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

Harness risk stratification of diabetic patients with dengue in a cohort study

BACKGROUND

Identifying predictors of severe dengue (SD) is key for triage and management of patients as well as for advising travellers to countries where dengue is endemic. In this, meta-analyses have raised diabetes mellitus as a risk factor for SD and a prognostic factor for dengue-related mortality. The purpose of this study was to assess whether diabetic patients (DPs) are at increased risk for SD in comparison to non-diabetic patients (NDPs) in a setting of high prevalence of type 2 diabetes mellitus and increasing endemicity for dengue.

METHODS

In a cohort study conducted during the 2019 dengue epidemic on Reunion Island, we estimated the risk ratios (RR) of DPs for SD (WHO 2009 definition), hospitalisation, intensive care unit (ICU) admission, critical care need or death in the ICU, and scales rating severity or multiple organ dysfunction syndrome (MODS), among confirmed cases of dengue (positive RT-PCR or NS1 antigen).

RESULTS

In a Poisson regression model adjusted for age, gender and comorbidity, DPs were more likely to develop SD (adjusted RR: 1.46, 95%CI 1.10-1.95), to be hospitalised, admitted to the ICU, and need critical care or die in the ICU. Subgroup analyses identified female DPs, non-elderly DPs (< 65 years) and DPs with low Charlson score (< 3) to be at higher risk for SD, the two first subgroups trough more severe presentation (higher Simplified Acute Physiology Score-2 values; higher MODS scores, respectively). Male gender, age less than 65 years and mixed comorbidity were identified as prognostic factors for critical care need or death in the ICU, male and non-elderly DPs being more likely to develop MODS than their non-diabetic counterparts.

CONCLUSIONS

Together, these data highlight the role of diabetes mellitus in the progression from dengue to SD through higher severity per se or the event of MODS.