHF during the ward stay were included in the final analysis. Patients were considered hydrated when their blood pressure, capillary refilling time and urine output were normal and their mouths were moist with saliva. Baseline HCT was measured using a haematology analyser and, if dehydrated, the patient was hydrated before measurement. HCT was rechecked at the onset of the critical phase and two to three times a day until the critical phase ended. Patients were closely monitored with vital signs and urine output. Daily ultrasound scans were done in the early precritical phase and thrice-daily examinations were done when the platelet count was <100 000/μl to identify plasma leakage at the earliest possible instance. The onset of the critical phase was confirmed by the first ultrasonic detection of plasma leakage into pleural or peritoneal cavities.
Ultrasound scans were performed by trained registrars or consultants in internal medicine using a curvilinear probe (4.5 MHz) on the ultrasound machine (Z5, Mindray, Mahwah, NJ). Pleural effusions were looked at in the costodiaphragmatic recesses in an upright propped-up position. The hepatorenal pouch, splenorenal area and pouch of Douglas were screened for free fluid.
Definitions
Dengue haemorrhagic fever: dengue fever with evidence of plasma leakage.7,8
Early precritical phase: early days of fever with no ultrasound evidence of plasma leakage with a platelet count >100 000/μl.
Onset of the critical phase: the earliest instance where fluid extravasation is detected by ultrasonography in pleural or peritoneal cavities.
End of the critical phase: 48 h have elapsed from the onset of the critical phase and the patient has started feeling well with a return of appetite and polyuria.
Hepatitis: pain over the right hypochondrium and aspartate aminotransferase and alanine transaminase more than three times the upper limit of normal.
Statistical analysis
Haematocrit levels were normally distributed and assessed by the Shapiro–Wilk test, hence the data were described using mean (standard deviation [SD]). We used a 20% increase in HCT for severe plasma leakage and a 15% increase as moderate plasma leakage.12 A paired t-test was used to compare HCT levels at baseline, at the onset of the critical phase and during the critical phase. A p-value <0.05 was considered statistically significant. Data were entered and analysed and Figure 1 was generated using GraphPad Prism 10.1.2 (GraphPad Software, Boston, MA, USA).
HCT at baseline (red dots), at the onset of the critical phase (blue dots) and the maximum value during the critical phase (green dots). ****p<0.001, paired t-test. CP; critical phase.
Results
A total of 177 patients with confirmed dengue fever were initially recruited between February 2018 and June 2023, of which 48 (27%) progressed to DHF. Due to incomplete data for 5 patients, 43 patients (90%) were considered in the final analysis. The baseline characteristics and other clinical details are described in Table 1.
Demographics, clinical outcome and treatment metrics of the patients.
| Parameters | Values |
|---|---|
| Patients, N | 43 |
| Males, n (%) | 25 (58.1) |
| Age (years), median (IQR) | 33 (23–42) |
| Time to critical phase after fever onset (days), median (IQR) | 5 (4–6) |
| Patients entering the critical phase on day 4 or 5, n (%) | 26 (60.5) |
| Hepatitis, n (%) | 8 (18.6) |
| Other dengue-related organopathy, n (%) | 0 (0) |
| Post-recovery urinary tract infection, (%) | 1 (2.3) |
| Cannula site infections, n (%) | 2 (4.7) |
| Compensated shock, n (%) | 1 (2.3) |
| Fluid overload or intensive care unit admission, n (%) | 0 |
| Saline volume (ml) for haemodynamic stability, mean (SD) (range) | 4271 (265) (2900–5210) |
| Parameters | Values |
|---|---|
| Patients, N | 43 |
| Males, n (%) | 25 (58.1) |
| Age (years), median (IQR) | 33 (23–42) |
| Time to critical phase after fever onset (days), median (IQR) | 5 (4–6) |
| Patients entering the critical phase on day 4 or 5, n (%) | 26 (60.5) |
| Hepatitis, n (%) | 8 (18.6) |
| Other dengue-related organopathy, n (%) | 0 (0) |
| Post-recovery urinary tract infection, (%) | 1 (2.3) |
| Cannula site infections, n (%) | 2 (4.7) |
| Compensated shock, n (%) | 1 (2.3) |
| Fluid overload or intensive care unit admission, n (%) | 0 |
| Saline volume (ml) for haemodynamic stability, mean (SD) (range) | 4271 (265) (2900–5210) |
Demographics, clinical outcome and treatment metrics of the patients.
