Context Survival varies for patients with advanced dementia, and accurate prognostic tools have not been developed. A small proportion of patients admitted to hospice have dementia, in part because of the difficulty in predicting survival.

Aspiration pneumonia is a serious problem for the elderly institutionalized person, often requiring transfer to a hospital and a lengthy stay there. It is associated with a high mortality rate and is very costly to the health care system. The current study sought to determine the key predictors of aspiration pneumonia in a nursing home population with the hope that health care providers could identify those residents at highest risk and focus more efforts on prevention of this serious disease. A cross-sectional, retrospective analysis was done, using the Minimum Data Set (MDS) nursing home assessment data for three states (New York, Mississippi, Maine) from 1993 to 1994 (N = 102842). Nursing home residents were aged 65+. Standardized MDS summary scales and their component items were used, including: the Activities of Daily Living (ADL) scale, the cognitive performance scale (CPS), and the Resource Utilization Groups (RUGs). Results of these analyses showed the prevalence of pneumonia among this population was 3% (n = 3118). Results from the logistic regression models indicated 18 significant predictors of aspiration pneumonia. The strongest to weakest predictors of pneumonia were, respectively, suctioning use, COPD, CHF, presence of feeding tube, bedfast, high case mix index, delirium, weight loss, swallowing problems, urinary tract infections, mechanically altered diet, dependence for eating, bed mobility, locomotion, number of medications, and age, while both CVA and tracheotomy care were inversely predictive of pneumonia. The emergence of these significant predictors suggested a different pathogenesis of pneumonia in the elderly nursing home resident from the acute care patient or the outpatient. Nursing home residents have chronic medical conditions that gradually lead to "decompensation" in functional status, nutritional status, and pulmonary clearance. Dysphagia and aspiration are common complications of their medical conditions and may slowly worsen as their status deteriorates. Alternatively, a sudden adverse event may dramatically increase the amount aspirated or the ability to resist infection and lead to sudden decompensation. Clinical staff must identify residents with dysphagia and aspiration and work to prevent decline in functional status in all residents. They must be aware of the dangers of adverse events that lead to sudden inactivity or illness and increase the risk of aspiration pneumonia. Prevention of this disease whenever possible will reduce costs, improve health outcomes, and improve our quality of care.

Conclusions: Persons dying with advanced dementia admitted to nursing homes have different characteristics compared to those admitted to home care services. Their end-of-life experiences also differ in these two sites of care. However, palliative care was not optimal in either setting.

Several outcomes for nursing home residents improved after implementation of the RAI. Of the four conditions for which there are significant declines in prevalence or outcome changes, three are specifically addressed in the care planning guidelines incorporated the RAI system (all except stasis ulcer, although there is a RAP for decubitus ulcer). Pain, the only other condition with a significant result --an increase in baseline prevalence--also has no RAP. Although the changes might be ascribed otherwise, they support the premise that the RAI has directly contributed to improved outcomes for nursing home residents.

Background The COVID‐19 pandemic significantly disrupted nursing home (NH) care, including visitation restrictions, reduced staffing levels, and changes in routine care. These challenges may have led to increased behavioral symptoms, depression symptoms, and central nervous system (CNS)‐active medication use among long‐stay NH residents with dementia. Methods We conducted a retrospective, cross‐sectional study including Michigan long‐stay (≥100 days) NH residents aged ≥65 with dementia based on Minimum Data Set (MDS) assessments from January 1, 2018 to June 30, 2021. Residents with schizophrenia, Tourette syndrome, or Huntington's disease were excluded. Outcomes were the monthly prevalence of behavioral symptoms (i.e., Agitated Reactive Behavior Scale ≥ 1), depression symptoms (i.e., Patient Health Questionnaire [PHQ]—9 ≥ 10, reflecting at least moderate depression), and CNS‐active medication use (e.g., antipsychotics). Demographic, clinical, and facility characteristics were included. Using an interrupted time series design, we compared outcomes over two periods: Period 1: January 1, 2018–February 28, 2020 (pre‐COVID‐19) and Period 2: March 1, 2020–June 30, 2021 (during COVID‐19). Results We included 37,427 Michigan long‐stay NH residents with dementia. The majority were female, 80 years or older, White, and resided in a for‐profit NH facility. The percent of NH residents with moderate depression symptoms increased during COVID‐19 compared to pre‐COVID‐19 (4.0% vs 2.9%, slope change [SC] = 0.03, p < 0.05). Antidepressant, antianxiety, antipsychotic and opioid use increased during COVID‐19 compared to pre‐COVID‐19 (SC = 0.41, p < 0.001, SC = 0.17, p < 0.001, SC = 0.07, p < 0.05, and SC = 0.24, p < 0.001, respectively). No significant changes in hypnotic use or behavioral symptoms were observed. Conclusions Michigan long‐stay NH residents with dementia had a higher prevalence of depression symptoms and CNS active‐medication use during the COVID‐19 pandemic than before. During periods of increased isolation, facility‐level policies to regularly assess depression symptoms and appropriate CNS‐active medication use are warranted.