Enhancing intensive care outcome prediction with nutrition information available at admission: An analytic observational study conducted in 31 intensive care units throughout Australia and New Zealand

Abstract

Background At admission to an intensive care unit (ICU), patient data are collected to allow the objective estimation of risk of mortality. These risk estimates are used to stratify quality assurance projects and clinical research. The purpose of this study was to determine whether measures of nutrition status could add additional information to this risk stratification process. Methods At 31 ICU’s throughout Australia and New Zealand, in addition to routine patient data, the following measures of nutrition status were collected: Triceps Skinfold Thickness, Mid Arm Muscle Circumference (MAMC), Body Mass Index (BMI), the Subjective Global Assessment (SGA) of Muscle Wasting and SGA Fat Loss. Results 1,363 critically ill patients were enrolled. Controlling for severity of illness and other traditional risk factors, multivariable analysis revealed MAMC, SGA Muscle Wasting and SGA Fat Loss added significantly more information than BMI. With each measure, improved nutrition status was associated with improved outcome. Conclusion This analytic observational study demonstrates the existence of significant independent associations between a patient’s nutrition status and outcome from critical illness. Future research should focus on determining whether this relationship is causal. For example, can improving nutrition status before ICU admission result in improved outcomes from critical illness?