Evolution of end-of-life care at United States hospitals in the new millennium

Abstract

Objective: To examine the characteristics of United States hospitalizations that result in hospice transfers including the clinical and demographic features, and to determine distinctive factors associated with discharges to hospice (DTH).

Methods: The National Inpatient Sample (NIS) databases for 2000-2009 were queried for hospitalizations which resulted in transfers to hospice and expiration in the hospital. Yearly totals, as well as demographic and clinical features were tabulated for DTH hospitalizations. These characteristics were also compared with hospitalizations that ended with expiration using multivariate regression.

Results: The number of DTH per year increased 15 fold from 27,912 in 2000 to 420,882 in 2009. The median hospital stay decreased, while the median age, proportion of sepsis disease related groups (DRGs), and proportion of Medicare hospitalizations increased. Lung, gastrointestinal, hepato-biliary, and brain cancer were consistently the most prevalent malignancy DRGs. However, the initial preponderance of hospitalizations with any diagnosis of cancer was diminished by the end of the study. The adjusted odds ratio (95%CI) for the prediction of DTH (compared to hospital death) by any diagnosis of cancer decreased from 3.61 (3.52-3.71) to 2.02 (2.00-2.04) from the years 2000-2009. Female gender, age, and chronic obstructive pulmonary disease were predictors of discharge to hospice, while congestive heart failure was inversely associated.

Conclusions: Hospital discharges to hospice have increased over the past ten years, with a concomitant shift in clinical and demographic characteristics. A growing trend toward offering and adopting hospice care upon discharge from US hospitals will likely impact health care finance and quality of care measures.