Running Head: ADVANCE DIRECTIVE ADVOCACY Advance Directive Advocacy: Empowering Homeless Clients to Express Their End-of-Life Wishes Melissa Morrison, RN, CNRN, DNP-student Arizona State University 2 ADVANCE DIRECTIVE ADVOCACY Abstract The homeless population is aging. Homeless individuals have distinct barriers to equitable endof-life care. Completion of an advance directive would address many of these barriers. A review of the literature indicates that homeless individuals, once educated on the purpose and significance of advance directives, complete them at a higher rate than non-homeless people. One such educational intervention was performed in the setting of a medical respite center and a day resource center, both of which serve the homeless. The goal of the intervention was to increase advance-directive completion by clients via improving knowledge and attitudes about the documents, and to establish an agency protocol for future clients. Twenty-eight clients and 12 staff members completed a pre- and post-test measuring attitudes and knowledge relating to advance directives, their responses measured on a Likert scale. A McNemar’s test discerned an increase in knowledge about a living will (p = .013), while a Wilcoxon’s signed rank test determined that attitudes were improved but not significantly. Two clients completed directives following the intervention. A self-selected participant pool and a one-on-one educational session resulted in more document completion versus a convenience sample. Implications for practice include an inexpensive intervention that does not require a medically trained facilitator, enabling a broad application to a variety of settings, with the goal of empowering a traditionally disenfranchised population to make health decisions related to end-of-life care. Future undertakings should look at creating a validated instrument for measuring knowledge and attitudes about directives among disadvantaged populations. Keywords: homeless, advance directive, end of life 3 ADVANCE DIRECTIVE ADVOCACY Advance Directive Advocacy: Empowering Homeless Clients to Express Their End-of-Life Wishes Homeless individuals suffer the same chronic diseases as their domiciled counterparts, but they have higher mortality rates and less access to care. While homeless people die younger on average, the aging of the Baby Boomer generation means the aging of the homeless population. However, as a disadvantaged population and a distinct culture, they have unique needs at end of life. Barriers to addressing these needs have resulted in a disparity in care. Reduced access to care is a result of lack of health insurance among the homeless, and a culture of mistrust of the medical establishment, as well as pragmatic factors such as no fixed address and limited access to transportation (Song et al., 2006). Health equity not only means equal access to life-extending and life-enhancing care, but to quality end-of-life (EOL) care, as well. For most, that means care that fulfills the wishes of the dying patient as reasonably as is possible including the relief of suffering, both physical and spiritual. The purpose of this paper is to discuss the results of an evidence-based applied research project aimed at improving homeless patients’ access to and participation in quality treatment with the formalization of EOL wishes in an advanced directive (AD). Background and Significance Homelessness is defined as a person (or family) who spends most nights in a place not designed for regular accommodation, such as a car, campground or abandoned building; lives in an emergency shelter; or is leaving a place where he or she lived for less than 90 days after being in one of the prior situations (United States Interagency Council on Homelessness [USICH], n.d.). Nationally, 564,708 people were homeless in America on a given night in January 2015, 68% of whom were over the age of 24 (U.S. Department of Housing and Urban Development ADVANCE DIRECTIVE ADVOCACY 4 [HUD], 2015). Nearly 48,000 were veterans. As the general population ages, the percentage of elderly homeless (defined as older than 65) is expected to do likewise, increasing 33% by 2020, to 58,770 (National Health Care for the Homeless Council [NHCHC], 2013). In Arizona, an estimated 29,170 people experienced homelessness in 2014 – 1 for every 227 residents (Department of Economic Security [DES], 2014). Of them, 36% were over the age of 45. From 2011 to 2014, homelessness among people over the age of 62 in Maricopa County increased 53% (Esperanza, 2015). The majority of Arizona’s homeless in 2014, 79%, were single men, who as a demographic experienced a higher rate of addiction and physical or mental comorbidities – 23 percent and 56 percent, respectively (DES, 2014). This signifies additional challenges at end of life. Just over one-fourth of the American homeless population is estimated to suffer severe mental illness (SMI), defined as a prolonged or recurrent mental illness that impairs activities of daily living and requires long-term treatment (Substance Abuse and Mental Health Services Administration [SAMHSA], 2011; Blatt & Crawford, 2015). As many as 49% of older homeless individuals have SMI (SAMHSA, 2011). Half of individuals with SMI also suffer at least two chronic medical illnesses and are at higher risk of death from cancer, cardiovascular disease, respiratory and gastrointestinal illness (Blatt & Crawford, 2015). SMI is also associated with concurrent drug addiction. For example, 21 to 61 percent of people with bipolar disorder are addicted to substances, compared with 3 to 13 percent of the general population (Blatt & Crawford, 2015). Approximately half of all homeless individuals (and 70% of homeless veterans) are estimated to be substance abusers (USICH, 2013). Heart disease is the leading cause of death among older homeless adults, as it is for all Americans (NHCHC, 2011; Jones et al., 2009; Centers for Disease Control [CDC], 2015). ADVANCE DIRECTIVE ADVOCACY 5 However, homeless men ages 45 to 64 are 40% to 50% more likely to die from heart disease than the general population (NHCHC, 2011). Lower respiratory diseases are the fourth leading cause of death in the general population and affect homeless individuals at twice the rate (NHCHC, 2011). Most chronic diseases, such as congestive heart failure (CHF) and chronic pulmonary obstructive disease (COPD), progress in stages, allowing – theoretically, at least – time for sufferers to determine their goals of care and discuss them with future surrogate decision-makers and medical providers should they reach a stage where they are unable to express them personally (Hemani & Letizia, 2008). The CDC recommends completing an AD in order to do so (CDC, 2015). An AD is typically defined as comprising two documents: a living will and/or a durable healthcare power of attorney, also known as a medical power of attorney (MPOA) or healthcare proxy (American Cancer Society [ACA], 2015). A living will describes the type of medical treatment a person would want in the event of a terminal illness or vegetative state (ACA, 2015). It addresses the patient’s desire for such interventions as feeding-tube placement, dialysis, and ventilators, as well as whether he or she would want cardiopulmonary resuscitation (CPR) in the case of heart or breath cessation. It also describes whether the patient would want aggressive or palliative care in such circumstances. An MPOA names a person the patient would want to make healthcare decisions on his or her behalf (ACA, 2015). Both are legal documents that typically require a witness or notary to validate. In terms of their attitudes about the dying process, homeless individuals express many of the same fears as their domiciled counterparts, such as inadequate pain control and being kept alive futilely on life support. However, some concerns specific to the homeless include fear of their bodies not being found or identifiable, improper body disposal, and of dying anonymously. ADVANCE DIRECTIVE ADVOCACY 6 “My fear is being found on the street, but no one knowing how to help me or who I am,” one person said (Song et al., 2006, p. 437). Homeless patients are also less likely to desire foregoing life-sustaining treatment if comatose or dying – 37% and 31%, compared to 78% to 94% in the general population (Song et al., 2010; Cagle, 2009). A clearly stated and accessible AD would be a significant first step in addressing many of the above concerns. As written statements of a person’s medical wishes, ADs are meant to speak for the person when he or she is no longer capable of doing so. In the case of homeless individuals, who typically are socially isolated and are less likely to have readily available legal proxies, ADs are arguably of greater significance (Song et al., 2010). Eighty-six percent of hospitalized patients over 80 years old with ADs had their EOL wishes respected compared to 30% without them (Detering, Hancock, Reade, and Silvester, 2010). Of patients alive six months after hospital discharge, 94% percent were “very satisfied” with being listened to in the hospital compared to 52% who did not have ADs (Detering et al., 2010). However, only 26% of all Americans have actually completed an AD (National Hospice and Palliative Care Organization [NHPCO], 2013). The Patient Self Determination Act (PSDA), passed by Congress in 1990 and enacted in 1991, mandates that healthcare institutions that receive Medicare and Medicaid funding provide written information to patients about their rights under state law to make decisions about their medical care, including their right to complete an AD. The act also requires organizations to document in the medical record whether the patient has an AD, as well as educate staff and community about them (American Bar Association, n.d.). In addition to a legal obligation, nurses are ethically bound to advocate on behalf of EOL patients for a death that is “congruent with the values and desires of the dying person” (American Nurses Association, 2010, p. 31). ADVANCE DIRECTIVE ADVOCACY 7 Significant fiscal costs are attributed to underuse of ADs. Care of chronic illness in the last two years of life accounts for 32% of Medicare spending (Dartmouth Atlas of Healthcare, 2015). In 2011, an estimated $205 billion, or 13%, of total health spending was spent in the last year of life (Aldridge and Kelley, 2015). The Dartmouth Atlas of Healthcare (2015) attributed much of the cost burden to repeat hospitalizations and drew a connection to the incongruence of patients’ stated preferences (a majority wish to die at home) and actual outcomes (55% die in a hospital). Politically, EOL discussions between providers and patients were set back when potential Medicare reimbursement under the proposed Affordable Care Act was characterized by prominent politicians as “death panels” and the provision was removed from the eventual act (Millman, 2014). However, as of January 1, 2016, in part due to stakeholder influence including the American Medical Association, the Centers for Medicare & Medicaid Services (CMS) began reimbursing for advance-care planning discussions (CMS.gov, 2015). While individual providers are exempt from the PSDA, primary care providers (PCPs) are key loci for introducing ADs. The American Academy of Family Physicians (AAFP) recommends the patient’s PCP broach the topic early, such as when discussing preventative care at a routine check-up (Spoelhof & Elliott, 2012). Homeless individuals, however, typically lack a PCP (White & Newman, 2015). Like many Americans, they are often not given the opportunity to fill out an AD until they experience a health crisis. Other barriers originate from the provider side: Nurses, for example, often feel unprepared or resistant to educating patients about ADs (Goodwin, Kiehl & Peterson, 2002). Because SMI is more common among homeless people, providers may be uncertain about whether the patient has decision-making capacity (Massachusetts Department of Mental Health, n.d.). ADVANCE DIRECTIVE ADVOCACY 8 Other factors influence who has an AD. An inverse relationship has been noted between completing an AD and lower levels of education, socioeconomic status, insurance status, and minority status (Song et al., 2008; NHPCO, 2013). Nursing home residents and discharged hospice patients are more likely to have an AD than the general population: 65% and 88%, respectively (Jones, Moss, & Harris-Kojetin, 2011). The older the patient, the more likely he or she is to have one; for example, 77% of nursing home residents over the age of 85 reported an AD, compared to 36% under the age of 65. Also, in general, community-dwelling residents over 65 years old are more likely to have an AD than all community-dwelling adults, 37% vs. 15% (Jones, Moss, & Harris-Kojetin, 2011). A medical respite center (MRC) in Maricopa County is currently seeking a policy and tool for AD completion. Other than asking new admissions for code status, the center acknowledges it does not have a means of introducing and addressing ADs. A recently opened sister clinic to provide continued healthcare to discharged respite patients also presents opportunities to include AD discussions as part of comprehensive care. Similarly, a Maricopa County day resource center (DRC) for homeless individuals 55 years or older expressed interest in providing information on the topic. The director cited a recent incident when a patient suffered a heart attack and died on the premises as an example of the fragile health many clients exhibit. Social work, pastoral counseling, and nursing care is provided at the center, allowing varied opportunities to introduce the topic of ADs to clients. Problem Statement Elderly and chronically ill homeless stand to benefit significantly from completing an AD and have been shown amenable to doing so, but barriers to completion are significant. To explore how best to do so led to the PICOT question: ADVANCE DIRECTIVE ADVOCACY 9 In homeless patients (P), how does an educational session explaining the purposes of an AD (I) compared to no education (C) affect rates of completion (O) within eight weeks of the intervention (T)? Search Strategy An exhaustive search of the literature was conducted, comprising four databases: CINAHL, PsychInfo, JSTOR, and PubMed, as well as the National Guidelines Clearing House and the Cochrane Library. In addition, grey literature was searched. The search terms were: “homeless,” “advance directive(s),” “mentally ill,” “schizophrenia,” “addiction,” “poverty,” “low-income,” and “advance care planning,” alone and in combination. Yields generally gleaned less than 10 studies, with the exception of “advance directives” alone, which resulted in over 4,400 results in CINAHL. When that search was limited to results available in full text, English, post-2009 publication, and randomized control trials (RCTs) with adults only, studies were winnowed to two. In addition, ancestry searches of relevant articles harvested several additional studies. Of the four databases combed, ten studies were selected for inclusion, addressing each element of the PICOT (Appendix A). The studies comprise two systematic reviews, three RCTs, two cohort studies, and three cross-sectional studies. All studies were in English and published between 2008 and 2015. Appraisal and Conclusions Each study appears to be of reasonable quality and potentially contributory towards increasing AD completion among homeless patients in Arizona. Two are of Level I evidence, three of Level II, and five of Level III. Internal validity was based on generally few confounding variables, with the exception of race, despite the complexity of homeless culture and the myriad ADVANCE DIRECTIVE ADVOCACY 10 demographic factors that could or have been shown to influence AD completion (Appendix B). The populations studied supported external validity: in half of the studies, they were homeless or recently homeless. In the majority, participants were ethnically mixed. Instrument reliability was consistent, primarily measuring nominal demographic data and ordinal data such as Likert scale responses (Appendix A). Likert scales were employed by several studies to measure attitudes. No bias was noted across the studies (Appendix A). Independent and dependent variables were generally homogenous across the studies. One exception was the Payne, Prentice-Dunn & Allen study (2010), whose hypothesis was not supported by the evidence. The study was included because it shed light on how the method of delivering the educational content can affect participant response. Another was the Ganzini, Socherman, Duckart, and Shores (2010) study, which found veterans with schizophrenia and cancer received comparable or better EOL care than their counterparts with cancer but no schizophrenia. It was included because it addressed mental illness’s prominence in homeless culture. However, in terms of resources, veterans have a well-placed infrastructure for medical care and, more recently, for housing (Arizona Veterans Standdown Alliance [AVSA], n.d.). Therefore, it is reasonable to assume they will not be as prominent as non-veterans in the MRC’s or DRC’s patient base. Independent variables primarily comprised demographic characteristics, and the quality of the educational intervention. As the main dependent variable, completion of ADs was straightforward. Outcomes were also homogeneous – AD completion or willingness to do so, which allowed for synthesis. Conclusions/Discussion Generally, while homeless older adults recognize the importance of EOL planning, they do not regard it as a priority, in part because they are focused on fulfilling the needs of their ADVANCE DIRECTIVE ADVOCACY 11 immediate future (Ko & Nelson-Becker, 2014). However, when offered the opportunity to complete an AD, they do so at the same or higher rates as the general population. The number of homeless individuals who opt to complete one ranges from 27% to 59%. (Song et al., 2008; Song et al., 2010; Leung, Nayyar, Sachdeva, Song, and Hwang, 2015). One study speculated that homeless persons are more responsive than their domiciled counterparts to interventions aimed at promoting ADs, possibly because those who are both chronically ill and homeless are doubly motivated to make known their EOL wishes (Leung et al., 2015). Low-income elderly patients, many of whom were recently homeless, are more willing to complete an AD when they rate their health as “poor” or “fair” (versus “good” or “better” health) and if they have an available healthcare proxy (Ko, Lee & Hong, 2015). Conclusions about ethnicity and AD completion were difficult to draw. Whites were more likely to have an AD: 28%, compared to 21.4% of African-Americans and 10% of Hispanics (Ko & Lee, 2014). However, when knowledge was controlled for, race/ethnicity became nonsignificant. Two later studies, including one by the same lead researchers, found no significance in race/ethnicity among its White, Hispanic and Black participants (Ko, Lee & Hong, 2015; Leung et al., 2015). Other factors noted to result in higher incidence of AD completion were prior ICU admission and higher income levels, as well as an association between filling out an AD and knowing what kind of EOL care the person desired but had not informed anyone of (Ko & Lee, 2013, Leung et al., 2015). A counselor-guided intervention was more effective than a self-guided intervention at a rate of 37.9% compared to 12.8% (Song et al., 2010). The former intervention consisted of an appointment with a hospice social worker offered at the same location as initial recruitment. ADVANCE DIRECTIVE ADVOCACY 12 Counselors were specially trained in the EOL concerns specific to homeless persons. Each encounter averaged 45 minutes. An earlier, smaller study by the same lead researchers noted results of 59% AD completion in the counselor-guided group compared to 30% in the selfguided group (Song et al., 2008). The content of the educational intervention may influence outcomes. Payne, PrenticeDunn and Allen (2010) posited that a 90-minute presentation that explicitly outlined the threats of not having an AD on EOL care would result in more completed ADs compared to a 90-minute presentation that discussed healthy aging but did not mention ADs. However, the opposite was the case. The authors attributed this unexpected result to the possibility that overwhelmingly threatening information may cause recipients to deny the content of the message and retreat into “fatalistic thinking” (p. 57). Song et al.’s 2008 study used an AD adapted to the homeless population called HELP (Honoring End of Life Preferences) and included an educational session, either written or in person, about how to fill it out. Clients who completed it were more likely to report carrying contact information in case of emergency and a decrease in worry about receiving appropriate care if sick or dying (Song et al., 2008). They also reported a decrease in pessimism about the worth of ADs, which the authors interpreted as an enhancement of clients’ autonomy in a healthcare setting. Participants who completed an AD said their doing so made them feel at peace (75%) and think more about their health (65%), according to one study (Leung et al., 2015). Initiatives tailored to a specific culture have proved successful. A 6-month pilot test to reduce disparities among completion of living wills and MPOAs for healthcare among African Americans and black immigrants in Minnesota improved results when the intervention was ADVANCE DIRECTIVE ADVOCACY 13 adapted to the target population based upon their feedback. For example, rather than requiring only one individual be named as MPOA, the care coordinators allowed clients to name multiple family members (Agency for Healthcare Research and Quality, 2013). At the end of the pilot test, AD completion rate improved to 32% from 24.8% and narrowed the completion-rate gap between Blacks and Whites to 19.9% from 25.7%. Available studies drew differing conclusions about whom homeless people chose as surrogate decision-makers. Some favored unrelated contacts such as service