scientiﬁc and quality improvement project poster...

Scientific and Quality Improvement Project Poster Abstracts

Advance Care Planning Conversations:Measuring Registered Nurses’ Self-EfficacyBefore and After an EducationalIntervention (QI701)Joyce Borgfeld, MSN, The University of Texas Healthat San Antonio, San Antonio, TX. Kristen Overbaugh,PhD RN ACNS CHPN, La Salle University, Philadel-phia, PA. Ana Fernandez, DNP, American SentinelUniversity, Aurora, CO. Andrea Berndt, PhD, The Uni-versity of Texas Health at San Antonio, San Antonio,TX.

Objectives� Describe advance care planning from theperspective of the registered nurse.

� Articulate changes in registered nurse advancecare planning self-efficacy following evidence-based training.

Background. Advance care planning (ACP) improvespatient outcomes at the end-of-life. Effective ACP re-quires specialized knowledge and communicationskills lacking in traditional nursing education and bynurses practicing in our local community. Substantialgaps in the literature exist in understanding nurses’ACP self-efficacy and the effects of innovative teachingapproaches on this outcome.Aim Statement. The purpose of this quality improve-ment project was to evaluate changes in registerednurses’ (RNs) ACP self-efficacy after participating ina multimodal evidence-based educational program de-signed to improve ACP conversations.Methods. This project was a comparative study using anon-experimental Before-After design with an equiva-lent control group. Participants completed a demo-graphic survey and the Advance Care Planning Self-Efficacy (ACP-SE) scale before and after a programcombining lecture, video-taped demonstration using asimulated patient scenario, and role-play activities. An-alyses were conducted using descriptive statistics andWilcoxon Signed Rank test. Bivariate correlations be-tween a global single-item measure of ACP-SE and themean score of the items on theACP-SE scale were exam-ined using Spearman Rho correlation coefficient toassess the validity of the ACP-SE scale, originally devel-oped for physicians, in this sample of RNs (N¼45).Results. Findings demonstrated a statistically signifi-cant improvement in ACP-SE scores following the pro-gram, z¼-3.65, p<.000, with a medium effect size(r¼.39). The median score on the ACP-SE scaleincreased from pre-activity (Md¼65) to post-activity

(Md¼74). There were strong, positive significant cor-relations between the global single-item and meanscore of ACP-SE pre-activity, rs¼.73, and post-activity,rs¼.78, supporting the validity of the ACP-SE scale inthis sample of RNs.

Conclusions and Implications. Findings demon-strated that a multimodal educational programimproved nurses’ self-efficacy in conducting ACP. Asnurses are uniquely positioned to facilitate ACP, addi-tional research is needed to determine if increasedACP-SE improves patient and family outcomes and de-creases healthcare costs.

Promoting Upstream Integration ofPalliative Care in Elderly Heart FailurePatients (QI702)Harmony Brown, DNP APRN, University of Utah Hos-pitals and Clinics, Salt Lake City, UT.

Objectives1. Summarize understanding and deployment of

the ARISE guideline.2. Articulate possible provider barriers to early inte-

gration of Palliative Care.Background. Heart failure (HF) is a devastating dis-ease that inordinately affects the elderly. AlthoughPalliative Care (PC) can be an invaluable tool in pro-moting symptom management and quality of life(QOL) for end-stage heart failure (ESHF) patients,it is a resource that is often underutilized by HF clini-cians because many providers do not know when inthe disease trajectory to refer patients for PC.Aim Statement. The purpose of this quality improve-ment project was to assess clinicians’ understanding ofPC to determine barriers preventing appropriate re-ferrals for elderly patients with stage 3 and 4 HF.This information was then used to develop an evi-dence-based guideline for promoting earlier PC refer-rals. Finally, the developed guideline was disseminatedas a resource for HF providers/clinicians.Methods. Baselinedatawere collected from the facilityPC team about the number of inpatient (IP) and outpa-tient (OP) referrals for PC that were received from theHF attending physicians for three months prior to theARISE guideline launch, and for three monthsfollowing the deployment of the guideline. A pretest-posttest design was the approach employed to measurechanges in provider attitude toward PC, frequency ofPC referrals, and barriers to consulting PC.

456 Journal of Pain and Symptom Management Vol. 57 No. 2 February 2019

http://crossmark.crossref.org/dialog/?doi=10.1016/j.jpainsymman.2018.12.210&domain=pdf


Results. Referral data from the PC team indicatedthat although IP referrals decreased slightly (17%)in the 3 months after the guideline launch, OP refer-rals increased by 40%. In general, pilot feedback fromthe ARISE guideline was positive with 100% of clini-cians indicating they felt the guideline was ‘‘very’’ or‘‘extremely’’ useable. In addition, 75% of cliniciansin the posttest indicated they anticipated using theguideline in the future.Conclusions and Implications. It is hoped that thisproject helped HF clinicians better understand thebenefits that PC can provide for their patients, andthat the ARISE guideline may be a viable tool to pro-mote upstream integration of PC into the care ofelderly ESHF patients and their families.

