Abstract
Background: This study aimed to determine the agreement between the modified Nutrition Risk in Critically ill Score (mNUTRIC) and the Subjective Global Assessment (SGA) and compare their ability in discriminating and quantifying mortality risk independently and in combination. Methods: Between August 2015 and October 2016, all patients in a Singaporean hospital received the SGA within 48 hours of intensive care unit admission. Nutrition status was dichotomized into presence or absence of malnutrition. The mNUTRIC of patients was retrospectively calculated at the end of the study, and high mNUTRIC was defined as scores ≥5. Results: There were 439 patients and 67.9% had high mNUTRIC, whereas only 28% were malnourished. Hospital mortality was 29.6%, and none was lost to follow-up. Although both tools had poor agreement (κ statistics: 0.13, P < .001), they had similar discriminative value for hospital mortality (C-statistics [95% confidence interval (CI)], 0.66 [0.62–0.70] for high mNUTRIC and 0.61 [0.56–0.66] for malnutrition, P = .12). However, a high mNUTRIC was associated with higher adjusted odds for hospital mortality compared with malnutrition (adjusted odds ratio [95% CI], 5.32 [2.15–13.17], P < .001, and 4.27 [1.03–17.71], P = .046, respectively). Combination of both tools showed malnutrition and high mNUTRIC were associated with the highest adjusted odds for hospital mortality (14.43 [5.38–38.78], P < .001). Conclusion: The mNUTRIC and SGA had poor agreement. Although they individually provided a fair discriminative value for hospital mortality, the combination of these approaches is a better discriminator to quantify mortality risk.
Get full access to this article
View all access options for this article.