providers, friends, and romantic partners (Song et al., 2006). Many expressed a preference for their physicians over family members, because the former had more expertise about the patient’s situation (Ko & Nelson-Becker, 2014). However, a later and larger study found that 52% of homeless participants who opted to fill out an AD named a sibling as a surrogate decision-maker, 32% a child, and 22% a parent (Leung et al., 2015). When homeless patients were asked if they would want CPR if they were in a permanent coma, over one-third said that they would; non-white participants were less likely than whites to specify no CPR (Leung et al., 2015; Song et al., 2010). Though it did not provide comparative numbers with non-homeless, one study concluded that homeless persons are more likely to want life-sustaining interventions such as CPR (Leung et al., 2015). The homeless culture is a complex one whose elements cannot be easily parsed. Ethnicity, religiosity, comfort with discussing death, and past experiences with healthcare are among the many factors that influence a homeless person’s EOL wishes, just as they are for a non-homeless individual. A study of EOL attitudes among older low-income residents of supportive housing, one-third of whom had been homeless, found that greater religiosity resulted in higher likelihood of desiring treatment at end of life compared to less religious counterparts ADVANCE DIRECTIVE ADVOCACY 14 (Ko & Lee, 2014). Likewise, residents who were less comfortable discussing death were more desirous of aggressive treatment at end of life. Alternatively, the Ko & Lee (2014) study found no difference in EOL preferences based on ethnicity. Although the AAFP recommends the primary care setting for EOL discussions (Spoelhof and Elliott, 2012), homeless patients are less likely to access regular non-emergent care than domiciled patients. As an alternative, the NHCHC’s Respite Care Providers Network (RCPN) recommends medical respite centers as ideal places to introduce ADs. Such centers often have social workers who can assist patients in identifying appropriate individuals, including help locating relatives, to be surrogate decision-makers, and to assist the patient in notarizing and filing the documents (NHCHC, 2011). Most studies did not specify the type of AD used, with a few notable exceptions. Song et al. (2008) utilized an AD called HELP (Honoring End of Life Preferences), designed specifically for marginalized groups and adapted for the purposes of the study to the homeless. A systematic review of 16 studies about the effectiveness of educational interventions on AD completion noted the Physician Orders for Life Sustaining Treatment (POLST) as a successful EOL plan that is an adjunct to an AD (Durbin, Fish, Bachman & Smith, 2010). It is for patients with serious illness to delineate their wishes for current treatment. Its video component is useful for patients with limited literacy (POLST, 2015). The review also referenced Five Wishes, which is reasonably priced and is also available in Spanish (Aging With Dignity, n.d.). Leung et al. (2015) used “My Living Will, An Ontario Advance Directive,” adapted from the Study of End-of-Life Preferences among Homeless People (SELPH) Advance Directive (2015). Both HELP and SELPH are based on Chochinov’s dignity-conserving care model (Song et al., 2008; Leung et al., 2015). Finally, the NHCHC (2011) includes in its clinical recommendations for respite care a ADVANCE DIRECTIVE ADVOCACY 15 living will that appears designed for its population, with a section listing features for identification of the signee’s body (such as tattoos or scars) and a section called “Maintaining My Dignity,” which includes statements such as “I would want to be remembered as a person who:” (p. 47) and “These are my concerns about a relationship I have:” (p. 48). Potential Application to Practice ADs are meaningful tools for delivering desired care at end of life. Medical providers such as nurse practitioners (NPs), as well as non-medical providers, such as social workers, can increase the likelihood of homeless patients completing one by educating such patients on their purpose and offering the opportunity to fill one out at each encounter. Several studies have found that a simple one-time intervention – an educational session on the purpose and meaning of an AD – has been effective. It is something that can be done in a nonhospital setting by a variety of staff members who have received training on the subject. NPs can effect a practice change by training team members to deliver the education to clients. On the recipient end, studies show that homeless patients are amenable to filling out ADs. Establishing trust is imperative to overcoming the experience-based biases this population may have against healthcare providers. Providers of care to homeless clients must examine their own biases as well and approach homeless patients with suspended judgment. At least one study demonstrated that, once trust is established, homeless individuals expressed a preference for physicians, rather than family members, to make EOL decisions for them (Leung et al., 2015). This is another impetus for providers who have regular contact with homeless patients to introduce the topic early in their relationship so wishes can be formalized by the time they are needed. The Affordable Care Act has increased access of previously uninsured Americans to healthcare, primarily through the expansion of Medicaid, in which the state of 16 ADVANCE DIRECTIVE ADVOCACY Arizona is participating. Homeless individuals now have the potential to begin receiving primary care before their medical issues evolve to late stages requiring hospitalization. The practical aspect of accessing ADs when they are needed is a hurdle that remains largely unaddressed in current research. One study, however, noted that all participants agreed to let the authors file their completed ADs at the two hospitals in its region most likely to provide care to homeless patients: a VA hospital and a county hospital (Song et al., 2010). This represents a potential means of storing and accessing completed ADs. While some studies included specific AD examples adapted to the homeless population, none made explicit recommendations about which AD was most effective, by what means the ADs were adapted to the homeless, and what in particular (i.e., literacy, health literacy, language barriers) needed to be adapted. This too remains an area for further study. Summary Homeless individuals face significant barriers in accessing EOL care. Providing such care to this population will only gain in importance as homeless people age and are diagnosed with the same diseases as the general population, but without the same level of social and economic support. ADs have been shown to be an effective way of conveying EOL wishes when the patient is unable to express them. Among the homeless, ADs may also serve to empower a population saddled disproportionately with both greater disease burden and less access to healthcare. ADVANCE DIRECTIVE ADVOCACY 17 Applied Clinical Project Methods and Results The applied clinical project to increase AD completion among homeless clients was prepared and delivered to staff and clients of the two participating agencies within a two-week period, in accordance with approved ethical guidelines. The two agencies consisted of a medical respite center (MRC) and a day resource center (DRC) for individuals 55 years or older; both agencies exclusively serve people experiencing homelessness. Project execution of the evidencebased intervention varied somewhat according to each site’s environmental and organizational culture, but comprised essentially the same content. EBP Model and Theoretical Model Rosswurm and Larrabee’s (1999) evidenced-based model (Appendix C) is applicable to a primary-care setting and has been used to effectively establish practice changes that result in measurable outcomes. Step 1, assess the need for a change in practice, was accomplished by the MRC, which cited the need for a tool and policy for both the existing center and the planned outpatient primary-care clinic. The DRC likewise expressed a need to formalize clients’ EOL wishes given their age and relative health status. Step 2, linking the problems, interventions and outcomes, involved identifying educational sessions as a viable means of encouraging ADcompletion among this population. Step 3, synthesizing the best evidence, included an exhaustive search of the literature, the identification of studies comparing the effects of an educational session to usual care, and the determination that the former is feasible, beneficial and minimally risky. Step 4, designing a practice change, is in progress, as each site has begun to informally incorporate AD education into its respective client interactions. Step 5, implementing and evaluating the change, began with conducting the educational intervention separately for staff and clients at each site, a total of four presentations. Pre- and post-surveys were completed 18 ADVANCE DIRECTIVE ADVOCACY to measure changes in attitude and knowledge about ADs, and completed ADs eight weeks postintervention were tallied. Step 6, integrating and maintaining the change, is in progress. It will likely involve working with major stakeholders to formally adapt the practice of educating clients for institutional use, including, in the case of the MRC, creation of a printed policy. King’s Theory of Goal Attainment (King, 2007) provided an appropriate foundation on which an educational intervention to encourage AD completion among homeless patients could be built. The Theory of Goal Attainment is drawn from King’s (2007) conceptual framework of individuals existing and interacting amid three strata: personal, interpersonal, and social systems (Appendix D). Within this framework, the nurse-client relationship is key. “[Individuals] have a right to accept or reject health care. … Health professionals have a responsibility to share information that helps individuals make informed decisions about their health” (King, 2007, p. 109). Perception is a mutual concept drawn from the personal realm: The nurse must perceive when health education is appropriate for the client, while the client must perceive whether or not to make a decision (Goodwin et al., 2002). Interaction, a concept from the interpersonal stratum, is reflected in the nurse-patient education process early on and, later, in the nurse’s advocacy of the EOL wishes the patient has expressed (Goodwin et al., 2002). From the social stratum, power is reflected in that which the newly informed patient exercises when he or she decides whether to complete an AD and, if so, what is contained therein (Goodwin et al., 2002). On a macro level, King’s theory as it applies to AD completion rests on the ethical linchpin of patient selfdetermination. Project Methods Ethics ADVANCE DIRECTIVE ADVOCACY 19 Arizona State University’s Institutional Review Board approval was sought and granted (Appendix E). A consent was