Implications of Discordance in World Health Organization 1997 and 2009 Dengue Classifications in Adult Dengue

Background: Revised dengue guidelines were published by the World Health Organization (WHO) in 2009 addressing severe dengue cases not classified by dengue hemorrhagic fever (DHF) and shock syndrome (DSS). Methods and Principal Findings: We conducted a retrospective cohort study to compare WHO 2009 and 1997 classifications using 1278 adult dengue cases confirmed by polymerase chain reaction assay from Singapore epidemics in 2004 and 2007 (predominantly serotype 1 and 2 respectively). DHF occurred in 14.3%, DSS 2.7% and severe dengue 16.0%. The two WHO dengue classifications were discordant in defining severe disease (p <0.001). Five DSS patients (15%) were classified as non-severe dengue without warning signs. Of severe dengue patients, 107 did not fulfil DHF criteria. Of these, 14.9% had self-resolving isolated elevated aminotransferases, 18.7% gastrointestinal bleeding without hemodynamic compromise and 56.1% plasma leakage with isolated tachycardia. We compared both guidelines against requirement for intensive care including the single death in this series: all six had severe dengue; only four had DHF as two lacked bleeding manifestations but had plasma leakage. Increasing length of hospitalization was noted among severe cases with both classifications but the trend was only statistically significant for WHO 2009. Length of hospitalization was significantly longer for severe plasma leakage compared with severe bleeding or organ impairment. Requirement for hospitalization increased using WHO 2009 from 17.0% to 51.3%. Conclusions: While the WHO 2009 dengue classification is clinically useful, we propose retaining criteria for plasma leakage and hemodynamic compromise from WHO 1997, and refining definitions of severe bleeding and organ impairment to improve clinical relevance having found that differences in these accounted for the discordance between classifications. Findings from our retrospective study may be limited by the study site - a tertiary referral center in a hyperendemic country - and should be evaluated in a wider range of geographic settings.