Risk indicators for hospitalization during the last year of life

Objective. High levels of hospital expenditures for older people during their last year of life are widely documented. However, evidence of the association between prospectively measured indicators and subsequent hospitalization is sparse. This article investigates the pattern of hospitalization for a sample of Medicare enrollees during their last year of life. Data Sources. Data from the Longitudinal Study of Aging,a national study of persons age 70 and older, are used. Only data on decedents are used. Study Design. We determine individual characteristics (including functional status, evidence of disease, living arrangement, and prior hospitalization) shortly before the last year of life. A distinction is made between terminal and nonterminal admissions. National estimates and regression analyses using survey weights are conducted. Principal Findings. The likelihood of any use is high regardless of age, functional status, or the presence of major diseases. Although only a few indicators are associated with having a terminal stay, a number of indicators are associated with nonterminal use. Nonterminal stays and total nights hospitalized are positively associated with prior evidence of disease, prior hospitalization, and age, although the probability of nonterminal use decreases with age for persons over 82 years old. The relationship between use and functional status depends on whether persons lived alone, were institutionalized, or had private health insurance. Conclusions. This study demonstrates that while it is difficult to predict who will be admitted to the hospital at the time of death, a number of characteristics existing before the last year of life are associated with nonterminal hospitalization and total nights hospitalized during the last year of life.