provided both to staff and to clients, describing the project, its purpose, and what would be asked of those who chose to participate (Appendix F). Pre- and postsurveys were anonymous; participants provided a unique identification number to protect their identity. Risk to participants was defined as potential discomfort with the topic of EOL decisionmaking and status of relationships when asked to consider potential MPOAs. Participants were excluded if they did not speak or read English. Snacks were provided, but no financial remuneration was provided for participation. Setting The intervention was performed at two sites, the MRC and the DRC. Both are located in the state’s urban capital within four miles of each other; project participants could potentially qualify for services at both sites. While by definition the MRC’s clients have recently suffered a health crisis, clients of the DRC also experience frequent health needs, necessitating the presence of a nurse with a dedicated exam room who is available during week days. The DRC is open daily during daytime hours. It features a common room of tables and chairs where clients gather and converse or use the communal computers. Lunch is served daily. Clients are free to come and go at will. By contrast, the MRC is a more structured environment. Its primary purpose is healing and so residents are expected to visit the on-site medical clinic at appointed times for daily medications, wound care, or to see the physician or physician assistant. Ancillary services are also provided, such as those designed to find clients secure housing upon discharge. Clients live on the premises, in gender-segregated dorms, with two private rooms reserved for isolation or hospice patients. All meals are provided in a communal dining area, and elective activities, such ADVANCE DIRECTIVE ADVOCACY 20 as art classes, are offered, as are occasional field trips for clients healthy enough to make them. A chapel and common room with a television and several computers are also on site, as well as a shaded smoking patio in the parking lot, which is a popular gathering place. Clients must earn the privilege of leaving the facility, such as to walk next door to the convenience store. The nurses’ station is located on the same floor as patient dorms and medical clinic; glassed in, it is referred to as “the bubble.” Management’s and case workers’ offices are located on the ground floor. Staff-only sites are generally accessible only by badge. A security guard is always present on the campus. Organizational Culture Both sites’ leadership expressed enthusiasm for the evidence-based project and both had key individuals who articulated a need for it. The organizational culture of each site differed in significant ways, however. The DRC is the smaller of the two, with a staff of approximately 10 individuals. It had experienced a leadership change within the year, and staff were acclimating to a new executive director whose background was in fundraising, in contrast to the prior director, whose background was church-based. Furthermore, in the weeks prior to and during the execution of the intervention, the staff had experienced some emotional disequilibrium, including the terminal medical diagnosis of one employee and the expected departure of others in response to leadership changes. Physically, staff frequently encounter clients, who occupy a great room around which staff offices are clustered. When not engaged in private client consultations, the pastor and the case managers frequently leave their doors open, facilitating access. The executive staff occupies a front office with a view of the great room. The MRC, by contrast, is a larger organization, occupying a former two-story office complex. Staff and client areas are kept separated by various means, including locking the door ADVANCE DIRECTIVE ADVOCACY 21 of the medical clinic during posted closing times and badge entry required to access administrative areas. Clients also require permission to leave the premises. As a result, clientstaff interaction is more structured, facilitated by appointment more so than chance encounters. Also, in contrast to most of the DRC, the MRC’s staff either wears uniforms (scrubs, lab coats) or business-casual dress, making them visually distinctive from clients. Participants Intervention participants comprised staff and selected clients at each facility. At the DRC, the intervention was presented at a weekly staff meeting, which included the executive director, two case workers, and the pastor (who is also the housing coordinator). The presentation was followed by one in the center’s great room for the clients who happened to be gathered there, a circumstance arranged primarily by a key stakeholder who had helped to coordinate the project at the site. Although the event was advertised approximately one week prior with a posted flyer, any participant who may have been deliberately present for the session was not identified. The MRC presentation was conducted at the monthly nurses’ meeting. As with the DRC, the staff presentation was followed by one for clients, this one in an unoccupied community room. Unlike the DRC, however, participants were pre-screened to gauge interest in attending by a key stakeholder who helped coordinate the project at the site. Intervention The educational intervention, an approximately 15-minute explanation of ADs and their applicability to the needs of homeless clients, was performed separately for staff and clients at each site, a total of four presentations. Each presentation proceeded similarly, with the distribution of photocopied releases (one for staff, one for clients) and pre-survey (the same for both staff and clients). The presentation was delivered using an easel and large note pad. Upon ADVANCE DIRECTIVE ADVOCACY 22 conclusion of the presentation, the post-survey was distributed. The post-survey contained the same seven questions as the pre-survey, minus the demographic information. Outcome measures The pre- and post-survey was drawn from Murphy et al.’s (1996) study, “Ethnicity and advance care directives,” which developed an instrument titled “Ethnicity and attitudes toward advance care directives questionnaire.” Questions measuring knowledge of, attitude toward, and possession of an advance directive were excerpted for the intervention’s pre- and post-survey (Appendix G). They totaled seven questions, plus four demographic questions (gender, age, race, and marital status). The knowledge questions were definition-based: “What is a living will?” and “What is a medical power of attorney?” Responses were scored on a nominal scale: either correct or incorrect. The attitude questions’ responses were measured on a four-point Likert scale (i.e., 1 = agree strongly, to 4 = disagree strongly). For the purposes of the intervention, the wording of one question was adapted from “Doctors should …” to “Medical providers such as doctors, nurse practitioners, and physician assistants should ….” The questionnaire also asked if the respondent was already in possession of an AD. The Murphy et al. (1996) study noted the scale was derived from previously validated scales as well as new scales, which underwent testing for internal and external validity, including extensive pilot testing. Lead author Sheila T. Murphy did not respond to a personal email communication about specific coefficients relating to reliability and validity. Data collection and analysis plan Data was collected on site at the conclusion of each presentation. Eight client pre-surveys were distributed at the MRC, and eight post-surveys were collected, plus a ninth following a oneon-one presentation to a client who had later expressed interest. Six pre-surveys were distributed to MRC staff and six post-surveys collected. At the DRC, 31 pre-surveys were distributed to ADVANCE DIRECTIVE ADVOCACY 23 clients and 19 returned. At the DRC, seven pre-surveys were distributed to staff and six postsurveys collected. Data was first entered into an Excel spreadsheet to preserve it and to provide a basic idea of demographics and survey responses. The Excel spreadsheet was subsequently imported into SPSS 22, where data were analyzed for errors and for case summaries. Thirteen cases were thrown out because they could not be matched, leaving 40 matched pairs: 24 from the DRC and 16 from the MRC. A McNemar’s test was used to analyze improvements in knowledge about a living will and medical power of attorney. A Wilcoxon signed-rank test was used to analyze changes in attitudes about advance directives. Proposed budget Cost outlays for the intervention itself totaled approximately $200 for office supplies, snacks and photocopies (Appendix H). This included compiling and binding booklets containing examples of collected ADs for focus groups from each agency to peruse and rate. An additional $120 was spent by the project’s author on an online training course in facilitating AD discussions in order to better inform the presentation. The agencies’ budget for sustaining the delivery of AD facilitation is virtually free for materials. The directives themselves are downloadable at no cost from various World Wide Web sites, as are wallet cards and access to sites that provide supplementary information for staff and clients who seek it. Photocopying downloaded documents and completed documents for inclusion in patient files incur a minor cost, though one that is unlikely to be prohibitive, even given each agencies’ presumed frugal budgets. Staff time required to deliver education to current and future clients is less easily monetized, though perhaps more significant. ADVANCE DIRECTIVE ADVOCACY 24 Project Results A total of 92 surveys were collected: 51 pre-surveys were distributed and 41 were returned. A total of 14 staff members were surveyed, and 37 clients. Twenty-seven respondents were female; the remaining 23 were male. Average age was 57.5 years (SD 10.0). Thirty-two, or 65.3%, of respondents were Caucasian; 10, or 20.4%, were Hispanic; the remainder, or 14.3%, were African-American (Appendix I). Staff members were divided equally between agencies (seven participants at each site). Women comprised slightly over two-thirds of surveyed staff. All staff were Caucasian with the exception of one African-American member. Average age was 47.8 years (SD 13.1). More participants were married versus single. Approximately two-thirds stated they already had ADs. Among clients, nearly three-quarters, 72.3%, were DRC clients. Respondents were nearly equally divided in terms of gender (19 males versus 18 females). Average age was 60.9 years (SD 5.7). Over half, 56%, were Caucasian, with Hispanics comprising 28% and AfricanAmericans 17%. The vast majority, 89%, were unmarried. Likewise, most – 85% -- did not currently have an AD. A McNemar’s test determined that the intervention significantly improved knowledge of a living will (p = .013, N = 40), but not of the MPOA. A Wilcoxon signed-rank test revealed no statistically significant improvement in attitudes about AD