DEVELOPMENT OF AN INTEGRATED SCREENER FOR UNDERNUTRITION WITHIN A COMPREHENSIVE GERIATRIC ASSESSMENT SYSTEM

Abstract

Objective: To develop a screener for the presence of undernutrition in older adults in acute care utilizing items within a comprehensive geriatric assessment (CGA) instrument (the interRAI Acute Care). Design: Prospective cohort study and retrospective medical record review of nutritional assessment data. Setting: Acute care tertiary teaching hospital in Brisbane, Australia. Participants: Five hundred fifty-seven general medical patients aged 70 and older admitted to the hospital. Measurements: Prevalence of geriatric syndromes at admission; measures of functional status (activities of daily living), cognition, behavioural symptoms, social support, community assistance services, health conditions, medications and other medical treatments, weight, body mass index (BMI), mode of nutritional intake; demographic variables and Subjective Global Assessment (SGA) of nutritional status. These measures were tested for their prediction of undernutrition using a logistic regression model and decision tree analysis. Results: The following variables were significant independent predictors of undernutrition on admission, after adjustment for age and gender: (i) feeling sad/depressed (OR: 3.494 [1.124-10.864]; p<0.05); (ii) short term memory recalling ability (OR: 3.325 [1.152-9.594]; p<0.05); (iii) weight loss of 5% or more in the last 30 days or 10% or more in the last 180 days (OR: 2.877 [0.983-8.416]; p=0.05); (iv) fatigue (OR: 3.494 [1.414-43.205]; p<0.05). Decision tree analysis revealed two models most predictive of undernutrition: (i) short term memory recalling ability and depression (AUC 72.8% [95% CI: 65%-80.6%]); (ii) short term memory recalling ability and recent weight loss (5% or more in the last 30 days or 10% or more in the last 180 days) (AUC: 74.8% [95% CI: 65.9% - 83.6%]). Conclusion: Several measures within the interRAI-AC may be used as part of a screener for undernutrition in acute hospital patients aged 70 years or older. The combination of short term memory recalling ability and percentage weight loss provides the most statistically robust screener for undernutrition within the interRAI-AC.