Melioidosis in people living with diabetes; clinical presentation, clinical course and implications for patient management

Abstract

Despite the well-established link between diabetes mellitus and melioidosis, the precise impact of diabetes, its complications, and its therapy on the presentation and clinical course of melioidosis is incompletely defined. The influence of glycaemic control on the diverse clinical manifestations and the clinical course of melioidosis in patients with diabetes has also not been examined in detail. We examined all cases of culture-confirmed melioidosis in Far North Queensland, Australia between October 1, 2016, and April 30, 2024. We hoped to define the impact of diabetes, its control and its therapy on the patients' presentation and their clinical course. There were 321 cases of culture-confirmed melioidosis during the study period; the patients' median (interquartile range (IQR)) age was 57 (46-69) years, 212/321 (66 %) were male, 130/321 (41 %) identified as First Nations Australians. Diabetes was the most common risk factor for melioidosis in the cohort (163/321, 51 %); in 19/163 (12 %) this was a new diagnosis. The median (IQR)) glycosylated haemoglobin prior to presentation was 9.1 % (7.2-11.5) and 96/162 (59 %) with complete data had established macrovascular or microvascular complications. People with diabetes were more likely - than people without diabetes - to have involvement of the liver (odds ratio (OR) 95 % confidence interval (CI): 9.68 (2.21-42.46), p = 0.003), the spleen (OR (95 % CI): 7.32 (1.64-32.80), p = 0.009) or to have disseminated disease (OR (95 % CI): 2.93 (1.26-6.78), p = 0.01). However, people with diabetes were no more likely than people without diabetes to require intensive care unit admission (OR (95 % CI): 0.82 (0.47-1.42), p = 0.48) or to die before hospital discharge (12/163 (7 %) versus 19/158 (12 %), OR (95 % CI): 0.58 (0.27-1.24), p = 0.16). Only 58/163 (36 %) with diabetes had specialist endocrinology review during their hospitalisation and only 22/72 (31 %) with accessible data had good glycaemic control (glycosylated haemoglobin ≤7 %) in the 12 months after discharge, increasing their risk of subsequent diabetic complications. Of the 151 people with diabetes surviving their hospitalisation, 26 (17 %) died, at a median (IQR) of 1.0 (0.40-4.1) years after discharge. Of the individuals with diabetes who had completed five years of follow up, 21/60 (35 %) had died at a median (IQR) age of 67 (51-84) years. Individuals with diabetes and melioidosis are more likely to have liver and spleen abscesses and disseminated disease than individuals without diabetes, manifestations that appear to be linked directly to glycaemic control. In Australia's well-resourced health system <10 % of patients with diabetes and melioidosis will die from their infection. However, five-year all-cause mortality in individuals with diabetes who survive their melioidosis is greater than 30 %, emphasising the importance of close, holistic multidisciplinary follow-up to ensure their optimal long-term health outcomes.