NUTRIC-S proposal: Using SAPS 3 for mortality prediction in nutritional risk ICU patients

• Main Authors: Diogo Oliveira Toledo, João Manoel Silva Junior, Jerusa Márcia Toloi, Thaisa de Assis, Luciana Martins Serra, Paula Glaucieli do Carmo, Flavia Julie do Amaral Pfeilsticker, Dyaiane Marques dos Santos, Branca Jardini de Freitas, Amanda M.R. de Oliveira, Daren K. Heyland
• Format: Article
• Language: English
• Published: Elsevier 2020-06-01
• Series: Clinical Nutrition Experimental
• Subjects: NUTRIC; SAPS 3; Critically ill patients; Mortality
• Online Access: http://www.sciencedirect.com/science/article/pii/S235293931930137X

Summary: Objectives: The Nutrition Risk in the Critically Ill score (NUTRIC) identify patients who will benefit the most from optimal nutrition therapy and it is composed by the variables age, number of comorbidities, days from hospital to Intensive Care Unit (ICU) admission, SOFA (Sequential Organ Failure Assessment), APACHE II (Acute Physiology and Chronic Health Evaluation) and interleukin-6 level (IL-6) as an optional variable. However, the APACHE II has been superseded in clinical practice by the Simplified Acute Physiology Score 3 (SAPS 3) as a measure of illness severity. Therefore, this study aimed to propose a new modification on NUTRIC scoring system by replacing APACHE II for SAPS 3, entitled as NUTRIC-S score. Method: This was a prospective observational study, carried out at two tertiary care, surgical-medical ICUs. Patients at least 18 years of age were enrolled within 24h of admission to the ICU and were followed up to hospital discharge. Patients who remained less than 24h in the ICU and those readmitted during the study period were excluded. The NUTRIC-S score predictive ability in differentiating survivors from non-survivors by using SAPS 3 instead of APACHE II, was compared to NUTRIC score (without IL-6). In order to detect the SAPS 3 cut-off points, that would represent the similar APACHE II cut-off points used in NUTRIC score, a linear regression model was performed between the two variables for the sample. Results: A total of 281 patients were enrolled in this study. The patients average age was 62.6 ± 17.0 years-old; 51.2% were female. The average BMI was 25.8 ± 5.5 Kg/m², classified into 38.8% healthy weight range, 27.4% obese, 19.9% underweight and 13.9% overweight. The correlation between NUTRIC and NUTRIC-S was R2 = 0.75 (CI 95% 0.69–0.80, p < 0.001). The ROC curve for predicting hospital mortality was 0.62 for the NUTRIC versus 0.70 for the NUTRIC-S (the difference between the areas was 0.08, 95% CI - 0.01–0.09, p = 0.14). The area under the ROC curve for APACHE II was 0.65, while for SAPS 3 it was 0.71 and 0.68 for SOFA. Patients who had NUTRIC-S score ≥5 presented a higher probability of hospital mortality (HR = 1.76, 95% CI 1.16 to 2.66, p = 0.008) whereas no such relationship was observed with NUTRIC (HR = 1.08,95% CI 0.97 to 1.21, p = 0.17). Conclusion: Compared to the NUTRIC score, the proposed NUTRIC-S scoring system is feasible and may be superior with respect to mortality prediction. NUTRIC-S can be used in those settings where APACHE II is not available.