completion following the educational intervention. The result was the same when run for clients only, as well as for total participants (clients and staff) (Appendix J). While attitudes did not change significantly, they did more positively favor ADs in two of the four areas explored post-intervention (Appendix K). Following the intervention, increased disagreement registered with the statements “It is best to wait until the situation arises to make decisions about 25 ADVANCE DIRECTIVE ADVOCACY treating very serious illness, rather than writing them down in advance,” z = -.666, p = .506, and “It is not necessary for people to write down their wishes about medical care because their family will know what to do when the time comes,” z = -1.137, p = .256. Two clients, both at the MRC, completed documents. One of the two completed an AD at the conclusion of the presentation. A second client, who received a one-on-one educational session later that same day, completed an MPOA eight weeks later. Discussion The literature indicates that, when given the opportunity, homeless individuals are as likely or more likely than the general population to complete an AD (Song et al., 2008; Song et al., 2010; Leung et al., 2015). Two of nine clients, 22%, completed the document at the MRC; none did at the DRC. Results thus fell short of 26%, the percent of the general population with ADs (NHPCO, 2013). Implicit in the studies is the idea that informing homeless individuals about the purpose of ADs increases their knowledge and reshapes attitudes, thus presumably leading to completion of the document. In this regard, the intervention was partially successful: It significantly increased knowledge about one of the two documents comprising an AD and insignificantly improved attitudes. Project strengths include the relative simplicity and inexpensiveness of providing the intervention. Feedback from staff at both agencies was positive about the content and delivery of the intervention, which lays the groundwork for sustainability by staff in delivering future interventions to clients. Limitations include the fact that client participation, while voluntary, was not necessarily motivated by interest at the DRC, where the presentation was given to those who happened to be present at the time. Comprehension of the survey may have been limited by poor eyesight, poor 26 ADVANCE DIRECTIVE ADVOCACY literacy, limited comprehension of English, and physical or mental infirmities. The sample size, 40, was small. It was observed that African-American clients were less willing to participate, and so participation may not accurately reflect the racial makeup of total clients. Furthermore, while a strength of the proposed intervention is its applicability beyond strictly medical staff, only nurses received the education at the MRC, potentially limiting the types of staff capable and willing to sustain further education. Conclusion A simple and inexpensive educational intervention resulted in a significant knowledge increase among staff and clients of two agencies serving the homeless, as well as the completion of two ADs by clients. The intervention’s basis in King’s (2007) Theory of Goal Attainment guided the interaction between presenter and clients and provided the foundation for future interaction between agency staff and clients in a manner adapted to the social context in which homeless older adults and those who are chronically and/or acutely ill dwell. Based upon minimally significant results, the intervention requires further finessing in order to be truly effective in the goal of expressing and therefore facilitating EOL wishes by a marginalized segment of the population. ADVANCE DIRECTIVE ADVOCACY 27 Organizational/Health Policy Impact and Sustainability The project to increase AD completion by homeless clients of two agencies achieved modest statistical success and holds promise for clinical sustainability, though certain challenges must be overcome. The circumstance of performing the project at two sites, which share broad similarities but differ in key ways, allowed insight into factors that may help or hinder impact and sustainability when attempting an educational intervention to shape knowledge and attitudes about ADs with the intention of increasing their completion. Impact of project at practice sites The project resulted in one statistically significant result, the improvement of knowledge of a living will (p = .013, N = 40). Two of nine clients (22%) at the MRC completed an AD. The intervention for both nursing staff and clients resulted in the later completion of one of those documents. “Thanks to [the intervention], we had them [the directives] on hand and the nurse could answer the patient’s initial questions,” the nurse manager stated (S. Bone, personal communication, January 13, 2016). A formal policy is being developed to make the introduction and discussion of the documents a consistent part of the new-patient intake process and subsequent encounters. The policy draft is tentatively scheduled to be completed by May 2016 for stakeholder review. At the DRC, no documents were completed. Clinically, discussion of the project (before its actual execution) had an immediate influence on institutional process, with intake staff preemptively asking new clients about ADs and referring them to social workers to complete, if appropriate. By that method, the DRC reported the completion of approximately 10 to 12 documents (O. Munoz, personal communication, November 23, 2015). However, this informal process appeared to lapse over time, in part due to staff changes, discussed in more detail below. ADVANCE DIRECTIVE ADVOCACY 28 Financial implications of project Cost outlays for the intervention itself totaled approximately $200, with an additional $120 spend by the project’s author on an optional online training module in order to better inform the presentation. Cost to the agencies of incorporating AD education into their policies is minimal, as the documents, wallet cards, and supplementary information are all available free of charge via the Internet. Staff time required to deliver education to current and future clients, however, incurs a cost in terms of lost opportunities to perform other client-related tasks. As the DRC situation illustrates, advance-care planning has less priority than other tasks: Staff losses have required remaining staff to take on more duties, with a result that AD discussions with clients have been tabled. One reason is certainly that the documents are prepared for use at an unknown future time, while more pressing matters, such as putting clients in stable housing, demand immediate attention. Another reason may be that the financial benefit to implementing the project is not remunerative to the participating agencies; any financial boon is likely to be seen diffusely, in the reduction of healthcare costs associated with EOL care. However, given the minimal cost outlays, the cost-to-benefit ratio would not be expected to be prohibitively negative. Impact of current policy As of January 1, 2016, the Centers for Medicare and Medicaid Services (CMS) began reimbursing providers for advance-care planning discussions with patients. It is too early to tell whether such discussions have increased as a result, but some experts expect that they will (AHC Media, 2016). The American Academy of Family Physicians (AAFP), for example, officially encourages its members to discuss advance-care planning in a non-emergent setting (AAFP, 2015). Because many previously uninsured individuals in Arizona now qualify for Medicaid ADVANCE DIRECTIVE ADVOCACY 29 coverage since its expansion in 2014 as part of the implementation of the Affordable Care Act, more clients can be expected to seek non-emergent medical care, providing more opportunities for advance-care planning discussions, including the completion of ADs. Leader and innovator role Both agencies were approached about performing an educational intervention for staff and clients on the subject of ADs. Both responded with enthusiasm. The MRC additionally requested that the intervention culminate in a written policy for addressing the subject with clients. Barriers to implementation included arranging time with staff; agency upheaval; and variable client interest. Arranging staff time was most challenging at the MRC. While the nursing staff was readily available and allotted time at a monthly meeting for the intervention, scheduling time with other disciplines proved more difficult. Because advance-care planning can be done by non-medical staff – indeed, that is one of its strengths – it was important that social workers, the intake coordinator, and the volunteer coordinator receive the education and be included in the drafting of the policy. Perhaps due to the agency’s small size and high patient census at the time of the intervention, scheduling time with the intake and volunteer coordinators took several weeks of attempts via email and text. Because stakeholder buy-in is key to sustaining any eventual policy (Porter-O’Grady & Malloch, 2015), assiduous continued efforts to schedule time with the remaining employees will be attempted, even if it means delaying the drafting of a policy past the self-determined deadline of May 2016. As mentioned above, the DRC experienced significant staff changes, beginning with new leadership and continuing with the loss, both expected and unexpected, of key staff, including the intake coordinator and the chaplain, the latter of whom served a dual role as the housing ADVANCE DIRECTIVE ADVOCACY 30 coordinator. Regular updates will continue to be sought from remaining key stakeholders in order to assess a time that is more opportune for revisiting formal implementation of a sustainable process and policy. One measure of successful project implementation was the number of directives completed by clients, post-intervention. Clients did so solely at the MRC, where the client intervention group was self-selected. (A nurse there also changed her MPOA as a result of the intervention [personal communication, S. Bone, December 7, 2015].) One of the lessons learned from the opportunity to present at two sites is that the self-selected audience was more amenable to receiving information about advance-care planning than the convenience sample, which was the case at the DRC, where no directives were completed as a result of the intervention. This knowledge will inform future project implementation at other sites. Sustainability plan At the DRC, an initial push to discuss ADs with new clients and route those interested to social workers for completion lapsed over time. The organization, whose employees originally numbered approximately 10 at the time of the intervention subsequently lost at least four staff members, including one to death. Those remaining have had to temporarily take on additional duties, challenging their ability to discuss ADs with clients (personal communication, O. Munoz, February 8, 2016). Although remaining staff have expressed a commitment to