Exploring End-of-Life Nutrition Care for Residents of Long-Term Care: A Retrospective Chart Review

Introduction: Eating challenges are prevalent in long-term care (LTC) and have been associated with mortality, particularly in older adults with advanced conditions such as dementia. However, changes to eating habits and nutrition care practices to support LTC residents who experience difficulties with eating have not been described within the context of end-of-life decline. Despite a general move towards palliative approaches to care in LTC, research on the implementation of comfort-focused nutrition care practices is lacking to date and the implications of a palliative approach to nutrition care is unknown. Nutrition care is an ideal target for palliative-focused initiatives because of the deeper meaning often placed on food and mealtimes as symbolic of relationships and care and the additional emotional challenges associated with health decline and death. When poorly addressed, severe eating challenges can cause distress among residents and care partners and contribute to decreased quality of life for residents and care providers. As such, best practices to support residents with severe eating challenges moving towards the end of life are needed to maintain quality of life for residents and provide positive care experiences for care providers, but the end-of-life nutrition care experience must first be understood. Towards this end, this dissertation aims to describe longitudinal changes in various aspects of the nutrition care experience for LTC residents approaching the end of life. \nMethods: Data for this study were obtained from a retrospective resident chart review of a convenience sample of deceased residents from 18 LTC homes in southern Ontario, Canada. Two sampling frames, the “Better tArgeting, Better outcomes for frail ELderly patients (BABEL)” study and a chain of homes were accessed to meet sample size requirements. The latter included random selection of decedents for chart review. Eligible participants were over the age of 65 years at death and living in residence for at least 6 months. Data were collected from electronic charts. Admission forms provided information on demographics (e.g., age at death, sex, length of admission), and resident assessments completed approximately 6 months prior to death provided functional and health characteristics (e.g., cognitive performance, health instability) and diagnoses. Nutrition-related information (e.g., nutrition interventions, eating challenges) were obtained from monthly weight records, progress notes, and care plans, narrowed to four time points at 6-, 3-, 1-month and 2 weeks prior to the date of death. The point at which the first mention of end-of-life decline occurred and the details around this event were also noted. Study 1 described the eating challenges documented at each time point: a mixed effects logistic regression model tested longitudinal within-resident differences in eating challenges, and an unconditional repeated measures latent class analysis identified patterns of change in multiple eating challenges over time. Study 2 described the nutrition care interventions used to support residents and identified resident-level and time-dependent factors associated with comfort-focused nutrition care orders using a mixed methods logistic regression analysis. Study 3 determined which nutrition- and non-nutrition-related challenges factored into the first documentation of suspected end-of-life decline from the care provider perspective and assessed the association of each factor with time to death using a Cox proportional hazards regression analysis. Study 4 described dietitian referrals documented at each of the four time points and identified time-dependent factors, including comfort-focused orders, that were associated with dietitian referrals. \nResults: The sample was comprised of 164 LTC residents (mean age 88.3±7.5 years at death; 61% female) and had a high level of health challenges and functional impairments, including 79.3% with moderate/severe cognitive impairment and 58.5% who left food uneaten at most meals 6 months prior to death. Study 1: All but 1 resident (99.4%, n=163) exhibited one or more eating challenges in the 6 months prior to death. Likelihood of severe and complex eating challenges (e.g., poor appetite, dysphagia, refusing to eat, decreased food intake) increased in the month preceding death compared to 6 months prior. A significant increase in lethargy at mealtimes was found as early as 3 months (Odds ratio [OR] = 1.78, 95% confidence interval [CI] = [1.06, 3.00]) compared to 6 months prior to death, and increased in a stepwise fashion at 1 month (OR = 4.09 [2.47, 6.76]) and 2 weeks (OR = 8.24 [4.92, 13.78]) before death. Four patterns of eating challenges were identified with the repeated measures latent class analysis, labeled for the most prominent features defining each group over time: “refusing” (20.1%), “complex” (18.9%), “progressive” (30.5%), and “end-stage” (30.5%). Study 2: Restorative nutrition interventions (e.g., physical assistance, oral nutritional supplementation) to support oral nutrient intake were also common (99.4% of the sample) in the last 6 months of life, and nearly half of the sample had nutrition care plan modifications between each time point. Despite a high need for nutritional interventions, only 30.5% of the sample received a comfort-focused nutrition care order, and most often these occurred within two weeks of death. Comfort orders were more likely with health instability (OR = 4.35 [1.49, 13.76]), within two weeks of death (OR = 5.50 [1.70, 17.11]), following an end-of-life conversation (OR = 5.66 [2.83, 11.33]), with discontinued nutrition interventions (OR = 6.31 [1.75, 22.72]), with the co-occurrence of other care plan modifications (OR = 1.48 [1.10, 1.98]), and with a higher number of eating challenges (OR=1.19 [1.02, 1.38]), especially dysphagia (OR = 2.59 [1.09, 6.17]), at the preceding time point. Study 3: Several of the eating challenges cited above were documented by care providers to suggest suspected end-of-life decline. Of 7 eating challenges cited, only dysphagia was significantly associated with increased risk of death (HR = 2.99; 95% CI = 1.41, 6.33). Study 4: Dietitian referrals were common (73% of participants) in the last six months of life and were significantly associated with number of eating challenges (OR = 1.38 [1.25, 1.54]). A higher frequency of comfort-focused nutrition care orders was found when a dietitian was referred (24.7%) compared to when not referred (11.9 %) in the final month of life, though a causal association could not be established. \nConclusion: Findings demonstrate that nutrition care is an important aspect of the end-of-life experience for LTC residents and that there is potential room for growth in supporting residents with severe and complex eating challenges. Eating challenges were common and increased in prevalence with proximity to death requiring several modifications to nutrition care plans, while comfort-focused nutrition care plans were used infrequently and often late in the end-of-life trajectory. Given the common occurrence of eating challenges with proximity to death and the evidence that they are considered by care providers as informal signals of end-of-life decline in current practice, changes to eating habits (e.g., decreased food intake) may be useful targets for initiating timely comfort-focused nutrition care conversations. The involvement of dietitians at key moments when comfort-focused nutrition care may be appropriate suggests that dietitians are well-positioned to champion palliative approaches to nutrition care within the multidisciplinary LTC team. Future work should explore residents’ and care providers attitudes towards comfort-focused nutrition care to develop best practices that align with their needs and preferences to support quality of life across the final months of life.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.