A Rapid Qualitative Synthesis of Insightsfrom the Veterans Administration Life-Sustaining Treatment Decisions Initiative(LSTDI) National Implementation (QI703)Cati Brown-Johnson, PhD, Stanford School of Medi-cine, Stanford, CA. Natalie Lo, OT MPH, Center forInnovation to Implementation, Los Altos, CA. KarleenGiannitrapani, PhD, VA Health Services Research andDevelopment Service/Stanford, Palo Alto, CA. JillLowery, PsyD, Veterans Health Administration, Wash-ington, D.C. Mary Beth Foglia, PhD MA RN, VeteransHealth Administration, Washington, D.C. Anne Wall-ing, MD, University of California, Los Angeles, Los An-geles, CA. David Bekelman, MD MPH, University ofColorado, Denver, CO. Karl Lorenz, MD MS MSHS,Stanford - VA Palo Alto Health Care System, PaloAlto, CA.

Objectives1. Describe the core elements of the Life-Sustaining

Treatment Decisions Initiative.2. Describe implementation barriers and facilita-

tors from the perspective of multiplestakeholders.

3. Define essentials of a lightning report rapid qual-itative synthesis methodological approach.

Background. The Veterans Health Administration’sLife-Sustaining Treatment Decisions Initiative(LSTDI) was initiated to elicit, document, and respectseriously ill Veterans’ goals, values, and preferencesfor life-sustaining treatments. To characterize imple-mentation barriers and facilitators, we interviewedstakeholders from four geographically diverse demon-stration sites (Salt Lake, Black Hills, Madison, andJames Lovell), and identified strategies to enhancecontinuing national rollout.Aim Statement. To characterize major contextualfactors affecting implementation of goals of care in adiverse health system.

Methods. We conducted 16 semi-structured inter-views (4/site) with providers and clinical leaders iden-tified using snowball sampling and characterizedthemes using rapid qualitative synthesis (RQS) whichincludes: pre-planning with evaluation partners; datagathering; data synthesis using Plus/Delta debriefingderived from LEAN techniques; report creation, withexecutive summary and findings that reflect Plus/Delta, refined with stakeholder input.Results. Respondents represented nursing, primarycare, social work, non-clinical administration, andleadership. Local issues were common, ‘‘Our badgessay one VA. But then everybody jokes that when you’veseen one VA you’ve literally only seen one VA.’’ Never-theless, general themes encompassed facilitators ofhaving clinical champions, national support, outpa-tient palliative care, and strong local advisory over-sight. Barriers included limited physician support,low provider self-efficacy for communication, lack ofownership by specialty/primary care clinicians,‘‘We’re not having great luck with the subspecialty pro-viders,’’ poor support (e.g., time and funding) fortraining, and staff/trainee turnover and inexperience.Insights included that middle manager outreach over-comes senior leader resistance, champions cultivategrass roots support for training, a dedicated traineris critical, and that small group training, and culti-vating long-term-care superusers, abets implementa-tion. Local implementation coordinators succeededthrough repeated incremental efforts that encom-passed train-the-trainer, peer mentoring, and culti-vating onsite experts/advocates.Conclusions and Implications. Site leadersdescribed critical strategies to support goals of careimplementation and improvement in an extremelylarge, diverse national health system. RQS methodol-ogy can provide just-in-time information to informongoing implementation of complex initiatives.

The Discrepancy Between Palliative andOncology Providers in Palliative Referrals(QI704)Bethany Calkins, MD MMM HMDC, VA Western NewYork Healthcare System, Buffalo, NY. Jaclyn Schneider,MD, VAWestern New York Healthcare System, Buffalo,NY. Raed Al Yacoub, MD, University at Buffalo, Buf-falo, NY. Vrinda Bhasin, MD, University at Buffalo,Buffalo, NY. Alan DeMaine, DO, University at Buffalo,Buffalo, NY. Rym Lamrad, MD, State University of NewYork at Buffalo, Minot, ND.

Objectives1. Justify the need for palliative care consultation to

outpatient oncologists by meeting with them on aregular basis in order to increase the rate ofreferrals.

Vol. 57 No. 2 February 2019 457Scientific and Quality Improvement Project Poster Abstracts



2. Identify other patient-centered outcomes as aresult of proactive palliative care consultation,including rate of completion of advance direc-tives, rate of state-authorized portable orders(SAPO) completion, and rates of in-hospitaland community deaths.

