RESUMO
Increasing awareness of the importance of pathology stewardship in reducing low value care has led to scrutiny of appropriate laboratory test ordering. The objective of this study is to investigate the value of a commonly ordered laboratory test, the C-reactive protein (CRP), in decisions regarding diagnosis, management and disposition of emergency department (ED) patients with suspected sepsis. Retrospective chart reviews were performed on 1716 adult patients with suspected sepsis presenting to the Townsville University Hospital ED between 1 January 2021 and 30 June 2021. Suspected sepsis was defined as the emergency clinicians' decision to perform a blood culture. A CRP value of 10 mg/L or higher was defined as an elevated CRP. The primary outcome of interest was commencement of antibiotics in ED. Secondary outcomes include hospital admission (ward and ICU), hospital length of stay, mortality, documentation of indication for CRP testing, test parameters of CRP in detecting culture-positive bacteraemia and rates of bacteraemia with presumptive ED diagnosis. This study found no significant association between CRP values and antibiotic commencement (p=0.222), ward admission (p=0.071), ICU admission (p=0.248), hospital length of stay (p=0.164) or mortality (p=1.000). CRP had an area under the curve of 0.58 (95% CI 0.51-0.66) for detecting culture-positive bacteraemia. Sensitivity and specificity of CRP were 62.5% and 47.7%, respectively, at a threshold of 46 mg/dL. CRP testing is of little value in ED patients with suspected sepsis as it does not influence decision making about diagnosis, management, or disposition. Avoiding CRP testing in this patient cohort can contribute to pathology stewardship and optimal use of finite healthcare resources.