Worldwide, several clinical screening tools (including EWS, SOFA, qSOFA, SIRS and CURB65) are used on admission to identify patients at risk of sepsis-related mortality. However their ability to accurately predict mortality remains controversial. The purpose of this study was to evaluate whether the predictive performance of these tools is affected by the timing of their use.


A systematic literature search was performed using PubMed, identifying studies in adult patients with a suspected admission diagnosis of infection, sepsis or pneumonia, in which screening tools were used to predict mortality. Meta-regression analysis was performed on included studies to identify factors affecting the tool’s ability to predict mortality, with a focus on score timing.


From 3901 abstracts screened, 49 studies met inclusion criteria, comprising 421,006 patients and 13 clinical screening tools. No significant difference was found between any predictive tool and mortality. Of all variables considered (size of study, setting, diagnostic group, timing), only the timing related to admission affected predictive value of the tool. Studies that reported using purely physiological measures were less sensitive, marginally more specific, but had generally poorer predictive ability than those which included additional biochemical measures.


Clinicians must recognise that the performance of clinical screening tools is largely related to when they are used, not the individual tool. Given such tools are used on admission to identify risk of sepsis-related mortality, future studies must consider admission scores rather than those further into patients’ admission, to avoid over-reporting any tool’s predictive ability.

  • This is an open-access article distributed under the terms of the Creative Commons Attribution License.

Article metrics loading...

Loading full text...

Full text loading...

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error