Dengue—How Best to Classify It

Since the 1970s, dengue has been classified as dengue fever and dengue hemorrhagic fever. In 2009, the World Health Organization issued a new, severity-based clinical classification which differs greatly from the previous classification.

Abstract

Dengue has emerged as a major public health problem worldwide. Dengue virus infection causes a wide range of clinical manifestations. Since the 1970s, clinical dengue has been classified according to the World Health Organization guideline as dengue fever and dengue hemorrhagic fever. The classification has been criticized with regard to its usefulness and its applicability. In 2009, the World Health Organization issued a new guideline that classifies clinical dengue as dengue and severe dengue. The 2009 classification differs significantly from the previous classification in both conceptual and practical levels. The impacts of the new classification on clinical practice, dengue research, and public health policy are discussed.

The exponential increase in published dengue research in recent years emphasizes the global importance of this infectious disease. Dengue is an important health threat to a large proportion of the worldwide population, with an estimated 3.6 billion persons at risk of this infection [1–3]. In addition to well-known areas of endemicity in Southeast Asia, dengue has become endemic in South and Central America. Dengue has now been reported in areas not previously affected and areas where the disease was thought to be controlled [1–3]. Furthermore, dengue, previously considered to be largely a pediatric illness, has been increasingly reported in adults [4–7]. The expansion of dengue geographically and demographically underscores the need for an appraisal of the clinical classifications with regard to their applicability and usefulness in clinical management and research.

Historically, dengue was considered to be a debilitating but not fatal illness. During the late 1960s, outbreaks of fatal dengue hemorrhagic fever in children in Southeast Asia changed this perception [8–11]. Clinical information from these outbreaks formed the basis for a dengue clinical classification published in the World Health Organization (WHO) guideline in 1975 and updated in 1997 [12, 13]. Clinical dengue is classified as dengue fever (DF) and dengue hemorrhagic fever (DHF) (Figure 1). Dengue fever is defined as a febrile illness with at least 2 clinical findings, including nausea, vomiting, headache, arthralgia, retro-orbital pain, rash, myalgia, hemorrhagic manifestations, and leukopenia. Because of the lack of specificity of these clinical signs and symptoms, laboratory or epidemiological evidence of dengue virus infection is required for a confirmed diagnosis. The definition of DHF consists of 4 clinical criteria: fever, a hemorrhagic tendency (spontaneous bleeding or a positive tourniquet test result), thrombocytopenia (platelet count, ≤100000 cells/mm³), and plasma leakage as shown by pleural effusion, ascites, or ≥20% hemoconcentration. Of importance, the specificity of this syndrome means that a diagnosis of DHF does not require laboratory evidence of dengue virus infection.

Figure 1.

Dengue classification according to the World Health Organization guidelines issued in 1975 and 1997. Dengue is classified as (1) undifferentiated fever, (2) dengue fever (DF), and (3) dengue hemorrhagic fever (DHF). In addition to fever and at least 2 clinical findings, diagnosis of DF requires epidemiological or laboratory evidence supporting a dengue virus infection. To meet a case definition of DHF, all 4 criteria are required: (1) fever, (2) hemorrhagic manifestations, (3) thrombocytopenia (platelet count, ≤100000 platelets/mm³), and (4) evidence of plasma leakage. Diagnosis of DHF does not require laboratory support.

Several criticisms have been raised against the case definition of DHF. Reports have argued that the case definition of DHF is too rigid and too difficult to apply in primary care or resource-limited settings [14–18]. Another criticism has been that the case definition fails to identify a significant proportion of severe dengue cases. For example, severe manifestations, such as encephalopathy and hepatic failure, are not included in the DHF case definition. It has also been questioned whether the case definitions, which were developed primarily on the basis of data derived from pediatric cases in southeast Asia, are applicable to other regions and populations. Consequently, a multicenter study to define clinical dengue was conducted (the Dengue Control [DENCO] study), and on the basis of its findings, a new clinical classification of dengue disease has been published in the most recent WHO guideline and subsequently promulgated [19, 20]. Although an effort to publish the findings is underway, the supporting data for the new classification have been, up to this point, largely inaccessible for evaluation [21]. A recent follow-up study from the same investigators indicates that the revised classification is undergoing evaluation regarding its applicability and usefulness [22].

