LAUSR

OBJECTIVE To evaluate the substitution of Acute Physiology and Chronic Health Evaluation II (APACHE II) by Simplified Acute Physiology Score 3 (SAPS 3) as a severity marker in the modified version of the NUTrition RIsk in the Critically ill score (mNUTRIC); without interleukin 6) based on an analysis of its discriminative ability for in-hospital mortality prediction. METHODS This retrospective cohort study evaluated 1,516 adult patients admitted to an intensive care unit of a private general hospital from April 2017 to January 2018. Performance evaluation included Fleiss' Kappa and Pearson correlation analysis. The discriminative ability for estimating in-hospital mortality was assessed with the Receiver Operating Characteristic curve. RESULTS The sample was randomly divided into two-thirds for model development (n = 1,025; age 72 [57 - 83]; 52.4% male) and one-third for performance evaluation (n = 490; age 72 [57 - 83]; 50.8% male). The agreement with mNUTRIC was Kappa of 0.563 (p < 0.001), and the correlation between the instruments was Pearson correlation of 0.804 (p < 0.001). The tool showed good performance in predicting in-hospital mortality (area under the curve 0.825 [0.787 - 0.863] p < 0.001). CONCLUSION The substitution of APACHE II by SAPS 3 as a severity marker in the mNUTRIC score showed good performance in predicting in-hospital mortality. These data provide the first evidence regarding the validity of the substitution of APACHE II by SAPS 3 in the mNUTRIC as a marker of severity. Multicentric studies and additional analyses of nutritional adequacy parameters are required.