restoring regular discussion of ADs with clients, there is currently no formal process in place to do so. As stated above, key stakeholders will be queried intermittently to identify a time and willingness to attempt establishment of a consistent policy for AD discussion. At the MRC, a formal policy is in development that will likely entail introducing the topic with the initial intake interview and revisiting it in successive client-staff encounters, such ADVANCE DIRECTIVE ADVOCACY 31 as medical and social-work appointments. The agency recently converted its electronic medical record (EMR) to the Athena system, whose charting is configured to ask new clients about the presence of a directive. At a new client interview, the intake coordinator was observed asking the client whether she had such a document. Later, the intake coordinator stated that he formerly did not routinely ask when using the agency’s old EMR system, which was not similarly designed. So structurally, the agency now has technology in place that is conducive to sustainability. As noted above, nursing staff have begun to implement the as-yet-determined policy informally, while stakeholder buy-in from other key agency employees will continue to be sought. Implications for further application, study and research The presentation has since been delivered to interdisciplinary staff at a third agency that provides medical services to homeless patients. The clinic’s medical director had requested the presentation with the stated intention of instituting a more consistent approach to advance-care planning. A potential future site for implementation is the interscholastic, interprofessional clinic that operates at the same facility, but on Saturdays, and with different providers. A 90-minute variation of the intervention was also presented prior to formal implementation at a state conference about homelessness. Nationally, conferences directed at healthcare providers and the underserved are opportunities to present the project. Examples include the annual National Healthcare for the Homeless Conference; the Southwest Regional Nurse Practitioner Symposium; the American Nursing Association annual conference; and the National Hospice and Palliative Care Association conference. The somewhat lackluster results of the intervention’s initial iteration suggest that further honing is in order, with the goal of determining more effective means of increasing AD ADVANCE DIRECTIVE ADVOCACY 32 completion via such methods as improving knowledge of and attitudes about the documents. Developing a pre- and post-survey instrument specific to the population is an area for further research. Collaborators potentially include the two agencies that participated in the project implementation, as well as the interscholastic clinic, which has a notable research emphasis. Gaps identified during project No instrument measuring homeless clients’ attitudes and knowledge of ADs was identified during the initial and subsequent literature searches. The instrument used in the intervention was adapted from a tool derived for ethnic minorities. The culture specific to the homeless population – for example, a higher likelihood of an SMI diagnosis and/or substance abuse – may require a tool that takes such conditions into account. Additionally, the population targeted by the intervention was older, which may have meant the possibility of sensory deficits such as impaired eyesight, that could have negatively affected comprehension of the instrument. The results of the applied project did not reflect those promised by the literature. One potential factor has been described above: existing client interest versus a convenience sample. Other potential gaps include language limitations and staff resources. The intervention and preand post-survey were conducted in English, which automatically excluded participants for whom English is not their primary language. According to the literature, a one-on-one counseling session is more successful than a self-directed one (Leung et al., 2015; Song et al, 2010). Other than the single one-on-one session with the MRC client described above, agencies may not have the resources to dedicate to similar encounters. Finally, the project does not include homeless individuals who do not receive social services and who therefore do not have ties to an agency whose staff could provide the education ADVANCE DIRECTIVE ADVOCACY 33 and opportunity to complete an AD. To truly address health equity, all homeless individuals should be given the opportunity to complete a directive. Conclusion: Project summary and significance Executed within the structure described by Rosswurm and Larrabee (1999), the applied clinical project resulted in an increase in knowledge of a living will; the completion of ADs by two clients and one staff member as a direct result, and as many as 12 clients, as an indirect result; and the development of a formal policy at one of the two participating agencies. The results, while concrete, suggest the intervention can be improved. Two potential ways identified from both the literature and the project are assessing client interest before delivering the education and one-on-one counseling, if feasible. King’s Theory of Goal Attainment (Goodwin et al., 2002) supports the former method with its tenet that the nurse must perceive that the client is ready for the education and subsequent decision-making. Institutional culture will influence the project’s sustainability. Within the more structured environment of the MRC, the development of a formal policy promises more consistent adherence; while at the DRC, where staff upheaval has challenged new practices from taking root, maintaining the commitment of key stakeholders until the turbulence resolves may be a more effective strategy. Academically, the design of an instrument measuring attitudes and knowledge of ADs that is specific to the homeless culture may elicit more useful data and remains an area ripe for further research. The aging of the homeless population, expanded Medicare coverage for Arizona residents, and CMS reimbursement for advance-planning discussions create conditions favorable to increased completion by homeless patients of ADs. However, the small size and budget of ADVANCE DIRECTIVE ADVOCACY 34 many agencies that cater to the homeless mean such agencies tend to be buffeted more by destabilizing forces including staff changes, which is a challenge for sustainability. In summary, educating homeless individuals and the social-services staff who interact with them is a proven way to address health inequities in EOL care. The intervention must be delivered, however, in a way that is tailored to accommodate the exigencies of the population, as well as of the agencies that serve them. There are compelling reasons such strategies should be honed: Completing an AD empowers a medically and socially vulnerable section of the population to voice their preferences, assuage their fears, and reduce expensive and undesired medical care at the end of life. The act of writing down one’s EOL wishes is simple, inexpensive, and, when utilized effectively, empowering. 35 ADVANCE DIRECTIVE ADVOCACY References Agency for Healthcare Research and Quality. (2013). Care coordinators engage in culturally sensitive discussions about advance directives with seniors, increasing completing rates and reducing disparities between African Americans and Whites. U.S. Department of Health and Human Services. Retrieved from https://innovations.ahrq.gov/profiles/carecoordinators-engage-culturally-sensitive-discussions-about-advance-directives-seniors Aging With Dignity. (n.d.). Retrieved from https://www.agingwithdignity.org/five-wishes.php AHC Media. (January 1, 2016). The challenge of advance care planning billing. Retrieved from www.ahcmedia.com/articles/136977-challenges-still-persist-in-advance-careplanning-billing Aldridge, M.D., & Kelley, A.S. (2015). The myth regarding the high cost of end-of-life care. American Journal of Public Health 105(12), 2411-2415. doi: 10.2105/AJPH.2015.302889 American Academy of Family Physicians. (2015). Ethics and advance planning for end-of-life care. Retrieved from http://www.aafp.org/about/policies/all/planning-care.html American Bar Association. (n.d.). Law for older Americans. Retrieved from http://www.americanbar.org/groups/public_education/resources/law_issues_for_consume rs/patient_self_determination_act.html American Cancer Society. (2015). Types of advance directives. Retrieved from http://www.cancer.org/treatment/findingandpayingfortreatment/understandingfinancialan dlegalmatters/advancedirectives/advance-directives-types-of-advance-health-caredirectives American Nurses Association. (2010). Registered nurses roles and responsibilities in providing ADVANCE DIRECTIVE ADVOCACY 36 expert care and counseling at end of life. Position Statement. Retrieved from http://www.nursingworld.org/MainMenuCategories/EthicsStandards/Ethics-PositionStatements/etpain14426.pdf Arizona Veterans Standdown Alliance, (n.d.) Retrieved from http://www.arizonastanddown.org Blatt, L., & Crawford, S. (2015). Palliative care in serious and persistent mental illness. PowerPoint presentation at 2015 AAHPN & HPNA Annual Assembly – American Academy of Hospice and Palliative Medicine & Hospice and Palliative Nurses Association. Cagle, J.G. (2009). Weathering the storm: Palliative care and elderly homeless persons. Journal of Housing for the Elderly 23, 29-46. doi: 10.1080/02763890802664588 Centers for Disease Control. (2014). Give peace of mind: advance care planning. Retrieved from http://www.cdc.gov/aging/advancecareplanning/ Centers for Disease Control. (2015). Leading causes of death. Retrieved from http://www.cdc.gov/nchs/fastats/leading-causes-of-death.htm CMS.gov. (October 30, 2015). Proposed policy, payment, and quality provisions changes to the Medicare physician fee schedule for calendar year 2016. Centers for Medicare & Medicaid Services. Retrieved from https://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2015-Fact-sheetsitems/2015-10-30-2.html Dartmouth Atlas of Healthcare. (2015) End-of-life care. Retrieved from http://www.dartmouthatlas.org/keyissues/issue.aspx?con=2944 Department of Economic Security. (2014). Homelessness in Arizona annual report 2014. Retrieved from ADVANCE DIRECTIVE ADVOCACY 37 https://www.azdes.gov/InternetFiles/Reports/pdf/des_annual_homeless_report_2014.pdf Detering, K.M., Hancock, A.D., Reade, M.C., & Silvester, W. (2010). The impact of advance care planning on end of life care in elderly patients: randomized controlled trial. BMJ: British Medical Journal 340(1345c). Doi: 10.1136/bmi.c1345 Durbin, C.R., Fish, A.F., Bachman, J.A., & Smith, K.V. (2010). Systematic review of educational interventions for improving advance directive completion. Journal of Nursing Scholarship 42(3), 234-241. Doi: 10.1111/j.1547-5069.2010.01357.x Esperanza, S.H. (2015). Arizona and the aging homeless population: Preparing for the future. Department of Economic Security. Retrieved from