METHODS: Twenty-four national-level public health laboratories performed routine diagnostic assays on a proficiency testing panel consisting of two modules. Module A contained serum samples spiked with cultured dengue virus (DENV) or chikungunya virus (CHIKV) for the detection of nucleic acid and DENV non-structural protein 1 (NS1) antigen. Module B contained human serum samples for the detection of anti-DENV antibodies.

RESULTS: Among 20 laboratories testing Module A, 17 (85%) correctly detected DENV RNA by reverse transcription polymerase chain reaction (RT-PCR), 18 (90%) correctly determined serotype and 19 (95%) correctly identified CHIKV by RT-PCR. Ten of 15 (66.7%) laboratories performing NS1 antigen assays obtained the correct results. In Module B, 18/23 (78.3%) and 20/20 (100%) of laboratories correctly detected anti-DENV IgM and IgG, respectively. Detection of acute/recent DENV infection by both molecular (RT-PCR) and serological methods (IgM) was available in 19/24 (79.2%) participating laboratories.

METHODS: This was a retrospective cohort study of confirmed dengue patients who were warded in Kuala Lumpur Hospital between December 2014 and January 2015. CK, AST, ALT, hematocrit, platelet count, WBC and serum albumin were taken upon ward admission and repeated at timed intervals. Composite indices based on admission AST and ALT were analyzed. Correlation coefficients and coefficients of determination were computed.

RESULTS: Among the 365 cases reviewed, twenty-two (6%) patients had severe dengue. AST and ALT were found to be good at identification of severe dengue. The AST2/ALT composite index was the most accurate (AUC 0.83; 95% CI 0.73 - 0.93). Optimal cutoff was 402 with a sensitivity of 59.1% (95% CI: 36.4 - 79.3%) and specificity of 92.4% (95% CI: 89.1 - 95.0%). Modified cutoff of 653 had a sensitivity of 40.9% (95% CI: 20.7 - 63.7%) and specificity of 97.4% (95% CI: 95.1 - 98.8%). Our analyses also suggested that several underlying biological processes represented by biomarkers tested were unrelated despite occurring in the same disease entity. Also, markers of plasma leakage were discordant and AST was likely hepatic in origin.

METHODS: This prospective cross-sectional study consecutively recruited 494 patients with suspected dengue from a health clinic in Malaysia. Both RDTs were performed onsite. The evaluated ELISA and reference tests were performed in a virology laboratory. The reference tests comprised of a reverse transcription-polymerase chain reaction and three ELISAs for the detection of dengue NS1 antigen, IgM and IgG antibodies, respectively. The diagnostic performance of evaluated tests was computed using STATA version 12.

RESULTS: The sensitivity and specificity of ViroTrack were 62.3% (95%CI 55.6-68.7) and 95.0% (95%CI 91.7-97.3), versus 66.5% (95%CI 60.0-72.6) and 95.4% (95%CI 92.1-97.6) for SD NS1 ELISA, and 52.4% (95%CI 45.7-59.1) and 97.7% (95%CI 95.1-99.2) for NS1 component of SD Bioline, respectively. The combination of the latter with its IgM and IgG components were able to increase test sensitivity to 82.4% (95%CI 76.8-87.1) with corresponding decrease in specificity to 87.4% (95%CI 82.8-91.2). Although a positive test on any of the NS1 assays would increase the probability of dengue to above 90% in a patient, a negative result would only reduce this probability to 23.0-29.3%. In contrast, this probability of false negative diagnosis would be further reduced to 14.7% (95%CI 11.4-18.6) if SD Bioline NS1/IgM/IgG combo was negative.

METHODS: Electronic searches were conducted in the Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, MEDLINE (complete), PubMed and Scopus. Eligible studies to be included in this review were cohort studies with participants confirmed by laboratory test for dengue infection and comparison among the different severity of dengue infection by using statistical models. The methodological quality of the paper was assessed by independent reviewers using QUADAS-2.

RESULTS: Twenty-six studies published from 1994 to 2017 were included. Most diagnostic models produced an accuracy of 75% to 80% except one with 86%. Two models predicting severe dengue according to the WHO 2009 classification have 86% accuracy. Both of these logistic regression models were applied during the first three days of illness, and their sensitivity and specificity were 91-100% and 79.3-86%, respectively. Another model which evaluated the 30-day mortality of dengue infection had an accuracy of 98.5%.

METHODS: In-depth individual interviews with thematic saturation were conducted between May and July 2018. The data was analyzed using thematic analysis.

RESULTS: Based on expert opinion, diagnosis of severe dengue is challenging as it depends on astute clinical interpretation of non-dengue-specific clinical and laboratory findings. A specific test that detects impending manifestation of severe dengue could 1) overcome failure in identifying severe disease for referral or admission, 2) facilitate timely and appropriate management of plasma leakage and bleeding, 3) overcome the lack of clinical expertise and laboratory diagnosis in rural health settings. The most important feature of any diagnostics for severe dengue is the point-of-care (POC) format where it can be performed at or near the bedside.