The new WHO guideline classifies dengue as dengue and severe dengue (Figure 2). The definition of (nonsevere) dengue is somewhat similar to that of DF: a combination of ≥2 signs and symptoms in a febrile individual in an area of endemicity. However, the signs and symptoms listed in the revised classification are more numerous and less specific than those listed in the 1997 classification. They include aches and pains, nausea, vomiting, and rash. In addition, abdominal pain and tenderness, persistent vomiting, clinical fluid accumulation, mucosal bleeding, lethargy, restlessness, and liver enlargement are included in the list of clinical findings suggestive of dengue and also serve as warning signs for potentially severe dengue. Severe dengue is classified as cases including any of the following: (1) plasma leakage leading to shock or respiratory distress, (2) severe bleeding, or (3) organ failure (eg, elevated liver enzyme levels, impaired consciousness, or heart failure).

Figure 2.

Dengue classification according to the World Health Organization guidelines issued in 2009. Dengue is classified as dengue with or without warning signs and severe dengue. Diagnosis of dengue requires the presence of fever and at least 2 clinical findings or any warning signs. Epidemiological or laboratory evidence is required to make the diagnosis. Severe dengue is defined as dengue with any of the following: (1) severe plasma leakage leading to shock or respiratory distress, (2) severe hemorrhage, or (3) any organ failure.

The 2009 clinical classification represents a significant departure from the prior classification. In contrast to the previous classification, which defines DHF as a clinical entity with plasma leakage as the cardinal feature that differentiates it from DF, the 2009 classification lists several clinical manifestations as qualifiers for severe dengue. The improvement in dengue-associated mortality over the past decades has been based on the understanding of the natural history of plasma leakage in DHF, which occurs around the time of defervescence and coincides with the nadir of the platelet count. Delayed detection and treatment of plasma leakage is the major cause of organ failure, listed as severe manifestations in the 2009 classification. By listing severe organ involvement as a criterion for severity separate from plasma leakage, the revised classification places emphasis on isolated organ failure as a common and significant cause of dengue severity. With the exception of liver failure, which has been reported to occur in ≤1% in moderate to severe adult dengue cases [23, 24], the frequencies of severe organ impairment in the absence of plasma leakage have not been well documented. Most reports of isolated severe organ impairment consist of case reports or small case series [25–31]. On the basis of our prospective study involving children, severe organ involvement in the absence of plasma leakage is, in fact, uncommon [32]. In an attempt to include these unusual dengue manifestations, the 2009 classification has unfortunately reduced the emphasis on the most important aspect of dengue disease and the major factor contributing to fatality.

The 2009 severity-based classification stems from an argument that DHF case definition does not effectively identify severe dengue. It is important to indicate that the DHF case definition is based on a combination of clinical and laboratory findings to define a clinical syndrome rather than disease severity. Consistent with this, we found in a recent analysis that patients with DHF are not uniformly severe [32]. However, the risk of severe disease as defined by a requirement for intervention was significantly higher among patients with DHF than among those with DF. Therefore, the case definition of DHF affords significant discriminating value for clinical severity. Indeed, a recent study comparing the 2 classifications revealed that the 2 classifications perform comparatively well in identifying severe dengue, defined according to requirement for intervention [33].

