Clinical value of Oxford acute severity of illness score in identifying quick sequential organ failure assessment-negative patients with sepsis

Abstract

Objective: The quick sequential organ failure assessment (qSOFA) score has been proposed as part of diagnostic criteria for sepsis, but there might be an underestimation of the incidence of qSOFA-negative sepsis according to a few recent studies. The purpose of this study was to evaluate the value of Oxford acute severity of illness score (OASIS) in identifying qSOFA-negative (qSOFA<2) patients with sepsis. Methods: Sepsis patients with negative qSOFA scores were selected from the Medical Information Mart for Intensive Care Ⅲ(MIMIC-Ⅲ), a database comprising de-identified health-related data from patients staying in the critical care units of Beth Israel Deaconess Medical Center between 2001 and 2012. Non-infectious patients with both qSOFA and SOFA scores less than 2 were enrolled as controls. Propensity score matching (PSM) analysis was used to reduce the effects of selection bias. Receiver operating characteristic (ROC) curve analysis was performed to assess the value of OASIS in discriminating qSOFA-negative patients with sepsis and to determine its optimal cut-off. Associations of OASIS with 28-day mortality after intensive care unit (ICU) admission, ICU mortality, and hospital mortality were further examined using multivariate Cox regression and multivariate logistic regression analysis. Results: 2 273 qSOFA-negative patients with sepsis and 3 342 non-sepsis controls were included finally with a PSM cohort consisted of 1 677 sepsis patients and 1 677 controls. Results of ROC analysis showed that the area under ROC curve was 0.753 [95% confidence interval (CI) 0.741-0.765] and the optimal OASIS threshold according to the Youden index was 26.5 and yielded a 67.2% sensitivity and 70.8% specificity. Multivariate regression analysis indicated that OASIS>26 was an independent risk factor for 28-day mortality (hazard ratio 2.80, 95% CI 2.15-3.65, P<0.01), ICU mortality (odds ratio 4.69, 95% CI 2.60-8.49, P<0.01), and hospital mortality (odds ratio 4.48, 95% CI 3.13-6.42, P<0.01). Analysis of the PSM cohort presented consistent results. Conclusions: OASIS had a good discriminative value to differentiate qSOFA-negative patients with sepsis from those without sepsis.