Clinical microbiology and infection : the official publication of the European Society of Clinical Microbiology and Infectious Diseases 2018 03 29() pii S1198-743X(18)30294-5
To identify sensitivity, specificity and predictive accuracy of quick sequential organ failure assessment (qSOFA) score and systemic inflammatory response syndrome (SIRS) criteria to predict in-hospital mortality in hospitalized patients with suspected infection.
This meta-analysis followed MOOSE consensus statement for conducting and reporting the results of systematic review. PubMed & EMBASE were searched for the observational studies which reported predictive utility of qSOFA score for predicting mortality in patients with suspected or proven infection with the following search words: ‘qSOFA’, ‘q-SOFA’, ‘quick- SOFA’, ‘Quick Sequential Organ Failure Assessment’, ‘quick SOFA’. Sensitivity, specificity, area under receiver operating characteristic curves (ROC) with 95% confidence interval of qSOFA and SIRS criteria for predicting in-hospital mortality was collected for each study and a 2×2 table was created for each study.
Data of 406802 patients from 45 observational studies have been included in this meta-analysis. Pooled sensitivity (95% CI) and specificity (95% CI) of qSOFA≥2 for predicting mortality in patients who are not in intensive care unit (ICU) is 0.48(0.41- 0.55) and 0.83(0.78- 0.87) respectively. Pooled sensitivity (95% CI) of qSOFA ≥2 for predicting mortality in patients (both ICU and non- ICU setting) with suspected infection is 0.56(0.47- 0.65) and pooled specificity (95% CI) is 0.78(0.71-0.83).
qSOFA has been found to be a poorly sensitive predictive marker for in-hospital mortality in hospitalized patients with suspected infection. It is reasonable to recommend developing another scoring system with higher sensitivity to identify high-risk patients with infection.