Objectives : Malnutrition is a common problem in geriatric patients that often goes unrecognized. Undernutrition is a primary health concern for older adults due to associations with increased mortality, complications, and length of hospital stay. Yet, there is no consensus on which malnutrition screening tool should be used for hospitalized older adults. Therefore, the objective of this study was to determine which screening tool is best to rapidly detect malnutrition in hospitalized older adults so that patient outcomes may be improved.
Methods : Older adult patients (n=211; ≥ 65 yrs old) were enrolled during acute hospitalization. Testing occurring within 72 hours of admission and included the following screening tools included: Malnutrition Screening Tool (MST), Mini Nutritional Assessment Short Form (MNA-SF), Malnutrition Universal Screening Tool (MUST), Nutrition Risk Screening 2002 (NRS-2002), and Geriatric Nutritional Risk Index (GNRI). These screening tools were compared to a malnutrition diagnostic tool, the Subjective Global Assessment (SGA).
Results : According to SGA, 49% of patients were at risk of being malnourished. The other screening tools indicated a wide range of malnutrition prevalence, from 18% (MST) to 76% (MNA-SF). MST (93%) and MUST (92%) were highest in sensitivity. NRS-2002 had moderately good sensitivity (71%). MNA-SF and GNRI had poor sensitivity, eliminating them as good screening tools for hospitalized elderly patients. Of the remaining tools, NRS-2002 had the highest specificity (77%). MST and MUST had poor specificity (31%, 39%, respectively), eliminating them as good screening tools for hospitalized elderly patients. The remaining screening tool, NRS-2002, had moderately good positive and negative predictive values (76%, 72%, respectively). It also had the highest kappa (0.479). Overall, NRS-2002 had the best agreement to SGA and showed moderately good sensitivity, specificity and predictive values.
Our data suggests NRS-2002 is the best malnutrition screening tool for rapid detection of malnutrition in elderly hospitalized patients, when compared to the diagnostic tool, SGA. Future research is needed to determine which screening tool is most effective for use in different settings. Additional research can assist in standardizing malnutrition criteria and care processes.
Funding Sources : National Dairy Council, National Institutes of Health-National Center for Advancing Translational Sciences, and UTMB Claude D. Pepper OAIC