Prediction of mortality in severe dengue cases

• Published in: BMC infectious diseases 2018-05, Vol.18 (1), p.232-232, Article 232
• Main Authors: Md-Sani, Saiful Safuan, Md-Noor, Julina, Han, Winn-Hui, Gan, Syang-Pyang, Rani, Nor-Salina, Tan, Hui-Loo, Rathakrishnan, Kanimoli, A-Shariffuddin, Mohd Azizuddin, Abd-Rahman, Marzilawati
• Format: Article
• Language: English
• Subjects: Adult; Alanine Transaminase - blood; Analysis; Area Under Curve; Bicarbonates; Bicarbonates - blood; Bleeding; Care and treatment; Chemical properties; Clinical medicine; Creatinine - blood; Data collection; Death; Dengue fever; Dengue hemorrhagic fever; Diagnosis; Disease; Epidemiology; Fatalities; Female; Hepatitis; Hospitals; Humans; Infections; Infectious diseases; Intensive care; Laboratories; Lactic acid; Lethargy; Malaysia; Male; Middle Aged; Mortality; Multivariate Analysis; Outcome and process assessment (Medical care); Patient outcomes; Patients; Physicians; Plasma; Predict; Prediction models; Public health; Pulse rate; Regression analysis; Regression models; Retrospective Studies; ROC Curve; Severe dengue; Severe Dengue - diagnosis; Severe Dengue - mortality; Statistical analysis; Studies; Survival Rate; Vector-borne diseases; Viral infections; Young Adult
• ISSN: 1471-2334; 1471-2334
• DOI: 10.1186/s12879-018-3141-6

Increasing incidence of dengue cases in Malaysia over the last few years has been paralleled by increased deaths. Mortality prediction models will therefore be useful in clinical management. The aim of this study is to identify factors at diagnosis of severe dengue that predicts mortality and assess predictive models based on these identified factors. This is a retrospective cohort study of confirmed severe dengue patients that were admitted in 2014 to Hospital Kuala Lumpur. Data on baseline characteristics, clinical parameters, and laboratory findings at diagnosis of severe dengue were collected. The outcome of interest is death among patients diagnosed with severe dengue. There were 199 patients with severe dengue included in the study. Multivariate analysis found lethargy, OR 3.84 (95% CI 1.23-12.03); bleeding, OR 8.88 (95% CI 2.91-27.15); pulse rate, OR 1.04 (95% CI 1.01-1.07); serum bicarbonate, OR 0.79 (95% CI 0.70-0.89) and serum lactate OR 1.27 (95% CI 1.09-1.47), to be statistically significant predictors of death. The regression equation to our model with the highest AUROC, 83.5 (95% CI 72.4-94.6), is: Log odds of death amongst severe dengue cases = - 1.021 - 0.220(Serum bicarbonate) + 0.001(ALT) + 0.067(Age) - 0.190(Gender). This study showed that a large proportion of severe dengue occurred early, whilst patients were still febrile. The best prediction model to predict death at recognition of severe dengue is a model that incorporates serum bicarbonate and ALT levels.