| Parameters | Values |
|---|---|
| Patients, N | 43 |
| Males, n (%) | 25 (58.1) |
| Age (years), median (IQR) | 33 (23–42) |
| Time to critical phase after fever onset (days), median (IQR) | 5 (4–6) |
| Patients entering the critical phase on day 4 or 5, n (%) | 26 (60.5) |
| Hepatitis, n (%) | 8 (18.6) |
| Other dengue-related organopathy, n (%) | 0 (0) |
| Post-recovery urinary tract infection, (%) | 1 (2.3) |
| Cannula site infections, n (%) | 2 (4.7) |
| Compensated shock, n (%) | 1 (2.3) |
| Fluid overload or intensive care unit admission, n (%) | 0 |
| Saline volume (ml) for haemodynamic stability, mean (SD) (range) | 4271 (265) (2900–5210) |
| Parameters | Values |
|---|---|
| Patients, N | 43 |
| Males, n (%) | 25 (58.1) |
| Age (years), median (IQR) | 33 (23–42) |
| Time to critical phase after fever onset (days), median (IQR) | 5 (4–6) |
| Patients entering the critical phase on day 4 or 5, n (%) | 26 (60.5) |
| Hepatitis, n (%) | 8 (18.6) |
| Other dengue-related organopathy, n (%) | 0 (0) |
| Post-recovery urinary tract infection, (%) | 1 (2.3) |
| Cannula site infections, n (%) | 2 (4.7) |
| Compensated shock, n (%) | 1 (2.3) |
| Fluid overload or intensive care unit admission, n (%) | 0 |
| Saline volume (ml) for haemodynamic stability, mean (SD) (range) | 4271 (265) (2900–5210) |
The Shapiro–Wilk statistic for baseline HCT (W=0.98, p=0.73) and peak HCT (W=0.99, p=0.93) indicated no significant deviation from normality. Mean HCT on admission and at the onset of the critical phase was 40.6% (SD 4.7) and 41.3% (SD 5.1), respectively, with no significant difference (p=0.14, paired t-test) (Figure 1). None of the participants in this sample showed a 20% increase in HCT at the onset of the critical phase, with the maximum observed increase being 17.0%. Only three (7.0%) participants had a ≥15% increase in HCT at the onset.
The mean maximum HCT during the critical phase was 43.5% (SD 5.4), which is significantly different from the baseline (p<0.001, paired t-test). However, no one demonstrated a 20% increase in HCT during the critical phase, with the maximum increase being 18.9%. Only seven (16%) patients had a ≥15% increase in HCT during the critical phase.
Ultrasound findings revealed that all patients had evidence of plasma leakage, with pleural effusions detected in 39.5% (17/43) of cases and ascites in 90.6% (39/43).
Discussion
A 20% increase in HCT was not seen in any patients entering the critical phase, and there was no difference between HCT at baseline and at the onset of the critical phase. HCT was used to diagnose the critical phase of DHF before bedside ultrasonography became routine, and it is still used in hospitals where bedside ultrasonography is unavailable and when it is challenging to transport ill patients to the radiology department.
There are two possibilities for this observation. First, the plasma leakage diagnosed by ultrasound scan may have been detected in the critical phase before significant haemoconcentration, which was detected by the HCT. Second, all patients were on intravenous fluid when plasma leakage was detected. The Sri Lankan guidelines recommend administering intravenous fluid once the platelet count drops to <100 000/μl.8 Therefore, haemoconcentration would have been treated by fluid administration.
