DEVELOPMENT AND PILOT TESTING OF THE END-OF-LIFE READINESS ASSESSMENT (ERA) SURVEY

Abstract An emerging conceptual framework on the relationship between serious and sensitive end-of-life (EOL) discussions and patient hope points to the mediating role that patient readiness may play. Additional research has also found that among some patients, engaging in and EOL conversation before they are ready may actually cause harm. Presently, health practitioners do not have a way to measure patient readiness. Therefore, the purpose of this study was to develop a survey assessment tool, rooted in research, to measure patient readiness to engage in an EOL discussion. The 16-item survey was initially developed by a gerontological researcher and clinical oncologist, and tested for face reliability and appropriateness among a geriatrician and three patients with chronic illness. In May 2018, the final version of the survey was pilot tested among 168 patients attending their regularly scheduled oncology appointment. On average, most participants identified as Caucasian (74%) females (61%) in partnerships (58%) and having a cancer diagnosis (64%). Bivariate analyses revealed that older age (60+, p=0.008), Caucasian race (p=0.04), and reported greater knowledge of community and supportive services (p=0.049) was significantly associated with increased readiness to engage in an EOL discussion. Additionally, a higher score on the Advance Care Planning Engagement Survey was significantly associated with increased readiness (p<0.001). Findings reveal patient groups that might be more appropriate targets for interventions, education, and resources and highlight the role of interprofessionals in outpatient health settings to “prime the pump” for these types of conversations.

MIDUS study, we utilize hybrid fixed effects models to assess the relationship between future aspirations and income.We find that, net of age, health, and a host of other timevarying factors, more positive future aspirations are indeed related to higher income over time, but that this relationship takes different shapes in different contexts.In particular, in lower quality neighborhoods, higher future aspirations lead to worse economic outcomes over the life course, while in higher quality neighborhoods, higher aspirations are indeed related to higher incomes.We thus argue that aspirations are only helpful in some contexts, and are inherently contextual not just in their sources but also in their effects.

END-OF-LIFE CARE CULTURAL COMPETENCE TRAINING FOR HOSPICE STAFF: FINDINGS FROM A NATIONAL HOSPICE SURVEY
Nathan Boucher, 1 Kimberly S. Johnson 2 , 1. Durham VA Health System,Durham,North Carolina,United States,2. Duke University,Durham,North Carolina,United States Compared to whites, racial/ethnic minorities are less likely to enroll in hospice and if they enroll, more likely to experience poor quality care.Building cultural competence (CC) among hospice staff is a strategy that may reduce these disparities.We conducted a national survey of hospices' practices to promote CC.A total of 197 hospices participated; most were not-for-profit (80%) with an average daily census over 100 (53%); 73% offered staff cultural competence training (CCT).There were no differences in characteristics of hospices who offered CCT and those that did not.Of hospices offering CCT, 54% held it annually.Most trainings were one hour (60%); content was delivered via web (58%) and/or lecture (57%).While over 90% of staff (i.e., nurses, social workers, and chaplains) completed CCT, a smaller proportion of medical directors (63%), senior leaders (70%) and board members (23%) did so.Most common (>70%) topics were: cross-cultural communication, death and illness beliefs, and spirituality's role, and healthcare disparities.The majority focused on African-Americans (83%), Hispanics (76%), and Asians (61%)-the most common U.S. minority groups.Almost 30% reported no assessment of effectiveness of CCT while 45% reported a quiz at the end.In this study, most hospices offered some CCT.CCT has been shown to improve healthcare providers' knowledge and skills in caring for diverse patients and is associated with increased patient satisfaction.Future research should evaluate effectiveness of CCT in improving the ability of hospices to deliver high quality end-of-life care to diverse groups of older adults.

