Severe leptospirosis in tropical Australia: Optimising intensive care unit management to reduce mortality.

<h4>Background</h4>Severe leptospirosis can have a case-fatality rate of over 50%, even with intensive care unit (ICU) support. Multiple strategies-including protective ventilation and early renal replacement therapy (RRT)-have been recommended to improve outcomes. However, management guidelines vary widely around the world and there is no consensus on the optimal approach.<h4>Methodology/principal findings</h4>All cases of leptospirosis admitted to the ICU of Cairns Hospital in tropical Australia between 1998 and 2018 were retrospectively reviewed. The patients' demographics, presentation, management and clinical course were examined. The 55 patients' median (interquartile range (IQR)) age was 47 (32-62) years and their median (IQR) APACHE III score was 67 (48-105). All 55 received appropriate antibiotic therapy, 45 (82%) within the first 6 hours. Acute kidney injury was present in 48/55 (87%), 18/55 (33%) required RRT, although this was usually not administered until traditional criteria for initiation were met. Moderate to severe acute respiratory distress syndrome developed in 37/55 (67%), 32/55 (58%) had pulmonary haemorrhage, and mechanical ventilation was required in 27/55 (49%). Vasopressor support was necessary in 34/55 (62%). Corticosteroids were prescribed in 20/55 (36%). The median (IQR) fluid balance in the initial three days of ICU care was +1493 (175-3567) ml. Only 2/55 (4%) died, both were elderly men with multiple comorbidities.<h4>Conclusion</h4>In patients with severe leptospirosis in tropical Australia, prompt ICU support that includes early antibiotics, protective ventilation strategies, conservative fluid resuscitation, traditional thresholds for RRT initiation and corticosteroid therapy is associated with a very low case-fatality rate. Prospective studies are required to establish the relative contributions of each of these interventions to optimal patient outcomes.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.