From 1975 to 2009, the WHO used DHF and DSS to describe severe dengue.13 In 2009, the WHO introduced severe dengue to describe patients with severe plasma leakage, haemorrhage and organ impairment.12 Both classifications are in use, as one did not replace the other, and to add to the confusion, in 2011 and 2012 the WHO issued two documents, a guideline from South East Asia and a handbook.7,9 The 2009 guidelines flag an increase in HCT (no value given) and a rapid decrease in platelets as warning signs.14 In the current 2011 WHO Southeast Asia guidelines, one criterion for diagnosing DHF stage 1 is an increase in HCT of ≥20% from the baseline (page 26).7 On page 46 of the same document, a 10–20% HCT increase despite oral rehydration indicates the necessity of intravenous fluid in DHF during the critical phase. The Pan American Health Organization (PAHO) guidelines, published in 2022, discuss an increase in HCT (no value given) as one risk factor. Fluid accumulation detected clinically or by ultrasound is another risk factor.15
Bleeding is not a characteristic feature of DHF, but plasma leakage from increased vascular permeability is.16 Although thrombocytopenia is not a consequence, the progressive increase in HCT results from plasma leakage. However, having thrombocytopenia as a cut-off point may be a helpful predictor for DHF, especially since the expected increase in HCT was not seen. Although a 20% HCT increase may not be sensitive enough to detect leakage, a decrease may signify concealed internal bleeding. Hence, measuring serial HCT may still be relevant.
Several studies have compared the WHO criteria and ultrasonography in the diagnosis of DHF. In a study in South India, evidence of plasma leakage in the form of pleural effusion and/or ascites was observed in 46 of 65 (70.8%) dengue patients screened with ultrasonography.17 Therefore, they would qualify to be labelled as DHF. However, only 35 of 65 (53.8%) patients fulfilled the WHO criteria for DHF. Eleven cases of DHF would have been misdiagnosed as dengue fever if an ultrasound scan was not done. Another study conducted in Bangkok found ultrasound evidence of plasma leakage in 12 of 17 (70.6%) patients who did not show significant haemoconcentration.18 In a study in West Java, >91% of those who had ultrasonographic evidence of plasma leakage had severe dengue compared with 31% without.10
Studies report that ultrasound evidence of plasma leakage can be detected earlier than haemoconcentration in DHF.18,19 In a study in Bangkok in children, ultrasound evidence of plasma leakage was detected 1 d before haemoconcentration in 20% of patients with DHF.18 In a Malaysian study, plasma leakage was detected at 5.16 d after the onset of fever and increased haemoconcentration was detected at 5.6 d after.19 Data from the above studies suggest that haemoconcentration occurs later than ultrasound evidence of plasma leakage. Thus the use of haemoconcentration may delay the diagnosis of DHF. The data from our study were consistent with other studies reporting that haemoconcentration is a poor indicator of plasma leakage compared with ultrasound.10
The small sample size is a limitation of our study. We could not use the same haematology analyser to do all complete blood counts or a single operator to do ultrasound scans. The critical phase was determined by the clinician carrying out the ultrasound scan, and this was not verified by a radiologist or a repeat scan by another operator.
Conclusions
The WHO criterion of a 20% increase in HCT may not be reliable for diagnosing DHF in hospitalised patients, particularly those on intravenous fluids. While our observational study does not establish causality, it highlights the need to revisit the WHO criterion of a 20% HCT increase for diagnosing plasma leakage in dengue patients. In settings where ultrasound is not readily available, a lower threshold for an increase in HCT may still be helpful, but further research is needed to confirm this.
Authors’ contributions
CS, RB and KW conceived the study and designed the study protocol. CS, HS, PW, RB and HS carried out the clinical assessment. CS, AS, KW and SS conducted the analysis and interpretation data. CS and SS drafted the manuscript. All authors critically revised the manuscript for intellectual content and read and approved the final manuscript. SS is the guarantor of the article.
Acknowledgements
None.
Funding
None.
Competing interests
None declared.
Ethical approval
Ethical clearance for the study was obtained from the Ethics Review Committee of the Faculty of Medicine and Allied Sciences, Rajarata University of Sri Lanka (ERC/2017/54). Written informed consent was obtained from the patients. Informed written consent was obtained from patients and parents if the patient was <18 y old.
Data availability
The data will be shared upon reasonable request to the corresponding author.
Comments