Change in end-of-life care for Medicare beneficiaries: site of death, place of care, and health care transitions in 2000, 2005, and 2009

Abstract

Importance: A recent Centers for Disease Control and Prevention report found that more persons die at home. This has been cited as evidence that persons dying in the United States are using more supportive care.

Objective: To describe changes in site of death, place of care, and health care transitions between 2000, 2005, and 2009.

Design, setting, and patients: Retrospective cohort study of a random 20% sample of fee-for-service Medicare beneficiaries, aged 66 years and older, who died in 2000 (n = 270,202), 2005 (n = 291,819), or 2009 (n = 286,282). A multivariable regression model examined outcomes in 2000 and 2009 after adjustment for sociodemographic characteristics. Based on billing data, patients were classified as having a medical diagnosis of cancer, chronic obstructive pulmonary disease, or dementia in the last 180 days of life.

Main outcome measures: Site of death, place of care, rates of health care transitions, and potentially burdensome transitions (eg, health care transitions in the last 3 days of life).

Results: Our random 20% sample included 848,303 fee-for-service Medicare decedents (mean age, 82.3 years; 57.9% female, 88.1% white). Comparing 2000, 2005, and 2009, the proportion of deaths in acute care hospitals decreased from 32.6% (95% CI, 32.4%-32.8%) to 26.9% (95% CI, 26.7%-27.1%) to 24.6% (95% CI, 24.5%-24.8%), respectively. However, intensive care unit (ICU) use in the last month of life increased from 24.3% (95% CI, 24.1%-24.5%) to 26.3% (95% CI, 26.1%-26.5%) to 29.2% (95% CI, 29.0%-29.3%). (Test of trend P value was <.001 for each variable.) Hospice use at the time of death increased from 21.6% (95% CI, 21.4%-21.7%) to 32.3% (95% CI, 32.1%-32.5%) to 42.2% (95% CI, 42.0%-42.4%), with 28.4% (95% CI, 27.9%-28.5%) using a hospice for 3 days or less in 2009. Of these late hospice referrals, 40.3% (95% CI, 39.7%-40.8%) were preceded by hospitalization with an ICU stay. The mean number of health care transitions in the last 90 days of life increased from 2.1 (interquartile range [IQR], 0-3.0) to 2.8 (IQR, 1.0-4.0) to 3.1 per decedent (IQR, 1.0-5.0). The percentage of patients experiencing transitions in the last 3 days of life increased from 10.3% (95% CI, 10.1%-10.4%) to 12.4% (95% CI, 12.3%-2.5%) to 14.2% (95% CI, 14.0%-14.3%).

Conclusion and relevance: Among Medicare beneficiaries who died in 2009 and 2005 compared with 2000, a lower proportion died in an acute care hospital, although both ICU use and the rate of health care transitions increased in the last month of life.

Figure

Medicare Service Types and Locations in the Last Month of Life for Medicare Fee-for-Service Decedents in 2009 Overall, nearly one-half of decedents experienced a transition in the last 2 weeks of life. Decedents with a diagnosis of cancer experienced increases in the use of hospice services, especially in the last week of life, while decedents with a diagnosis of chronic obstructive pulmonary disease (COPD) often transitioned to an acute care hospital. Decedents with dementia were predominantly in a nursing home with transitions to hospice services in last week of life.