Another important aspect of the 2009 classification that reflects its intended use as a case-management tool is the inclusion of nonspecific warning signs as diagnostic criteria for probable and potentially severe dengue cases. As such, a patient presenting with abdominal pain and vomiting in an area of endemicity during the dengue season will qualify as a probable dengue case. On the basis of the statistics from the outpatient department at Queen Sirikit National Institute for Child Health in Bangkok in 2008, the new case definition would have qualified an additional 39000 cases as probable dengue, whereas only 1600 suspected dengue cases were detected using a positive tourniquet test result and leukopenia as screening tools [34]. Among the latter, 76% were confirmed dengue cases by serologic testing (unpublished data). Patients with warning signs warrant close monitoring according to the new guideline, posing a significant burden for resource-poor countries where dengue is highly endemic. In addition, the requirement for laboratory confirmation is not clearly stated for severe dengue in the 2009 classification. Without laboratory confirmation, an unknown number of cases that qualify as severe dengue may not be dengue, because of the nonspecific nature of the criteria. In contrast, the DHF case definition excludes 99% of nondengue cases without the need for laboratory tests [32]. Implementation of the 2009 case definitions of probable and severe dengue without laboratory confirmation will certainly inflate the number of dengue cases and further complicate resource allocation and implementation of preventive measures.

The revised classification also poses significant problems for dengue research. Because the revised classification is designed primarily as a case-management tool, less emphasis is placed on the underlying pathophysiology. In contrast to the previous classification, which defined DHF as a clinical entity with a common pathophysiological process, the 2009 classification makes no such distinction. This reflects the notion that “dengue is one disease entity with different clinical presentations and often with unpredictable clinical evolution and outcome,” as indicated in the 2009 guideline [19, p. 11]. Consequently, severe dengue in the new classification includes cases with diverse and potentially unrelated underlying mechanisms. For instance, patients with highly elevated liver enzyme levels may be classified as having severe dengue regardless of the presence of other clinical findings (eg, plasma leakage or hemorrhage). This poses a problem for research studies that focus on specific manifestations (eg, plasma leakage) of dengue disease, because patients classified as having severe dengue may or may not have manifestation(s) relevant to a particular research question. This would require more-specific designations of severe dengue for research (eg, severe dengue with plasma leakage, severe dengue with liver failure, or severe dengue with heart failure).

Although the severity-based classification is the central feature of the proposed classification, most of these severe manifestations are not well defined in the guideline and others are based on arbitrary laboratory cutoff values. Indeed, the severity is based mostly on clinical judgment and will lead to heterogeneity in the classified groups because of variation in clinical practice. This is in contrast to the previous classification, which depends on objective findings only. Moreover, shock and severe organ involvement can be manifestations of the disease, complications of delayed or inappropriate treatment, underlying health conditions, or even adverse effects of medications. For example, fluid accumulation with respiratory distress, listed as a criterion for severity in the new classification, is often a complication of overtreatment with intravenous fluids. Because severity can be modified by intervention, which varies significantly from place to place, the 2009 definition of severe dengue will reflect the underlying disease severity and the regional and local clinical practice. It is questionable whether such a classification would be useful in research.

The dynamic nature of dengue demands close monitoring and repeated clinical and laboratory evaluations. Failure to obtain information, such as hematocrit and platelet counts, at the critical period is the primary reason for the difficulties in applying the DF-DHF classification. Unfortunately, this is prevalent even in recent prospective studies. For example, a subsequent study by the DENCO investigators showed that the maximum hematocrit reading was available in only 12% of cases [22]. The inconsistent application of the DF-DHF classification, largely because of the lack of critical information, has led to confusion and misunderstanding. It is imperative that research studies include sufficient detail on clinical data to justify their conclusions. As dengue continues to expand globally, the clinical classification will need to be reexamined and modified. However, loosening of the criteria for case classification and, in particular, incorporation of heterogeneous disease manifestations and mechanisms in a single classification, is not helpful. Despite its limitations, the DHF case classification has proved to be useful for advancing important observations on dengue disease pathogenesis, such as the importance of secondary dengue virus infection to the plasma leakage phenomenon, and has been instrumental in the development of treatment regimens that have saved numerous lives.