Comparison of the characteristics and clinical course of patients with bacteraemia due to Burkholderia pseudomallei, Staphylococcus aureus and Escherichia coli in tropical Australia

Abstract

Background

Burkholderia pseudomallei, Staphylococcus aureus and Escherichia coli are common causes of bacteraemia in the tropical Asia Pacific region. The 28-day mortality rate of B. pseudomallei bacteraemia in Thailand has been reported to be 66%, higher than the 28-day mortality rate of bacteraemia due to S. aureus (43%) and E. coli (19%) in that country. Individuals living in rural and remote locations in these Thai studies had an even poorer prognosis. The mortality rates of individuals with bacteraemia due to B. pseudomallei, S. aureus and E. coli in countries with well-resourced health systems like Australia have not been compared directly.

Methods

We examined all cases of bacteraemia due to B. pseudomallei (between 2016 and 2022) and S. aureus and E. coli (between 2016 and 2020) in the Far North Queensland region of tropical Australia. We compared the characteristics and clinical course of these patients. We also examined the contribution of age, gender, comorbidity, remote residence and First Nations Australian status to the patients’ 30-day mortality.

Results

In total there were 177 (10.1%) episodes of bacteraemia due to B. pseudomallei, 601 (34.3%) due to S. aureus and 974 (55.6%) due to E. coli. Individuals with B. pseudomallei bacteraemia were younger than individuals with bacteraemia due to the other pathogens (median (interquartile range) age: 58 (47–67) versus 65 (49–77), p < 0.0001), they were less likely to live remotely (67/177 (37.9%) versus 730/1575 (46.4%), p = 0.03) and they did not have a greater rate of severe comorbidity (Charlson Comorbidity Index ≥ 5; 59/177 (33.3%) versus 624/1572 (39.7%), p = 0.10). There were 133/1752 (7.6%) who died within 30 days, which included 19/177 (10.7%) with B. pseudomallei bacteraemia, 62/601 (10.3%) with S. aureus bacteraemia and 52/974 (5.3%) with E. coli bacteraemia. In multivariate analysis that included all 5 pre-specified patient characteristics, the year of presentation and the 3 pathogens (with E.coli as the reference), B. pseudomallei bacteraemia (hazard ratio (HR) (95% confidence interval (CI): 2.52 (1.48–4.31), p = 0.001), S. aureus bacteraemia (HR (95% CI): 2.47 (1.69–3.59), p < 0.0001), severe comorbidity (HR (95% CI): 2.53 (1.66–3.85), p < 0.0001) and age (divided by 10) (HR (95% CI): 1.19 (1.06–1.34), p = 0.001) were independently associated with a higher 30-day mortality, while a rural/remote presentation (HR (95% CI): 0.53 (0.37–0.78), p = 0.001) was independently associated with a lower 30-day mortality. Among patients with B. pseudomallei bacteraemia, there was no association between any other of the pre-specified patient factors and 30-day mortality.

Conclusions

A well-resourced hub and spoke model of health care was able to minimise the 30-day mortality of individuals with B. pseudomallei, S. aureus and E. coli bacteraemia in this region of remote tropical Australia. However, B. pseudomallei bacteraemia had a higher 30-day mortality than bacteraemia due to S. aureus or E. coli, highlighting the lethal potential of the organism.