To compare the predictive value of the simplified acute physiology score (SAPS) Ⅱ and the Oxford acute severity of illness score (OASIS) for short-term outcomes of patients in ICU and to provide evidence for selection of disease severity scoring system in clinical practice.

Data including baseline information, vital signs, and some laboratory test results of adult (age ≥ 18 years) ICU patients between January 2001 to December 2012 were extracted from a freely accessible critical care database (MIMIC-Ⅲ) and SAPS Ⅱ and OASIS scores of each patients were calculated according to the requirements of each scoring system. ICU mortality was our primary outcome and receiver operating characteristic (ROC) analysis was performed to determine the predictive performance by comparing the areas under ROC curves (AUC).

A total of 38 468 ICU adult patients were included finally, of which male patients accounted for 56.61% with a median age of 65.72 years old and an ICU mortality rate of 8.28% (3185/38468). When compared with the survivors, non-survivors had higher SAPS Ⅱ (survivors vs non-survivors 32 vs 51, H=3473.792, P<0.001) and OASIS scores (survivors vs non-survivors 30 vs 41, H=3422.382, P<0.001) and higher rates of mechanical ventilation (survivors vs non-survivors 22.76% vs 73.59%, χ²=3831.865, P<0.001). The ROC curve analysis showed that the AUC of SAPS Ⅱ score and OASIS score were 0.8147 (95%CI 0.8068-0.8226) and 0.8123 (95%CI 0.8042-0.8204), respectively, but the Hanley-McNeil test showed no significant difference of AUC between the two scoring systems (Z=0.686, P=0.4928).

There is no significant difference in the predictive value of OASIS scores and SAPS Ⅱ scores for short-term prognosis of adult ICU patients, suggesting that OASIS, a simpler scoring system, might be another option for disease severity scoreing in ICU.