https://des.az.gov/sites/default/files/Arizona_and_the_Aging_Homeless_Population.pdf Ganzini, L., Socherman, R., Duckart, J., & Shores, M. (2010). End-of-life care for veterans with schizophrenia and cancer. Psychiatric Services, 61(7), 725-728. doi:10.1176/appi.ps.61.7.725 Goodwin, Z., Kiehl, E.M., & Peterson, J.Z. (2002). King’s theory as foundation for an advance directive decision-making model. Nursing Science Quarterly 15(3), 237-241. Hemani, S., & Letizia, M. (2008). Providing palliative care in end-stage heart failure. Journal of Palliative Nursing 10(2), 100-105. Retrieved from http://www.medscape.com/viewarticle/573871_5 Jones, C.A., Perera, A., Chow, M., Ho, I., Nguyen, J., & Davachi, S. (2009). Cardiovascular disease risk among the poor and homeless -- what we know so far. Current Cardiology Review 5(1), 69-77. doi: 10.2174/157340309787048086 Jones, A.D., Moss, A.J., & Harris-Kojetin, L.D. (2011). Use of advance directives in long-term care populations. National Center for Health Statistics, NCHS Data Brief, No. 54. US ADVANCE DIRECTIVE ADVOCACY 38 Department of Health and Human Services. King, I.M. (2007). King’s conceptual system, theory of goal attainment, and transaction process in the 21st century. Nursing Science Quarterly 20(2), 109-116. Doi: 10.1177/0894318407299846 Ko, E., & Lee, J. (2014). End-of-life treatment preference among low-income older adults: a race/ethnicity comparison study. Journal of Health Care for the Poor and Underserved 25(3), 1021-1033. doi: 10.1353/hpu.2014.0129 Ko, E., Lee, J., & Hong, Y. (2015). Willingness to complete advance directives among lowincome older adults living in the USA. Health and Social Care in the Community. doi: 10.1111/hsc.12248 Ko, E., & Nelson-Becker, H. (2014). Does end-of-life decision making matter? Perspectives of older homeless adults. American Journal of Hospice & Palliative Medicine 31(2), 183188. doi: 10.1177/1049909113482176 Leung, A.K., Nayyar, D., Sachdeva, M., Song, J., & Hwang, S.W. (2015). Chronically homeless persons’ participation in an advance directive intervention: A cohort study. Palliative Medicine. doi: 10.1177/0269216315575679 Massachusetts Department of Mental Health (n.d.). Do It Your Way. Retrieved from http://www.promotingexcellence.org/mentalillness/index.html Millman, J. (2014). It’s time to bury the “death panel” myth for good. Is this the way to do it? The Washington Post. Retrieved from http://www.washingtonpost.com/blogs/wonkblog/wp/2014/09/17/its-time-to-bury-thedeath-panel-myth-for-good-is-this-the-way-to-do-it/ Murphy, S.T., Palmer, J.M., Azen, S., Frank, G., Michel, V., & Blackhall, L.J. (1996). Ethnicity ADVANCE DIRECTIVE ADVOCACY 39 and advance care directives. Journal of Law, Medicine & Ethics 24, 108-117. National Health Care for the Homeless Council. (2011). Clinical recommendations for the medical respite setting. Nashville, TN: Edgington, S. (Ed.). Available from www.nhchc.org. National Health Care for the Homeless Council. (2013). Aging and housing instability: Homelessness among older and elderly adults. In Focus: A Quarterly Research Review of the National HCH Council 2(1). Retrieved from http://www.nhchc.org/wpcontent/uploads/2011/09/infocus_september2013.pdf National Hospice and Palliative Care Organization. (2013). NHPCO’s facts and figures: Hospice care in America. Retrieved from http://www.nhpco.org/sites/default/files/public/ Statistics_Research/2013_Facts_Figures.pdf Payne, K.L., Prentice-Dunn, S., & Allen, R.S. (2010). A comparison of two interventions to increase completion of advance directives. Clinical Gerontologist 33, 49-61. Doi: 10.1080/07317110802678375 POLST (2015). For patients and family. Retrieved from http://www.polst.org/advance-careplanning/ Porter-O’Grady, T., & Malloch, K. (2015). Changing organizations, institutions, and government. In R.M. Patton, M.L. Zalon, & R. Ludwick (Eds.), Nurses making policy: From bedside to boardroom (pp. 238-276). New York, NY: Springer Publishing Company. Rosswurm, M.A., & Larrabee, J.H. (1999). A model for change to evidence-based practice. Image: Journal of Nursing Scholarship, 31(4), 317-348. Retrieved from http://www.library.armstrong.edu/eres/docs/eres/NURS4445- ADVANCE DIRECTIVE ADVOCACY 40 1_TAGGART/444502tagModelforChange.pdf Song, J., Bartels, D.M., Ratner, E.R., Alderton, L., Hudson, B., & Ahuwalia, J.S. (2006). Dying on the streets: Homeless persons’ concerns and desires about end of life care. Society of General Internal Medicine 22, 435-441. doi: 10.1007/s11606-006-0046-7 Song, J., Wall, M.M., Ratner, E.R., Bartels, D.M., Ulvestad, N., & Gelberg, L. (2008). Engaging homeless persons in end of life preparations. Journal of General Internal Medicine 23(12), 2031-2045. doi: 10.1007/s11606-008-0771-1 Song, J., Ratner, E.R., Wall, M.M., Bartels, D.M., Ulvestad, N., Petroskas, D., … Gelberg, L. (2010). Effect of an end-of-life planning intervention on the completion of advance directives in homeless persons. Annals of Internal Medicine, 153, 76-84. doi: 10.7326/0003-4819-153-2-201007200-00003 Spoelhof, G.D., & Elliott, B. (2012). Implementing advance directives in office practice. American Family Physician 85(5), 461-466. Retrieved from http://www.aafp.org/afp/2012/0301/p461.html#afp20120301p461-t2 Substance Abuse and Mental Health Services Administration. (2011). Current statistics on the prevalence and characteristics of people experiencing homelessness in the United States. Retrieved from http://homeless.samhsa.gov/ResourceFiles/hrc_factsheet.pdf United States Interagency Council on Homelessness. (n.d.). Defining homelessness and chronic homelessness. Retrieved from http://usich.gov/action/what-it-means-to-endhomelessness/defining-homelessness-and-chronic-homelessness/ United States Interagency Council on Homeless. (2013). Substance abuse. Retrieved from http://usich.gov/issue/substance_abuse U.S. Department of Housing and Urban Development. (November 2015). The 2015 annual ADVANCE DIRECTIVE ADVOCACY 41 homeless assessment report (AHAR) to Congress. Retrieved from https://www.hudexchange.info/resources/documents/2015-AHAR-Part-1.pdf White, B.M., & Newman, S.D. (2015). Access to primary care services among the homeless: A synthesis of the literature using the equity of access to medical care framework. Journal of Primary Care & Community Health 6(22), 77-87. doi: 10.1177/2150131914556122 42 ADVANCE DIRECTIVE ADVOCACY Appendix A Table 1 Evaluation Table Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Agency for Healthcare Research and Quality (2014) King’s Theory of Goal Attainment SR (RCTs; controlled trials, no randomization; observational studies; expert opinion) N = ns Recommendations: multiple Review NS Target population: Patients whose death in 12 mo would not be surprising; with chronic, lifelimiting illness experiencing worsening symptoms; >55 yo in any stage of health IV1: Help pt. ID a surrogate Pro’s: Highest LOE IV2: Incorporate pt.’s goals, preferences, and choices into the AD Con’s: Not specifically directed at HP Advance care planning, Michigan Quality Improvement Consortium USA Funding: Michigan Quality Improvement Consortium Bias: None noted Purpose: To improve advance care planning through EB clinical practice guidelines Application to Practice/Decision for Use Level 1 Harm: None noted IV3: Encourage pt. to complete an AD IV4: Encourage pt. to discuss preferences w/surrogate Age: ns Gender: ns Race: ns Setting: ns ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 43 ADVANCE DIRECTIVE ADVOCACY Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Detering, Hancock, Reade, & Silvester (2010) Roy’s Adaptation Model RCT N=309 (n=154 intervention: ACP Respecting Patient Choices model) IV1: Formal ACP using Respecting Patient Choices model questionnaire, review of MR T tests DV1: 85% EOL wishes known and followed c/t 30% control (p<0.001) The impact of advance care planning on end of life care in elderly patients: randomised controlled trial AUSTRALIA Funding: Australian Commonwealth and Victorian state governments (Respecting Patient Choices program) Bias: None noted Purpose: To investigate impact of ACP on EOL in elderly patients Survey of pt. satisfaction Impact of events scale IV2: UC Age: 84 (84 control) Gender: 54% M (41% control) Race: NS Setting: Melbourne university hospital Inclusion criteria: > 80 yo, legally competent, Englishspeaking, expected hospitalization >24 hours DV: Proportion of pts. who died and whose EOL wishes were respected Hospital anxiety and depression scale MannWhitney U tests X2 tests Fisher exact tests 83% satisfaction w/quality of death vs. 48% control (p=0.02) Application to Practice/Decision for Use Level II Pro’s: intervention by trained, nonmedical staff, Respecting Patient Choices has been implemented in US Harm: None noted < stress (p<0.001), single written intervention in adult clinic outpatients and hospitalized elderly (p < .05) Systematic review of educational interventions for improving advance directive completion USA, Canada Purpose: To analyze evidence on effectiveness of types of educational interventions resulting in completion of AD’s Age: 26-93 yo Gender: F>M Race: Wh, Bl Setting: US, Canada outpt. hospitals, inpt. hospitals, communitybased Funding: NS Bias: none noted Inclusion criteria: described educational interventions, provided info to calculate % of completed AD’s, publication 19912009 IV2: Combo educational (written + video, verbal + video, verbal + computer, written + verbal + video) Application to Practice/Decision for Use Level I Pro’s: Highest LOE, Con’s: Tepid conclusions about effectiveness of interventions Harm: None noted Combo written + verbal Combo written + computer DV: No. of completed AD’s ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 45 ADVANCE DIRECTIVE ADVOCACY Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Ganzini et al. (2010) Roy’s Adaptation Model Crosssectional N = 256 n = 60 IV1: SMI dx Percentage of AD’s and POLST Purpose: compare EOL care between veterans w/ and w/out schizophrenia who died of cancer Age: 64 ave. Gender: M Race: NS Setting: VA data from 7 medical centers in Oregon, Wash., Alaska DV: No. of completed AD’s or POLST’s Student’s t test, chi square test (2tailed) DV1: 58% vet’s w/SMI > had AD, more likely to have POLST (No significant difference from nonSMI vet’s.) End-of-life care for veterans with schizophrenia and cancer USA Funding: NS Bias: none noted Inclusion criteria: Vet’s who died of cancer betw. 2003 – 2008, schizophrenia or schizoaffective dx, attendance at min. 1 mental health OP appt., prescribed antipsychotic Application to Practice/Decision for Use Level III Pro’s: Addresses SMI’s influence on AD completion Con’s: May not be generalizable beyond veterans Harm: None noted ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 46 ADVANCE DIRECTIVE ADVOCACY Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Ko & Lee (2013) The Transtheoretical Model Randomized cross-sectional study N = 256 IV1: Completion of AD? Existence of AD? (questionnaire) Cronbach alpha DV1: 20% completion of AD Completion of advance directives among