For the half million older adults who die in U.S. nursing homes (NHs) each year, management of their complex physical, psychosocial and spiritual needs is often inadequate. Palliative care (PC) in the form of hospice care aims to address these needs. However, while there has been much growth in NH hospice use, there remains important barriers (policy and other) to its use and timely use. To extend PC to NH residents unable or unwilling to access hospice, a new model of NH PC provision has emerged-PC consultations provided through hospice-affiliated PC programs. Using a novel approach which addresses existing methodological challenges, the proposed R21 study will be the first to compare how PC consults in NHs are associated with key outcomes. It will also explore creation of a methodology to identify NH recipients of PC consults using data available for Medicare beneficiaries across the U.S., enabling population-based research and potentially advancing the study of other NH care. The study's long-term goal is to provide generalizable evidence on the effectiveness of PC consults in NHs. The overall objective of this application is to understand how the use of PC consults in NHs is associated with hospital and emergency room (ER) use at and near the end of life, and with feeding tube use. Our central hypothesis is that at the resident- and NH-level PC consults will be associated with lower hospital and ER use near and at the end of life, and for dementia decedents, with less feeding tube use. The proposed study will pursue the following three specific aims: 1) Using 2006-09 data on an estimated 800 NH decedents with known PC consults in Rhode Island and 2 North Carolina counties and on NH-propensity matched (non-consult) decedents, test differences in hospital and ER use, hospital deaths, and for dementia decedents, in feeding tube use;2) Using 2000-2010 data on NH use of PC consults, test how NH introduction of PC consults and increases in their use are associated with NH rates of end-of-life hospital and ER use, hospital deaths, and for dementia decedents, with rates of tube feeding;and 3) In order to generalize the results from Aim 1 (derived from our convenience sample of decedents), we will explore the possibility of identifying recipients of PC consults using an algorithm(s) derived and validated with Medicare claims and MDS data. This study is innovative because it uses uniquely available data from our collaborating providers and from Brown University (data from the NIA-funded "Shaping Long Term Care in America" program grant, #P01 AG027296) to enable the first comparative study of this model of PC in NHs;and, it explores a methodology to identify receipt of PC consults in NHs which enables population-based research, and potentially advances the study of other NH care. This study is significant because it has the potential to not only improve the quality of car for the very vulnerable populations of older Americans who reside in and die in NHs, but also to reduce Medicare costs through fewer end-of-life hospitalizations and ER visits.
The proposed study is highly relevant to public health because it will generate findings and a methodology capable of having an important positive impact on access to palliative care in nursing homes. Due to population aging and inadequate attention to the complex needs of aging populations, the World Health Organization believes the achievement of more widespread access to palliative care is an important public health issue. This proposal is consistent with the Mission of the National Institute on Aging in that it addresses the special needs of the vulnerable population of older Americans who end their lives in nursing homes.