P58 Online Test-Retest Reliability of the COAST Malnutrition Screening Tool

Abstract

<h3>Background</h3> In face-to-face testing, the Comprehensive Older Adult Screening Tool (COAST) demonstrated internal consistency reliability and validity as a screening tool for malnutrition risk in community-dwelling older adults and its ease of use suggests it is a practical tool. However, the test-retest reliability of COAST has not been established. <h3>Objective</h3> The aim of this study was to determine the test-retest reliability of the COAST malnutrition screening tool in an online setting. <h3>Study Design, Settings, Participants</h3> Older adults (>60 years) were recruited through https://www.researchmatch.org. Interested volunteers were emailed a Qualtrics link to the consent description and the 7-item COAST, which examines weight loss, decreased appetite, illness affecting food intake, diet quality, and intake of protein foods, at 2 timepoints, 2 weeks apart. The COAST was scored from 0 to 8 points, with 0-4 points denoting a high risk of malnutrition, 5-6 points indicating a moderate risk of malnutrition, and 7-8, a low risk of malnutrition. <h3>Measurable Outcomes/Analysis</h3> The primary outcome was the correlation of COAST scores at 2 timepoints. <h3>Results</h3> Of the 947 sent the study description, 111 older adults agreed to be contacted by email. Of these, 49 completed the COAST at 2 timepoints. The COAST test-retest reliability coefficient was 0.74, considered acceptable reliability for a malnutrition screening tool; 75% of participants were similarly categorized as high risk, moderate risk, or low risk of malnutrition at both timepoints. Inconsistent responses to the COAST question, "In general, how healthy is your overall diet?" explained 50% of the misclassifications suggesting a need to clarify this item. <h3>Conclusion</h3> The COAST malnutrition screening tool demonstrated acceptable test-retest reliability when used in an online format. As the tool was previously validated for face-to-face administration, future research to determine the test-retest reliability using face-to-face administration may be warranted.