Background. According to NCCN and ASCO guide-lines, there is evidence that early palliative care consul-tation in patients with advanced cancer led tosignificant improvement in quality of life, mood andlonger survival as compared to patients receiving stan-dard care.Aim Statement. To improve the rate of palliative careconsultation by 20% in patients with advanced cancerover a 15 month period at Buffalo Veterans AffairsMedical Center (VAMC).Methods. Buffalo VAMC is an academic center affili-ated with the University at Buffalo. We provide bothoutpatient palliative and oncology services. In 2017,we decided to seek proactive palliative care consulta-tions (PPCC) with advanced cancer patients; definedas stage 4 lung cancer, any pancreatic cancer, or anycancer with more than 1 emergency, hospital or ICUadmissions. Based on our root analysis our interven-tion was a meeting between palliative and oncologyservices once a week since 01/01/2017 to 3/30/2018at the VA and reviewing the oncology cases on weeklybasis with average 80-100 patients per week. From 163patients requested to be seen by palliative team, oncol-ogists approved 63 proactive palliative consults withdenial rate of 60%.Results. PPCC resulted in an increase of referrals by33% over a 15 month period. Of those patients seenby outpatient palliative care, 92% died in a commu-nity setting (hospice or home) while those identifiedbut denied PPCC died in the community setting only50% of the time. Interestingly, patients whose PPCCwere denied but were later seen by the palliativecare team during a hospital admission died in thecommunity 91% of the time. In addition, advanceddirectives and SAPO increased by 30% and 65%respectively.Conclusions and Implications. Seeking proactivepalliative care consultation in patients with advancedcancer led to an increased outpatient referral ratewith a simultaneous decrease in in-hospital mortality.

Repeal and Replace: OvercomingInstitutional Biases to Implement a Safeand Efficacious End-of-life Opioid Infusion(QI705)Christopher Cauto, MD, Penn State Hershey MedicalCenter, Hershey, PA. Jeffrey Kowaleski, MD, Penn State

Health, Hershey, PA. Thomas Vissering, RPh MBA,Penn State Hershey Medical Center, Hershey, PA.

Objectives1. Contrast pharmacology-based end-of-life opioid

infusion guidelines with commonly observedreal-life practice.

2. Illustrate the challenges to implementing andmeasuring the impact of an institution-wideopioid project.

Background. End-of-life opioids, when deliveredexclusively via rapidly-titrated infusions withoutbolus doses, can lead to ineffective symptom controlearly and toxic levels late. Despite best practiceguidelines based on steady-state pharmacology, ourinstitution regularly delivered morphine using onlyan ‘‘every thirty minute’’ titratable drip with 20mg/hmax. Multiple systemic issues including a collectivemindset resistant to change perpetuated this un-sound practice an average of nine times per monthin 2016.Aim Statement. Improve safety and efficacy in end-of-life infusion-based opioid treatmentsMethods. In 2017, through a series of presentations,we generated interdisciplinary buy-in to remove thedangerous order and build a consensus replacementorder set. In our plan, morphine or fentanyl infusionsare nurse-titrated using objective FLACC (pain) andRDOS (dyspnea) assessments after at least four half-lives. Infusion max doses were halved and linked toq15minute bolus doses equal to the infusion dose. Ed-ucation included narrated e-learning (nursing) withpre/post survey to assess understanding and in-person, including Grand Rounds, and via email(pharmacists and providers).Results. Pre-intervention, end-of-life patientsreceiving the rapidly-titratable morphine infusionoften reached high rates (median 550% of initial infu-sion rate) in a matter of hours (median ¼ 5.25) whilePRN boluses were available and used in only 35.7% ofcases (n¼ 28). Initial post-intervention analysisshowed safer maximum levels (median 160% of initialinfusion rate) with slower time to maximum infusionrate (median 29 hours) while PRN boluses were avail-able and used in 69.6% of cases (n ¼ 22).Conclusions and Implications. Institution-widechange was achieved by addressing barriers at everylevel of implementation of the opioid order setfrom pharmacist to nurse to ordering provider. Basicsafety and efficacy measures have improved, butmore in-depth case review analysis is needed toassess fidelity of provider ordering and nursingexecution.

458 Vol. 57 No. 2 February 2019Scientific and Quality Improvement Project Poster Abstracts


The PACT Conversation Trigger Tool:Assessing Reliability, Acceptability andValidity (QI706)Lauren Chiec, MD, McGaw Medical Center of North-western University, Chicago, IL. Katie Clepp, MAMEd, Northwestern University, Chicago, IL. KelseyGradwohl, BA, Northwestern University, Chicago, IL.Eytan Szmuilowicz, MD, Northwestern Medicine, Chi-cago, IL. Gordon Wood, MD MSCI FAAHPM, North-western University, Chicago, IL.

