Background: Death in the U.S. frequently occurs in institutions despite the overwhelming majority of persons who state that they prefer to die at home. Little research to date has examined how well individual preferences compare to actual site of death.
Objectives: Determine the concordance between preferred and actual place of death and examine independent predictors for concordance.
Design: Observational cohort study.
Setting: Three area hospitals including a safety net hospital, veterans' hospital, and academic tertiary referral center.
Patients: 458 adult patients admitted to the general medical service from 2003-2005.
Measurements: Patients were asked where they preferred to spend their last days of life. Data on date and actual site of death from 2005-2009 was collected from hospital records and death certificates.
Results: The majority of patients preferred to die at home (75% n = 343). Low income and being married were significantly associated with a preference to die at home compared to nursing home or inpatient hospice (OR 2.71 95% CI 1.30-5.67 and OR 2.44 95% CI 1.14-5.21 respectively). Of the 123 patients who died during the follow up period, most (66% n = 80) died in an institutional setting. Overall concordance between preferred and actual site of death was only 37% (n = 41). Female gender was significantly associated with concordance between preferred and actual site of death (OR 3.30 95% CI 1.25-8.72).
Conclusions: Concordance between preferred and actual site of death is low and female gender was the sole patient level variable associated with concordance.
Copyright © 2013 Society of Hospital Medicine.