Effect of Ownership on Hospice Service Use: 2005-2011

Abstract

Objectives: To assess differences in populations and service use according to hospice ownership, chain status, and agency size.

Design: Retrospective cohort study.

Setting: United States.

Participants: Medicare beneficiaries aged 65 and older enrolled in hospice during 2005 to 2011 N = 5,405,526).

Measurements: Hospice use according to ownership category (for-profit nonchain and chain, not-for-profit nonchain and chain, government) and agency size (0-50, 51-200, 201-400, ≥401 individuals discharged each year). Mean length of use, stays of 3 days or fewer, stays ending with live discharge, and decedents receiving no general inpatient care (GIP)- or continuous home care (CHC)-level hospice in the last 7 days of life.

Results: After adjusting for individual and geographic differences, for-profit nonchain and chain agencies had longer mean length of use (84.5 and 91.2 days, respectively) than other agency types (66.3-72.5 days), higher rates of live discharge (21.0% and 20.2% vs 14.6-15.9%), and lower proportions of stays of 3 days or fewer (13.9% and 14.7% vs 16.6-17.5%) (all P < .001). The proportion of decedents not receiving GIP- or CHC-level care before death was highest in for-profit chains (75.9%) and lowest in not-for-profit nonchains (63.2%). Smaller agencies had longer mean length of use, higher live discharge rates, lower rates of stays of 3 days or fewer, and higher rates of individuals receiving no GIP- or CHC-level care. There were considerable differences in patient traits and unadjusted service use between the nation's largest chains.

Conclusion: In addition to for-profit and not-for-profit hospice agencies differing according to important dimensions, there is substantial heterogeneity within these ownership categories, highlighting the need to consider factors such as agency size and chain affiliation in understanding variations in Medicare beneficiaries' hospice care.

Keywords: Medicare; hospice; ownership; profit; quality.

FIGURE 1. Adjusted Hospice Use Outcomes by Ownership Type and Agency Size, 2005–2011

Adjusted Hospice Use Outcomes by Ownership Type and Agency Size, 2005–2011. Figure 1 shows adjusted service use results across ownership types, stratified by agency size categories (0–50; 51–200; 201–400; 401+). Includes first hospice stay only for Medicare beneficiaries aged 65+ ending in death or discharge during the 2005–2011 study period. Figures are based on adjusted regression models that include: age, gender, race/ethnicity, terminal diagnosis category, HMO enrollment status, region of service use, urban/rural status, year, and ownership status of hospice agency. Hospice agencies categorized into one of the size categories for each study year (2005–2011), based on the number of discharged patients in that particular year. Comparisons are made within size categories, across ownership types. Length of use modeled using ordinary least squares regression; other outcomes modeled using logistic regression. All outcomes were significantly different across hospice ownership types (p<0.0001). FP stands for “for-profit”; NFP stands for “not-for-profit”.