Objectives1. Describe a novel approach to screening patients

for unmet advance care planning needs.2. Describe the reliability, acceptability and validity

of the PACT Trigger Tool.Background. Many patients who would benefit fromAdvance Care Planning (ACP) do not have discussionswith providers. One barrier is the lack of a standard-ized method to identify these patients.Aim Statement. As part of a larger initiative whoseaim was to increase ACP conversations, the Prefer-ence-Aligned Communication and Treatment(PACT) Trigger Tool was developed to help inpatientclinicians of any discipline identify patients with ACPneeds based on serious illness, care dependency andfunctional decline. This study evaluated the reliability,acceptability and validity of the tool.Methods. To assess inter-rater reliability, 100 patientswere double-scored across participating hospitals. Foracceptability, nurses (N¼185) completed a 10-itemsurvey using a Likert scale. Validity was measured us-ing retrospective chart reviews (N¼135).Results. The PACT Trigger Tool demonstrated mod-erate to high reliability. A factor analysis of the accept-ability survey showed that nurses found the tool useful(M¼4.07, SD¼0.73) and easy to complete (M¼4.24,SD¼0.67). Nurses who had been trained to use thetool found it easier to complete (F(1, 183)¼12.69,p<.001) and more useful (F(1, 183)¼26.42, p<.001)than those who had not. Chart reviews revealed thatpatients who triggered positive (n¼40) were signifi-cantly more likely to have a DNR order (43% vs11%) and palliative care consult (53% vs 20%) duringthe index admission and less likely to discharge home(53% vs 86%) compared with patients who triggerednegative (n¼95). Patients who triggered positive hadshorter survival times (M¼125 days, Mdn¼69 days)than those who triggered negative (M¼248 days,Mdn¼240 days; log rank test p<.001).Conclusions and Implications. The PACT TriggerTool is a reliable, acceptable, and a valid means ofidentifying hospitalized patients who may benefitfrom ACP discussions. A larger validity study and eval-uation of optimal ACP methods for patients whotrigger positive are needed.

Improving Patient-Reported OutcomesCollection Rate in an Outpatient PalliativeCare Clinic (QI707)Isaac Chua, MD, Dana-Farber Cancer Institute, Cam-bridge, MA. Jocelyn Siegel, MS, Dana-Farber CancerInstitute, Boston, MA. Victoria Hayne, MBA, Dana-Farber Cancer Institute, Boston, MA. NadineMcCleary, MD MPH, Dana-Farber Cancer Institute,Boston, MA.

Objectives1. Explain the significance of patient reported

outcome measures (PROMs) in the managementof advanced cancer patients.

2. List potential barriers for PROMs collection in anoutpatient palliative care clinic.

Background. From January 2018 to March 2018,Dana Farber Cancer Institute (DFCI) launched a pilotto collect Patient-Reported Outcomes of the CommonTerminology Criteria for Adverse Events (PRO-CTCAE) within the adult palliative care (APC) clinicusing smart tablets. We had a limited responseratedonly 20% of patients attempted to completethe questionnaire. After the pilot ended on March2018, smart tablets were discontinued and were re-placed with a paper version of PRO-CTCAE from Aprilto June 2018.Aim Statement. Our aim was to increase the patientattempt and collection rate of the paper PRO-CTCAEfrom 20% to 50%.Methods. Our primary outcome measure was the per-centage of paper PRO-CTCAE attempted andcollected. Eligible patients were established patientsscheduled to see a provider in the APC clinic. We im-plemented several Plan-Do-Study-Act (PDSA) cyclesincluding the implementation of the paper versionof the questionnaire, training and educating frontdesk staff, and posting provider reminders in examrooms. We used a statistical process control (SPC)chart to track percentage of attempted and collectedquestionnaires over time and to differentiate betweenspecial cause and common cause variation.Results. From April 2018 to June 2018, the PRO-CTCAE collection rate improved from 20% to 48%.Special cause variation was associated with implemen-tation of the paper version of the PRO-CTCAE andincreased front desk staff engagement. Increased pro-vider satisfaction was also associated with the paperversion of the PRO-CTCAE.Conclusions and Implications. Implementing ahigh-reliability process for collecting patient reportedoutcome measures in an outpatient palliative careclinic is complex and requires cohesive multi-disci-plinary teamwork, a user-friendly patient-facing andprovider-facing interface, and a streamlined workflow.The electronic version of PRO-CTCAE will resume in




September 2018. We will implement lessons learnedfrom the paper PRO-CTCAE implementation,including ongoing front desk staff engagement andan enhanced provider view in the electronic medicalrecord.

