Dengue clinical features and predictors of severity in the diabetic patient: a retrospective cohort study on Reunion island, 2019

Aim: Diabetes mellitus is associated with both the risks of severe dengue and dengue-related deaths, however the factors characterizing dengue in the diabetic patient are ill-recognized. The objective of this hospital-based cohort study was to identify the factors characterizing dengue and those able to predict dengue severity in the diabetic patient. MethodsWe retrospectively analysed demographic, clinical and biological parameters at admission in the cohort of patients who consulted at the university hospital between January and June 2019 with confirmed dengue. Bivariate and multivariate analyses were conducted. ResultsOf 936 patients, 184 patients (20%) were diabetic. One hundred and eighty-eight patients (20%) developed severe dengue according to the WHO 2009 definition. Diabetic patients were older and had more comorbidities than non-diabetics. In an age-adjusted logistic regression model, loss of appetite, altered mental status, high neutrophil to platelet ratios (>14.7), low haematocrit ([≤] 38%), upper-range serum creatinine (>100 {micro}mol/l) and high urea to creatinine ratio (>50) were indicative of dengue in the diabetic patient. In a modified Poisson regression model, four key independent variables were predictive of severe dengue in the diabetic patient: presence of diabetes complications, non-severe bleeding, altered mental status and cough. Among diabetes complications, diabetic retinopathy and neuropathy, but not diabetic nephropathy nor diabetic foot, were predictive of severe dengue. ConclusionAt hospital first presentation, dengue in the diabetic patient is characterized by deteriorations in appetite, mental and renal functioning, while severe dengue can be predicted by presence of diabetes complications, dengue-related non-severe haemorrhages, cough, and dengue-related encephalopathy.