Abstract

Given the great difficulty in accurately estimating the life expectancy of an individual with advanced dementia, many do not receive the appropriate palliative care they require. This study identifies the factors that are associated with the six-month survival of newly admitted nursing home residents with advanced dementia, as well as develops an index which helps predict the six-month mortality of such individuals. This study found that demographic, functional status, and other health variables are associated with the six-month survival of admission into a nursing home. As well, the presence of such factors can greatly increase the risk of mortality within these six months. These findings can ultimately aid in the effective provision of end of life care to individuals with advanced dementia who are in need. Background: The ability to accurately estimate the life expectancy of individuals with advanced dementia is important in guiding decision making surrounding the provision and delivery of palliative care. Given the gradual destruction of the individual's ability to communicate, learn and make judgments, many nursing home residents with advanced dementia die before they before they are able to receive palliative care. This focus of this study was to create a model to predict mortality of individuals with advanced dementia. This study developed and validated a model, based on various demographic, functional status and health factors, which can help predict the six-month mortality for nursing home residents with advanced dementia.

Method: This is a retrospective study of Minimum Data Set (MDS) data of all Medicaid and Medicare nursing home residents aged 65 years and older with advanced dementia in New York from June 1, 1994 to December 30, 1998 and in Michigan from October 1, 1998 to July 30, 2000. Advanced dementia is defined as having a diagnosis of dementia, a low Cognitive Performance Score and very severe impairment with the activities of daily living. A model was created using mortality data from the National Death Index and MDS information in order to determine which residents were alive beyond six months of their admission to a long-term care facility. Analysis was first conducted to determine the association between six-month survival and a variety of factors, including demographic variables, functional status variables, diagnosis variables and other health conditions variables. The final model included factors such as age, sex, the presence of cancer or heart disease, the need for oxygen therapy, shortness of breath, unstable condition, and functional status. Survival curves were then derived based on these risk scores, in order to ascertain the predictability of mortality within the six-month of admission to a nursing home

Findings: In the New York and Michigan nursing home cohorts, 28.3% and 35.1% of residents with advanced dementia died within six months of admission, respectively. Based on the results from the regression analysis, all independent factors included in the equation were significantly associated with survival, except for one diagnosis variable (the presence of a fracture in the previous 180 days) and one health condition variable (the presence of hallucinations or delusions). With respect to the index, it was observed that the risk of mortality increased with higher index scores. An index score of zero is associated with a 8.9% risk of death within six month of admission to the nursing home; a score is 1-2 is associated with a 10.8% risk; a score of 3-5 is associated with a 23.3% risk; a score of 6-8 is associated with a 40.4% risk; a score of 9-11 is associated with a risk of 57%; and a score greater than 12 is associated with a risk of 70%.

Conclusions: The study determines which factors are associated with surviving beyond six-months of admission to nursing home, as well as which factors contribute to increased mortality within the six months. The model to predict mortality is easy to use with MDS data and its ability to accurately predict death was found to be very good. Many nursing home residents with advanced dementia are dying before they have the chance to receive proper end-of-life care. As of yet, MDS is not used in Ontario long-term care analysis, however if implemented, this study, can be useful in helping to predict the survival of individuals with dementia and help guide the decision-making process for the proper delivery of palliative care to individuals who require it.

Reference: Mitchell S., Kiely D., Hamel M.B., Park P., Morris J., and Fries B. "Estimating prognosis for nursing home residents with advanced dementia". Journal of the American Medical Association, 2004; 291(22):99-111.