Increasing Awareness and Recognition ofPediatric Physician Orders for Life-Sustaining Treatment (POLST) at anAcademic Children’s Hospital: A QualityImprovement Project (QI708)Caitlin Dockweiler, MD, Kaiser Permanente, San Die-go, CA. Vishar Amin, MD, Kaiser Permanente South-ern California, San Diego, CA. Heather Pierce, MD,Rady Children’s Hospital/University of CaliforniaSan Diego, San Diego, CA. Ami Doshi, MD, Rady Chil-drens Hospital, San Diego, CA.

Objectives1. Describe the barriers to POLST recognition and

subsequent ordering of correct code status onadmission in the pediatric setting.

2. Identify possible strategies to increase documen-tation of POLST awareness and ordering of cor-rect code status on admission in the pediatricsetting.

Background. When children are admitted to ourinstitution using electronic admission orders, code sta-tus defaults to ‘‘full code.’’ POLST documents on fileare often not recognized by the admitting team. As aresult, there are instances in which code status is incor-rectly ordered on admission.Aim Statement. Increase documentation indicatingPOLST recognition by 5% and decrease inaccuratecode status orders in EMR by 5% within 1 year.Methods. Pre-data was collected via chart review ofpatients on the Palliative Care Registry in Epicadmitted within a two-month window. Documentationof the existence of POLST and whether the orderedcode status was consistent with the POLST form wasassessed. Intervention included adding a banner inEpic notifying admitting physicians of a POLST, witha corresponding hyperlink to POLST document.Post-data was collected by chart review of patients onthe Palliative Care Registry admitted within a two-month window after intervention was implemented.Results. Pre-intervention, 9% of available POLSTswere documented as recognized on admission. Threepatients (11%) had incorrect code status ordered onadmission. After intervention implementation, 12%of POLSTs were documented as recognized on admis-sion. One patient (5%) had the wrong code statusentered by the resident. This was recognized and cor-rected by the attending within several hours.Conclusions and Implications. After implementa-tion of the POLST banner, there were less instances

of the wrong code status being entered on admission,thus meeting smart aim of decreasing inaccurate codestatus orders. However, documentation of reviewingPOLST remains low. We feel this may be secondaryto the fact that the majority of POLSTs are filled outas full code, and banner alone may not be enoughto trigger documentation. One future aim for thisproject includes adding a line to the standard admis-sion H&P template indicating if POLST is presentand reviewed by primary team.

Addition of the Nurse Triage Role inImproving Inpatient Consultation Delivery:A Palliative Care Team Model QualityImprovement Initiative (QI709)Christina Fitch, DO MPH, Baystate Medical Center,Springfield, MA. Diane Dietzen, MD HMDC FACPFAAHPM, Baystate Medical Center, Springfield, MA.

Objectives1. Explain the process of incorporating a nurse

triage role into an inpatient Palliative Careconsultative service team.

2. Articulate benefits to the Palliative Care team aswell as the hospital system in instituting a nursetriage role.

3. Differentiate replicable benefits to other hospitalsystems and the financial implications of thismodel.

Background. The Inpatient Palliative Care team at a700+ bed academic medical center had been clini-cian led for years. The consults were placed bypaging the consult pager and were therefore triagedby clinicians seeing patients throughout the day. Thiswas inefficient and was identified as a contributor toburnout.Aim Statement. The nurse triage role was proposedas a means of improving the efficiency of the teamto see more patients, have time for more family meet-ings, and improve the wellbeing of the clinician teammembers.Methods. Palliative Care Quality Network (PCQN)data was used for number of consults seen and familymeetings, looking six months before initiation ofNurse Triage and six months after. Palliative Careteam members shared their written reflections onhow adding this team member improved their qualityof life at work.Results. Number of new consults per monthincreased from 66.5 patients to 81.6 patients. Averagenumber of family meetings per day increased from 1.1to 1.2. The prevailing themes shared by all interdisci-plinary team members in their narratives were: TheNurse Triage improved the coordination of care,improved communication between palliative teammembers and also with the referrers, and allowed




clinicians to focus on taking care of their patientswithout distractions. In turn, this improved the workenvironment for all team members, even non-clinicians.Conclusions and Implications. Though this datadoes not take into account the fluctuating FTE ofthe providers on the team, there is a trend to improve-ment in both quantitative parameters after initiationof the Nurse Triage role. Qualitative data was resound-ingly positive in terms of improved wellbeing andpride in the quality of the work done by the team.Nurse Triage is an effective possibly replicable modelto improve delivery of care in Palliative Care inpatientconsult teams.