lowincome older adults: does race/ethnicity matter? USA Funding: John A. Hartford Foundation Bias: none noted Purpose: to examine prevalence of AD completion and effects of race/ethnicity on AD completion in socioeconomically marginalized older adults Age: 71 ave. Gender: M/F Race: Wh, H, Bl Setting: 2 Calif. supportive housing facilities and a senior center Inclusion criteria: >60, Wh/H/Bl, cognitively competent IV2: Knowledge of AD’s IV3: Attitudes towards AD’s IV4: Healthrelated variables (having a PCP, recent ICU admission, chronic health conditions) IV5: Sociodemographic variables (age, gender, education, religion, income, recent homelessness) Behaviors in AD completion AD knowledge (questionnaire) attitudes toward AD’s and distrust toward local h/c (Health Care System Distrust Scale) social support (Lubben Support Network Scale-6) and living contexts health-related and sociodemo variables (dichotomous Chi-square test Hierarchical logistic regression Monte Carlo Markov Chain multiple imputation techniques DV2: Wh > AD knowledge than other races; those w/knowledge 15x > complete AD DV3: Wh > (+) attitudes towards AD than other races Application to Practice/Decision for Use Level III Pro’s: Included Spanish language participants, delineates factors influencing willingness to complete AD’s Con’s: Lower LOE Harm: None noted DV4: Previous ICU admission 4.3x > complete AD’s DV5: Wh (27.6%) AD completion c/t Bl (21.4%) and ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 47 ADVANCE DIRECTIVE ADVOCACY scale) H (9.8%). Higher income 2.4x > complete AD ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 48 ADVANCE DIRECTIVE ADVOCACY Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Ko, Lee & Hong (2015) King’s Theory of Goal Attainment Crosssectional N = 204 IV1: selfrated health Willingness to complete an AD t-test DV1: Fair to poor health= 53.7% willingness to complete AD’s (OR=1.43, 95% CI=1.071.90) c/t good to better health (46.2%) Willingness to complete advance directives among lowincome older adults living in the USA USA Funding: John A. Hartford Fdn. Bias: None noted Purpose: To assess willingness to complete AD’s among lowincome elderly who have not completed one already Age: 70.8 ave Gender: M/F Race: Wh, H, Bl Setting: 2 San Diego HUD public housing programs for lowincome adults, incl. previously homeless, 1 sr. ctr. Inclusion criteria: > 60, Wh/Bl/H, cognitively competent 32% of participants had lived on the IV2: Availability of h/c proxy IV3: attitudes re: AD’s IV4: social support s chi-square test hierarchical logistic regression Markow chain Monte Carlo multiple imputation DV2: Availability potential h/c proxy 73% willingness to complete AD c/t 57% Application to Practice/Decision for Use Level III Pro’s: Hispanic and Black participants equal voice as Whites, Spanishlanguage participants included, provides insight into willingness to fill out AD Con’s: Lower LOE Harm: None noted DV3: + attitude towards AD > AD willingness (t[202]=2.26, p=0.02) DV4: >social ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 49 ADVANCE DIRECTIVE ADVOCACY street, 25% in shelters, 14.6% in transitional housing. 47% w/friends or family because of housing issues support = > AD willingness (t[202]=2.11, p=0.04DV1: 72% willing to complete AD ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 50 ADVANCE DIRECTIVE ADVOCACY Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Leung, et al (2015) King’s Theory of Goal Attainment Cohort Study N = 205 Purpose: To determine rate of AD completion w/CG intervention, identify characteristics assoc. w/AD completion, and describe EOL preferences among chronically HP Age: 55 ave. Gender: M Race: 70% Wh, 8.8% Bl, 11.2% Asian, 6.8% NA Setting: Toronto homeless shelter IV1: sociodemo characteristics, health status, use of h/c services on completion rates No. of completed AD’s Chisquare tests, T tests, MannWhitney U tests, Wilcoxon signedrank test 50.2% completed AD (p > .05) Chronically homeless persons’ participation in an advance directive intervention: a cohort study Canada Funding: “No specific grant” Bias: None noted Inclusion criteria: English, shelter resident, decisional capacity IV2: Knowledge of EOL wishes but haven’t told anyone DV: No. of completed AD’s Surveys, postcompletion surveys Application to Practice/Decision for Use Level III DV1: Sociodemo characteristics, health status, h/c use not associated w/completion (p > .05) Pro’s: Inexpensive intervention, limned HP’s attitudes towards EOL, nonspecialist utilization broadens generalizability DV2 > complete AD (p=0.001) Con’s: lower LOE Additional findings: 61.2% named proxy (p > .05) Harm: None noted DV4: 94.1% want CPR (p > .05) ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 51 ADVANCE DIRECTIVE ADVOCACY Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Payne, PrenticeDunn & Allen (2010) Protection Motivation Theory Nonrandomized cohort study N = 42 (n = 19 PMT, n = 23 health aging) IV1: Protection Motivation Theory Intervention Health measure (SF20 physical functioning subscale) ANCOVA IV2: health aging intervention AD questionnaire DV: PMT intervention < health aging intervention in AD completion A comparison of two interventions to increase completion of advance directives USA Funding: NS Bias: none noted Socioemotional Selectivity Theory Purpose: To compare an intervention based on protection motivation theory to a general health aging intervention on AD completion Age: 69.7 ave. Gender: 83% F Race: 93% W, 7% Bl Setting: senior citizen organizations, nutrition sites, churches DV: AD Completion Demographic questionnaire Chisquare test Application to Practice/Decision for Use Level III Pro’s: Con’s: Small sample size, tepid results Harm: None noted Mental Status Questionnaire Inclusion criteria: No current AD, cognitively intact, > 50 ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 52 ADVANCE DIRECTIVE ADVOCACY Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Song, et al (2010) King’s Theory of Goal Attainment RCT N = 262 HP (n = 145 counselorguided group; n = 117 selfguided group) IV1: Counselorguided Rate of AD completion measured by completed doc’s 2-sided Fisher exact tests 26.7% completion rate (total) (95% CI) Effect of an end-of-life planning intervention on the completion of advance directives in homeless persons USA Funding: Nat’l Inst. For Nursing Research and Nat’l Ctr on Minority Health and Health Disparities Bias: ns Purpose: Determine whether HP will complete a counseling session on ACP and fill out an AD Age: 18-74 Gender: 74% M Race: 54% Bl, 32% Wh, 7% NA Setting: 8 Minneapolis sites/settings serving HP Inclusion criteria: speak English, decisional capacity, >18, homeless prior 6 mo IV2: Selfguided DV: No. of completed AD’s, attendance at counseling session BreslowDay test for homogeneity MantelHaenszel odds ratio DV1: 37.9% completion rate (counselorguided group) DV2: 12.8% self-guided group (p < .001) Application to Practice/Decision for Use Level II Pro’s: inexpensive intervention, similar population, Large sample size, social service agency sites similar to mine Con’s: SELPH AD apparently no longer available (unable to find further info) Harm: None noted ACP = Advance Care Planning, AD = Advance Directive, Bl = Black, CG = Counselor Guided, EOL = End of Life, H = Hispanic, HP = Homeless Population, NA = Native American, ns = not stated, OP = outpatient, PCP = Primary Care Physician, PMT = Protection Motivation Theory, POLST = Physician’s Orders for Life-Sustaining Treatment, RCT = Randomized Control Trial, SG = Self Guided, SR = Systematic Review, UC = Usual Care, Wh = White 53 ADVANCE DIRECTIVE ADVOCACY Title/Author Citation Theoretical Framework Study Design Sample/ Setting Variables Measurement Data Analysis Findings Song, Wall, et al (2008) King’s Theory of Goal Attainment RCT N = 59 IV1: SG + counseling Rate of AD completion Fisher’s exact tests Purpose: To determine whether HP will complete an AD n = 29 CG group 44% AD completion (p = .02) IV2: SG 3-mo f/u EOL-related knowledge, attitude, behaviors CochranArmitage trend tests Engaging homeless persons in end of life preparations n = 30 SG group DV1: No. of completed AD’s USA Funding: Nat’l Inst. Of Nursing Research, NIH Bias: one author consults on another NIH-funded project Age: ave. 45 Gender: 75% M Race: 50% Bl, 34% Wh, 7% NA, 3% H Setting: Drop-in center for homeless in St. Paul, MN DV2: Frequency of worry & death PROC GENMOD in SAS with a REPEATED statement DV1: 59% AD completion CG group 30% AD completion SG only group DV2: < from 50% to 12.5% among AD completers (p=.05) Application to Practice/Decision for Use Level II Pro’s: Similar setting, AD geared specifically to HP, provides AD example Con’s: Convenience sample, small sample, limited to Englishspeakers, minimal Hispanics in sample Harm: None noted < from 25% to 12.5% in non-AD completers (p<.05) Inclusion criteria: Englishspeaking, >18 yo, homeless AD = Advance Directive, B = Black, CS = Cross Sectional, EOL = End of Life, H = Hispanic, HC = Health Care, NA = Native American, RCT = Randomized Control Trial, SR = Systematic Review, UC = Usual Care, W = White ADVANCE DIRECTIVE ADVOCACY previous 6 mo’s AD = Advance Directive, B = Black, CS = Cross Sectional, EOL = End of Life, H = Hispanic, HC = Health Care, NA = Native American, RCT = Randomized Control Trial, SR = Systematic Review, UC = Usual Care, W = White 54 55 ADVANCE DIRECTIVE ADVOCACY Appendix B Table 2 Synthesis Table Year Study Design SR RCT CS Cohort Theoretical Framework Roy’s Adaptation Model King’s Th. Of Goal Attainment Transtheoretical Model Protection Motivation Th. Socioemotional selectivity Th. Country USA Canada Australia Demographics Age** Gender Race AHRQ Detering 2014 2010 Durbin 2010 X X Ganzini 2010 Ko/Lee 2013 Ko/Lee/Hong 2015 X X X Leung Payne 2015 2010 Song1 2010 Song2 2008 X X X X X X X 70 F>M W>B 18-74 45 M>F M>F B,W,NA B,W,NA, H X X X X X X X X X X X X X X X X X 29-93 F>M B, W 64 M X X X X ns M/F 84 M/F 71 M/F B, W, H 71 M/F B, W, H 55 M B,W,A AD = Advance Directive, B = Black, CS = Cross Sectional, EOL = End of Life, H = Hispanic, HC = Health Care, NA = Native American, RCT = Randomized Control Trial, SR = Systematic Review, UC = Usual Care, W = White 56 ADVANCE DIRECTIVE ADVOCACY Homeless Veterans SA IV UC Formal ACP Single edu. session Combo edu. Session Schizophrenia dx Attitudes towards AD Social support Self-rated health Availability hc proxy Race/ethnicity Protection motivation theory intervention Socioemotional selectivity theory intervention DV AD completion EOL wishes respected Pt./family satisfaction w/hospital Willingness to X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X AD = Advance Directive, B = Black, CS = Cross Sectional, EOL = End of Life, H = Hispanic, HC = Health Care, NA = Native American, RCT = Randomized Control Trial, SR = Systematic Review, UC = Usual Care, W = White 57 ADVANCE DIRECTIVE ADVOCACY complete AD Worry about death Outcomes X >AD completion >EOL wishes respected >pt./family hospital satisfaction HC proxy named >willingness to complete AD