Abstract
BACKGROUND & AIMS: The nutritional status of elderly patients requiring ICU admission is largely unknown. This study evaluated the prevalence of malnutrition in elderly patients (>65 years) admitted to the surgical and medical ICUs, agreement between assessment techniques and associations between malnutrition and adverse outcomes. METHODS: For this prospective cohort, nutritional status was classified concurrently using the Mini Nutrition Assessment (MNA), Subjective Global Assessment (SGA), Nutrition Risk Score 2002 (NRS 2002) and MNA-short form (MNA-SF). Demographic and relevant medical information were collected from the medical record prior to the nutrition interview and/or following hospital discharge. Descriptive statistics, inter-rater agreement and regression analyses were conducted. RESULTS: The average patient was 74.2 (±6.8) years of age with a mean APACHE II score of 11.9 (±3.6). Malnutrition was prevalent in 23-34% of patients (n = 260) with excellent agreement between raters. Compared to MNA, NRS 2002 had the highest sensitivity, while SGA and MNA-SF had higher specificity. Malnutrition at ICU admission was associated with longer hospital LOS, a lower propensity for being discharged home and a greater need for hospice care or death at discharge (all p values <0.05). These relationships were diminished when controlling for severity of illness. CONCLUSIONS: Future work in this elderly population needs to explore the role of disease acuity, inflammation and body composition in the nutrition assessment process and in the examination of outcomes.