Open-Ended Responses to BereavedSurveys: Best Practices from the VeteransHealth Administration and KaiserPermanente (QI710)Randall Gale, DrPH, VA Palo Alto Healthcare System,Palo Alto, CA. Karleen Giannitrapani, PhD, VA HealthServices Research and Development Service/Stanford,Palo Alto, CA. Caroline Gray, PhD, VA Palo AltoHealth Care System, Menlo Park, CA. Luisa Manfredi,JD MPH, VA Palo Alto Health Care System, MenloPark, CA. Joy Goebel, PhD RN FPCN, California StateUniversity Long Beach, Long Beach, CA. MargaretWang, PhD MPH, Kaiser Permanente, Oakland, CA.Daniel Johnson, MD FAAHPM, Kaiser Permanente,Aurora, CO. Karl Lorenz, MD MS MSHS, Stanford/VA Palo Alto Health Care System, Palo Alto, CA.

Objectives1. Recognize opportunities to collect and/or use

existing qualitative self-reported data in theirpractice settings for quality improvement.

2. Discuss how self-reported data can be used toeducate administrators and non-palliative careclinicians about end-of-life care.

Background. The U.S. Veterans Health Administra-tion’s (VA) Bereaved Family Survey (BFS) is admin-istered to the next-of-kin of Veterans who die in VAhospitals; Kaiser Permanente (KP) recently pilotedan abridged version of the BFS, adapted items forits members across inpatient and outpatient set-tings. Narrative responses to 2 BFS open-endedquestions informed identification of best end-of-lifepractices.Aim Statement. We compared VA and KP BFS narra-tive responses to identify best care practices across2 large integrated delivery systems.Methods. Content analysis of: Is there anything else thatyou would like to share about either:

1. The patient’s care during his/her last month of life?2. How the care could have been improved for the patient?

Results. A sample of responses to VA’s and 1,463 re-sponses to KP’s open-ended questions were reviewedto identify best practices. Responses confirmed thequantitative BFS structured content (i.e., multiple-choice items) and generally supported its domains.However, unique processes of care emerged. Forexample, using music therapy to calm and sooth Veter-ans was identified in the VA sample. Data suggests op-portunities and specific approaches for improvingquality of life at the end of life. Other processes ofcare to emerge from both data sets included frequentand timely updates to family and loved ones on pa-tients’ clinical status as death nears. Among KP re-sponses, it was noted that families appreciate morefrequent and timely interaction with clinicians withrespect to early information sharing and dialogueabout end-of-life process, what to expect, and howthey can help their loved one.Conclusions and Implications. Analysis of qualita-tive data affirmed the domain structure and compre-hensive nature of the BFS. It also provided uniqueinsights into best end of life care practices.

Department of Veteran Affairs Gold StatusPracticedAdvance Care Planning UsingGroup Visits (QI711)Kimberly Garner, MD MPH CHCQM, Central Arkan-sas Veterans Healthcare System, Little Rock, AR. JamieJensen, LCSW, Central Arkansas Veterans HealthcareSystem, North Little Rock, AR. Lisa Nabholz, MSNRN, Central Arkansas Veterans Healthcare System,North Little Rock, AR. Laura Taylor, LSCSW, VeteransHealth Administration, Washington, D.C. Darlene Try-tek, MSW, Veterans Health Administration, Canandai-gua, NY. Christy Husmann, MSW, VeteransAdministration, White River Junction, VT.

Objectives� Describe key components of advance care plan-ning group visits, including social worker andteam-based facilitation, clinic-based needs, andstaffing resources.

� Obtain person-centered tools and communica-tion skills to successfully facilitate advance careplanning group visits.

Background. Advance care planning (ACP) is ahealth behavior that requires person-centered educa-tion, support by trained professionals and motiva-tional strategies to promote goal-setting and actions.Group visits in the healthcare setting can effectively in-crease an individual’s knowledge, motivation and self-efficacy.Aim Statement. This session presents a best practiceand lessons learned from implementation of groupvisits focused on ACP.




