Table 1 Characteristics of included observational studies
Study | Data collection period | Country | n infection/non-infection | Clinical condition | Control group (non-infection) | PSP cut-off for infection/non-infection | AUC ROC if cut-off | Conclusion |
|---|---|---|---|---|---|---|---|---|
Keel et al. [14] | January 2002 to September 2006 | CH | 49/14 | ICU patients at day 5 post-admission for trauma | ICU patients without infection at day 5 post-admission for trauma | None | – | PSP levels in infected patients were > 15-fold increased over baseline |
Llewelyn et al. [15] | August 2010 to January 2011 | UK | 88a/94 | Unselected ICU or IMC patients | Unselected ICU or IMC patients without infection | 30 | 0.93 | PSP performed well as infection biomarker in patients with suspected infection at the time of admission |
Gukasjan et al. [17] | August 2007 to February 2010 | CH | 88/43b | ICU patients with or without secondary peritonitisb | ICU patients after elective major abdominal surgery without secondary peritonitisb | None | – | PSP accurately predicted severity and outcome of severe infection |
Klein et al. [18] | May 2012 to December 2012 | CH | 17/86 | ICU patients two days post-cardiac surgery | ICU patients two days post-cardiac surgery without infection | 48.1 | 0.77 | PSP levels significantly associated with the presence of infection |
Guadiana-Romualdo et al. [13] | October 2013 to November 2013 | E | 129/23 | Unselected ER patients | Unselected ER patients without infection | 41.5 | 0 84 | PSP performed well as infection biomarker |