DEVELOPMENT AND PILOT TESTING OF THE END-OF-LIFE READINESS ASSESSMENT (ERA) SURVEY
Alexis M. Coulourides Kogan, 1 and Julie Taguchi 2 , 1. University of Southern California, Keck School of Medicine,Los Angeles,California,United States,Santa Barbara,California,United States An emerging conceptual framework on the relationship between serious and sensitive end-of-life (EOL) discussions and patient hope points to the mediating role that patient readiness may play.Additional research has also found that among some patients, engaging in and EOL conversation before they are ready may actually cause harm.Presently, health practitioners do not have a way to measure patient readiness.Therefore, the purpose of this study was to develop a survey assessment tool, rooted in research, to measure patient readiness to engage in an EOL discussion.The 16-item survey was initially developed by a gerontological researcher and clinical oncologist, and tested for face reliability and appropriateness among a geriatrician and three patients with chronic illness.In May 2018, the final version of the survey was pilot tested among 168 patients attending their regularly scheduled oncology appointment.On average, most participants identified as Caucasian (74%) females (61%) in partnerships (58%) and having a cancer diagnosis (64%).Bivariate analyses revealed that older age (60+, p=0.008),Caucasian race (p=0.04), and reported greater knowledge of community and supportive services (p=0.049) was significantly associated with increased readiness to engage in an EOL discussion.Additionally, a higher score on the Advance Care Planning Engagement Survey was significantly associated with increased readiness (p<0.001).Findings reveal patient groups that might be more appropriate targets for interventions, education, and resources and highlight the of interprofessionals in outpatient health settings to "prime the pump" for these types of conversations.

HOSPICE VISIT INTENSITY BY PHYSICIANS AND NURSE PRACTITIONERS ON THE GENERAL INPATIENT LEVEL OF CARE
Thomas J. Christian, 1 Joan Teno, 2 Pedro L. Gozalo, 3  The Medicare Hospice Benefit's General Inpatient (GIP) level of care provides short-term services for pain and symptom management in an inpatient facility that cannot be managed in the patient's home.Relatively little is known about how beneficiaries utilize services during GIP care.Among a cohort of Medicare hospice beneficiaries utilizing GIP during Federal Fiscal Year 2014 (FY2014), we used 100% Medicare hospice and Part B claims to identify physician and nurse practitioner services concurrent with GIP dates.We estimated logistic regression models to determine the likelihood a beneficiary never receives physician or nurse practitioner services.We found that among the 1.5 million GIP days serviced in FY2014, more than half (52.4%) lacked any recorded physician or nurse practitioner services.Absence rates for these services were particularly high among hospice GIP days provided in inpatient facilities (69.1% missing services), long-term care hospitals (84.3% missing services), and skilled nursing facilities (85.3% missing services).Moreover, one in five hospice episodes having at least three sequential GIP days lacked any physician or nurse practitioner services.Relative to hospice inpatient units, rates of absence were higher among episodes beginning in long-term care hospitals [59.3% long-term care hospital vs. 11.5% hospice inpatient units; AOR 9.65 95% CI 7.47-12.46]and skilled nursing facilities [51.3% skilled nursing facility vs. 11.5% hospice inpatient units; AOR 5.98, 95% CI 5.63-6.36].More in depth research and monitoring is needed to further understand dimensions of GIP care provision, to ensure that hospice beneficiaries are receiving adequate services regardless of their inpatient setting.Nursing homes (NH) must implement best practices to improve care.The purpose of this study is to understand NH characteristics that help or hinder implementation of the Goals of Care (GOC) intervention, an evidence-based decision aid to guide decision-making in advanced dementia.Study design was a cross-sectional staff survey at 11 NHs in North Carolina that participated in the GOC trial.Questions measured the dependent variable of implementation effectiveness (IE) (the consistency and quality of use of the GOC intervention).NH organizational characteristics were measured using publicly available data and administrator surveys.Averages were obtained, IE and NH factors above the average were ranked as high and vice versa.Analysis consisted of pattern matching logic, predicted results are compared to actual results, using within-case and cross-case analyses.NHs with high IE were expected to have the majority (five or more) of NH characteristics to be high to confirm the within (within-case analysis) and between (cross-case analysis) relationships.Within-case analysis expected results were met in 5 NHs, with 3 highs (IE high and five or more characteristics were high) and 2 lows (IE low and five or more characteristics were low).Among three NHs with high IE, the following characteristics were high: Medicare and SNF/ICF beds.Across 6 of the 11 NHs, high total beds and percent of White residents were related to high IE.NHs with enhanced resources may have fewer challenges implementing innovations compared to resource constrained NHs.Implementation strategies should account for resource limitations to promote successful implementation.

USING LONGITUDINAL HEALTH INFORMATION TO ENHANCE MORTALITY PREDICTION
Megan K. Beckett, 1 Marc N. Elliott, 1 Doug Ritenour, 2 Laura A. Giordano, 2 and Debra Saliba 1 , 1. RAND Corporation,Santa Monica,California,United States,2. Health Services Advisory Group,Phoenix,Arizona,United States The Vulnerable Elders Survey-13-Health Outcomes Survey (VES-13-HOS) 2.5 is a simple, validated method to predict two-year mortality using older adults' responses to the Medicare HOS.We explore whether adding longitudinal