End-of-life Care in Nursing Homes: Does Organization of Work Make a Difference?

Thesis (Ph.D.)--University of Rochester. School of Medicine & Dentistry. Dept. of Community and Preventive Medicine, 2011.
Nursing home residents are frequently transferred to hospitals at the end-of life and many die there. Furthermore, many residents with severe cognitive impairment receive feeding tubes at the end-of-life. The prevalent use of these aggressive treatments at the end-of-life has been a public health and policy concern. This dissertation examines the influence of nursing home organization of work on two care processes: 1) hospitalization among residents at the end-of-life and 2) feeding tube insertion among end-of-life residents with severe cognitive impairment. Four measures of work organization are examined: hospice concentration (the percentage of residents receiving hospice in the last 30 days of life), staff perceived work effectiveness, use of direct care teams, and the use of consistent assignment. The analyses are conducted at the individual-level, employing seven datasets. These datasets include: the Medicare beneficiary file, Minimum Data Set, Medicare inpatient claims and Medicare hospice claims for 2005-2007; a nursing home staff survey (n=7,418) collected in 162 New York state facilities in 2006-2007; and the provider of services and area resource files for 2007. The relationships between hospice concentration and the care processes were tested on national samples of decedent nursing home residents. Instrumental variable models were employed to address the endogeneity between hospice concentration and the care processes. The associations between the other three measures—perceived work effectiveness, the use of direct care teams and the use of consistent assignment—and the two care processes were tested on samples of decedent nursing home residents in New York State. Multivariate logit models were fit to examine the associations. The findings show that residents in facilities with higher hospice concentration, higher (i.e. better) staff perceived work effectiveness and greater use of direct care teams are at lower risk of end-of-life hospitalization. The ass