Methods. To improve care delivery for our nation’sVeterans, the Department of Veterans Affairs (VA)developed the Diffusion of Excellence Initiative toidentify and spread practices developed through qual-ity improvement methods. One such practice isAdvance Care Planning via Group Visits (ACP-GV),which uses an interactive and patient-centered groupsession to engage Veterans in thinking about and plan-ning for future medical decisions. In these sessions,social workers, or other health professionals, facilitatediscussions for Veterans and their trusted others. Facil-itators emphasize that while completing an advancedirective is voluntary, it increases the chance that theircare aligns with their wishes and values and relievestrusted others of having to make these difficult deci-sions. In addition, ACP-GV increases the effectivenessof advance care planning through allowing Veteransto discuss and process these complex topics with otherVeterans in a group session.Results. To date, 34 VA Medical Centers (VAMCs)have adopted the ACP-GV practice and more than10,250 Veterans have attended ACP-GV sessions. Ofthose participants, approximately 18-20% developeda new advance care directive within one month ofthe session, and 86% set a smart goal to take addi-tional steps toward advance care planning. Continuedrollout of this innovative practice to VAMCs across thenation is ongoing.Conclusions and Implications. At the conclusion ofthe session, attendees will have practical guidance,techniques and tools for implementation of ACP dis-cussions using group visits in integrated (VA) or fee-for-service (Medicare) outpatient settings.

It’s Everyone’s Business: Capturing theConversation (QI712)Roshni Guerry, MD, Christiana Care Health System,Newark, DE. Darcy Burbage, MSN, Christiana CareHealth System, Newark, DE. Kristen Isaac, MPH,Christiana Care Health System, Newark, DE.

Objectives1. Describe an innovative way to increase goals of

care communication between providers.2. State the required components of goals of care

discussions.Background. Results of research studies show thatclinicians typically avoid discussing goals of care(GoC) and prognoses with patients. However, in or-der for patients facing serious illness to receive thecare they want that is consistent with their valuesand wishes, health care providers must be skilledat challenging conversations. Currently, the GoCdocumentation is variable between generalist pro-viders leaving the Palliative Care clinicians unclearas to if the discussions took place or what was

understood regarding quality of life goals. Becauseof this, a standardized GoC form was implementedin the EHR to help facilitate communication be-tween clinicians that would be accessible for subse-quent admissions and sudden changes in thepatient’s condition.Aim Statement. The purpose of this quality improve-ment project is to improve communication, collabora-tion and decision making about GoC betweenclinicians, patients and family members.Methods. The current standard of care is for clini-cians to review GoC with patients upon admissionand to document them in the GoC section of theEHR utilizing specific criteria. After an education ses-sion to all clinicians regarding the essential informa-tion to be included, GoC discussions were reviewedfor all palliative care consults and rated as good, inter-mediate, or poor. Monthly standardized e-mail mes-sages are sent to providers acknowledging gooddocumentation as well as to offer assistance to improvediscussion and documentation.Results. Good GoC discussions increased by over25% and patients with no GoC discussion decreasedby over 20% during the initial study intervention. Asecondary analysis of individual provider results is inprogress.Conclusions and Implications. Providing feedbackto clinicians helped to improve GoC discussions anddocumentation in the EHR. Additional recognitionas a GoC ambassador was sent to the managers ofthose clinicians who consistently performed at ahigh level. By educating providers regarding how tohave difficult discussions surrounding GoC documen-tation increased leading to care that aligns with the pa-tient’s wishes.

Capturing Wishes: A Novel Approach toGoals of Care Documentation for InpatientPalliative Care Consults (QI713)Brieze Keeley, MD, University of California, San Fran-cisco, San Francisco, CA. Jessi Humphreys, MD, Uni-versity of California, San Francisco, San Francisco,CA. Jennifer Olenik, MD, University of California,San Francisco, San Francisco, CA. Laura Schoenherr,MD, University of California, San Francisco, San Fran-cisco, CA. Giovanni Elia, MD FAAHPM, University ofCalifornia, San Francisco, San Francisco, CA.

Objectives1. Identify the need for improved goals of care

documentation in the inpatient electronic medi-cal record among patients seen by an inpatientpalliative care service.

2. Articulate a strategy for improving goals of caredocumentation among patients seen by an inpa-tient palliative care service.




Background. Patients with life-threatening illnessface critical decisions regarding goals of care (GOC)and treatment preferences (TP). Palliative Care (PC)consultants play a pivotal role in clarifying and docu-menting patients’ wishes to ensure they receivegoals-aligned care. However, inconsistencies in docu-mentation of GOC/TP in the electronic medical re-cord (EMR) can result in patients receivinginterventions incongruent with their preferences.Among a baseline sample of patients seen by the Uni-versity of California San Francisco (UCSF) PC servicefor GOC discussions in April and May 2017, only63.3% had any goals documented in the EMR, andonly 54.5% had both overall GOC (e.g., curative)and at least one specific TP specified in addition tocode status (e.g., artificial nutrition, dialysis, etc).This quality improvement study examined the impactof an easily accessible, highly reproducible EMR notetemplate on the consistency of GOC/TP documenta-tion for patients seen by the UCSF PC service.Aim Statement. This study aimed to increase docu-mentation of GOC/TP among patients seen forGOC by the UCSF PC service from 54% to 80% withthe use of a note template designed to integrate intothe Advance Care Planning (ACP) problem in Epic.Methods. Study authors designed an Epic note tem-plate to facilitate consistent documentation of GOC/TP. The PC service encouraged routine use of thedot phrase by all consulting PC physicians. Analysts as-sessed compliance at monthly intervals.Results. Among 640 patients seen by the UCSF PCservice between September 2017 and May 2018, 466(72.8%) were seen for GOC. Of these, 461 patients(98.9%) had documentation of both overall GOCand at least one TP in their ACP problem in Epic.Conclusions and Implications. Implementation ofan EMR note template increased consistency andclarity of GOC/TP documentation for patients seenby the PC consult service.

