Internal and external factors influencing registered dietitians' recommendations for feeding tube use among older adults with advanced dementia: an application of the social ecological model

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
University of Alabama Libraries

Background: While feeding tubes are commonly used to provide nutrition to patients with advanced dementia, research indicates that this fails to improve nutritional status or survival, and often yields harmful complications. As Registered Dietitians (RDs) are often consulted to provide clinical recommendations for older adults with advanced dementia, is important to understand factors influencing RDs’ feeding tube recommendations. Purpose: This study developed and validated a theory-based instrument to assess knowledge, beliefs, attitudes, and perceptions of RDs regarding feeding tube use among older adults with advanced dementia. Additionally, internal and external factors that influenced RDs’ recommendations were explored. Methods: The standardized survey development process included a comprehensive literature review, expert panel review, pilot testing, an efficacy survey, and test-retest analysis. A random sample of U.S. RDs was invited to participate. Exploratory factor and regression analyses determined factors associated with RDs’ feeding recommendations for people with advanced dementia. Results: Of the 662 RDs who completed the survey, 72.2% responded that they were unlikely to recommend feeding tubes for patients with advanced dementia. Factor analysis yielded five factors, each with adequate internal consistency: I) Knowledge Self-Efficacy, II) Religion/Spirituality/Culture, III) Personal Values, IV) Perceived Organization and Training, and V) Perceived Policy. Test-retest correlation coefficients ranged .602 - .812. The multivariate regression analysis included 580 RDs who were either likely or unlikely to recommend a feeding tube (‘neutral’ responses were removed), revealing five factors associated with RDs making evidence-based recommendations: Total Knowledge [OR = 1.40, 95% CI (1.26, 1.57)], Personal Values [OR = 1.30, 95% CI (1.19, 1.43)], Perceived Policy [OR 1.20, 95% CI (1.02, 1.40)], Perceived Organization and Training [OR = .87, 95% CI (.77, .99)], and working in long-term care or hospice settings [OR 3.68, 95% CI (1.51, 8.93)]. This model predicted 53.2% of the variance in RDs’ recommendations. Discussion: The instrument was deemed valid and reliable. Factor analysis indicated that internal and external factors influenced RDs’ recommendations, findings consistent with the Social Ecological Model. Most RDs made recommendations consistent with evidence-based guidelines, an encouraging finding. Work setting and RD knowledge were important modifiable influences, providing direction for future continuing professional education.

Electronic Thesis or Dissertation
Nutrition, Aging, Health sciences