Biomarkers for the diagnosis of sepsis in burns: systematic review and meta-analysis

Dr Andrew Li1, Mr Erwin Yii2, Dr Zhiliang Caleb Lin1, Dr Anthony Moussa1, Ms Lorena Romero3, Prof Allen Cheng4,5, Mr Cheng Lo1, Dr Alexander Padiglione4, Assoc Prof Heather Cleland1

1Victorian Adult Burns Service, The Alfred, Melbourne, Australia, 2Monash School of Medicine, Monash University, Australia, 3Ian Potter Library, The Alfred, Melbourne, Australia, 4Department of Infectious Diseases, The Alfred, Melbourne, Australia , 5Monash School of Public Health, Melbourne, Australia


Background: Sepsis is a common complication after burn injury and a leading cause of death. Prompt antibiotic administration is critical to survival, however early clinical diagnosis remains challenging. Biomarkers, such as blood cell counts and serum molecules, have been proposed as adjuncts to aid in delineating sepsis from the systemic inflammatory and hypermetabolic state following a burn injury. We aimed to evaluate the diagnostic accuracy of biomarkers for sepsis in burns patients.

Methods: We systematically reviewed clinical studies that assessed the diagnostic accuracy of biomarkers for sepsis in burns-specific populations. We searched Medline, Embase, Cochrane Central, Biosis Previews, Web of Science, and Medline In-Process from inception to December 2018, and undertook citation checking of included articles. Inclusion criteria encompassed sufficient data to construct a 2×2 contingency table, and a sepsis definition that incorporated either bacteraemia or a combination of infection, systemic inflammation, and organ dysfunction.

Results: 15 studies with 716 participants met our inclusion criteria. A considerable variety of biomarkers have been evaluated in burns populations, including procalcitonin (n=13), C-reactive protein (n=2), white cell count (n=2), lactate (n=1), presepsin (n=1), mid-regional pro-adrenomedullin (n=1), and endotoxin (n=1). Reported sensitivities ranged from 9.5% (lactate 3.9mmol/L) to 100% (procalcitonin 0.5ng/mL), and specificities ranged from 22.2% (procalcitonin 0.5ng/mL) to 100% (procalcitonin 10ng/mL). Descriptive synthesis, meta-analysis of suitable studies, and quality assessment will be presented.

Conclusion: A wide variety of biomarkers have been studied for the diagnosis of sepsis in burns patients. Studies to date vary in their methodology and reports of diagnostic accuracy.


Dr Andrew Li is a surgical resident at Alfred Health, and has worked with the Victorian Adult Burns Service. He completed his MBBS at Monash University in 2016.

Recent Comments