Dysglycaemia in Ebola virus disease: a retrospective analysis from the 2018 to 2020 outbreak

Abstract

BACKGROUND: Ebola virus disease (EVD) is associated with multisystem organ failure and high mortality. Severe hypoglycaemia is common, life-threatening, and correctable in critically ill patients, but glucose monitoring may be limited in EVD treatment units. METHODS: We conducted a retrospective review of patients admitted to EVD treatment units in Butembo and Katwa, Eastern DRC. Glucose measurements were done using a handheld glucometer at the bedside or using the Piccolo xpress Chemistry Analyzer on venous samples. FINDINGS: 384 patients (median age 30 years (interquartile range, IQR, 20-45), 57% female) and 6422 glucose measurements (median 11 per patient, IQR 4-22) were included in the analysis. Severe hypoglycaemia (≤2.2 mmol/L) and hyperglycaemia (>10 mmol/L) were recorded at least once during the ETU admission in 97 (25%) and 225 (59%) patients, respectively. A total of 2004 infusions of glucose-containing intravenous solutions were administered to 302 patients (79%) with a median cumulative dose of 175g (IQR 100-411). The overall case fatality rate was 157/384 (41%) and was 2.2-fold higher (95% CI 1.3-3.8) in patients with severe hypoglycaemia than those without hypoglycaemia (p = 0.0042). In a multivariable Cox proportional hazards model, periods of severe hypoglycaemia (adjusted hazard ratio (aHR) 6.2, 95% CI 3.2-12, p < 0.0001) and moderate hypoglycaemia (aHR 3.0, 95% CI 1.9-4.8, p < 0.0001) were associated with elevated mortality. INTERPRETATION: Hypoglycaemia is common in EVD, requires repeated correction with intravenous dextrose solutions, and is associated with mortality. FUNDING: This study was not supported by any specific funding.