The purpose of this study is to determine whether homeless individuals will complete a living will and whether guidance enhances rates of completion and to see whether filling out a living will has any impact on the care given at the end of life.

Much research and effort have been directed towards improving end-of-life (EOL) care in the US. However, there is consensus that large gaps in knowledge remain regarding quality EOL care, most notably because previous research has focused on narrowly defined populations. There is little known about individuals who are from different ethnic, cultural, and socioeconomic backgrounds. There is some evidence that these individuals may desire and experience EOL care differently and may encounter barriers to obtaining this care. What is not known is what interventions are effective in these populations. This study addresses EOL care for homeless people, the most extreme examples of these overlooked populations. They have the greatest risk of death in the US, have unique perspectives of good EOL care, encounter significant barriers to care, and lack the resources and relationships assumed necessary for good EOL care. Despite this, there are no interventional studies determining whether EOL care can be improved in this population. Encouraging individuals to document preferences for EOL care through Advance Directives (ADs), has been central to efforts to improve EOL care. ADs, however, have been subject to conceptual and empirical criticism, particularly that they may not ultimately affect EOL care received. However, these shortcomings have been mainly demonstrated in specific and similar populations, and has not been addressed in underserved populations. There are reasons to believe that ADs and advance care planning have different utility and efficacy for homeless persons and are much desired in this population. This proposal builds upon four years of research, including an NIH/NINR-funded investigation, regarding EOL care, dying, and death. This research is some of the first work defining the concerns of homeless persons regarding EOL care and confirmed the relevance and potential of ADs in this population to impact EOL care positively. We seek to test an intervention to improve EOL care for homeless persons: 300 participants will be randomized into two arms, guided intervention (GI) and minimal intervention (MI). The GI condition will expose participants to education, guidance and counseling, and an AD tool, designated HELP (Appendix A), specifically designed for underserved and estranged populations to enhance autonomy and elicit preferences of EOL care. The MI condition will expose participants to the provision of HELP and instructions, mimicking community standards and the requirements of the federal PSDA.

Primary Aim 1: To determine whether homeless individuals will complete an AD and whether guidance enhances rates of completion. Hypothesis 1: We hypothesize that the MI will result in greater completion rate in this population than most others studied, and that the GI will result in a significantly greater rate than the MI - we anticipate a 20% completion rate in the MI condition and a 50% rate in the GI condition. Primary Aim 2: To determine whether completion of ADs by homeless individuals lead to their use during county hospital care encounters over 18 months after the intervention. Hypothesis 2: We hypothesize that when completed ADs exist, they will be used or documented as part of clinical care at least 70% of the time during county health care encounters.

It is important to test appropriate, reproducible interventions in underserved populations, and insights from this project will be significant in several ways: they will help address the EOL concerns of homeless persons, and provide the basis to address the needs of others who are disenfranchised from loved ones and/or receive fragmented, episodic care, such as the rural and urban poor, homebound persons, or undocumented persons.

Inclusion Criteria:

• Homeless in the previous 6 months as defined by the federal McKinney Act
• At least 18 years old
• Speak English
• Able to give valid informed consent

Exclusion Criteria:

• Actively intoxicated
• Cognitive deficiencies using BLESSED tool