Reducing Medication Errors in HomeHospice to Improve Patient Safety (QI714)Trisha Kendall, RN, Gilchrist Services, Hunt Valley, MD.Carson Savage, BSN RN CHPN, Gilchrist Hospice Care,Hunt Valley, MD. Kimberly Schonfeld, RN CHPN, Gil-christ Hospice, Hunt Valley, MD. Lauren Garlow, RNCHPN, Gilchrist Services, Hunt Valley, MD.

Objectives1. Reduce medication errors in the home hospice

setting by focusing on performance improve-ment efforts in nurse practice and patient/care-giver practice.

2. Develop patient- and family-centered educationaltools for safe medication practices in a home hos-pice setting.

Background. Gilchrist Hospice Care’s average dailycensus is 433 in the home care setting, representing49% of the total served. In FY15, our medication errorrate was 0.61, with 55% in home care. A MedicationSafety Team uncovered root causes and focused effortson nurse centered medication reconciliation practicesand patient/caregiver education.Aim Statement. To reduce medication errors inhome hospice service by 20% by FY18.Methods. The Medication Safety Team includes ourhome care Medical Director, Clinical Nurse Specialist,and nurses from home care, triage, admissions, and af-ter-hours teams. Using the IHI model, the team devel-oped new educational tools for medication safety:Syringe Tool, Medication Record, PRN MedicationTracker. Feedback about the ease of use during a crisisto prevent medication errors was collected from pa-tients, family caregivers, and clinicians. The Patientand Family Advisory Council consulted on the visualrepresentation of the tools. A double check processwas initiated in the new First Dose Protocol, providingfamilies 24/7 access to our nurse help line. Our Med-ical Director trained nurses in polypharmacy, toimprove crucial conversations about reducing thenumber of medications taken.Results. The nurses and family members surveyed re-ported over 90% satisfaction with use of new educa-tion tools. The FY18 medication error rate is 0.34,representing a 44% decline over a three-year period.The errors in home care decreased 9% during thesame period.Conclusions and Implications. The new tools areintegrated into the hospice Caregiver Handbook andprovide cues about when to administer medications,how to safely check dosing, and provide clinicians aclear picture of medication usage between visits. Thedouble check process in triage has led to countless‘great catches’. These simple improvements to nursepractice and patient education have made a lastingimpact at the frontline of care to improve patientsafety and overall caregiver confidence.

Nursing Telephonic Intervention to ReduceNo-Show Rates for Outpatient OncologicPalliative Care (QI715)Arum Kim, MD FAAPMR, New York University, NewYork, NY. Akash Shah, MD, New York University, NewYork, NY. Deetta Vance, DNP, New York University,New York, NY.

Objectives1. Illustrate the potential value of a nursing tele-

phonic intervention to reduce no-show rates toan outpatient palliative care practice.

2. Identify care coordination needs of patients withserious illness that may be addressed through




telephonic follow-ups to potentially decrease EDutilization and hospitalizations.

Background. Missed appointments for palliative careclinics are reportedly as high as 50-70% and may beprevalent in cancer patients due to care coordinationconfusion, late referrals, and time burden. In our ur-ban academic cancer center, the no-show rate in2017 was 38% despite automated appointmentreminder alerts. This results in inefficient utilizationof limited palliative care resources.Aim Statement.

� Decrease no-show rates in our oncologic palliativecare clinic through implementation of a nursingtelephonic intervention.

� Provide follow up telephone calls to identifysymptom management and care coordinationneeds.

Methods. Patients were called the day before theirscheduled appointment to introduce the service fornew referrals, identify needs, and coordinate sched-uling changes. Follow-up calls were provided oneweek after the patient encounter to monitor interven-tions and triage for needs at home. No-show patientswere called to identify potential barriers to the visit.Results. A total of 408 patients, 202 new and 206follow-ups, were scheduled for palliative care appoint-ments over the course of 3 months. The mostcommonly represented oncologic disease groupsincluded GI and thoracic malignancies. 329 patientcalls were attempted, and 252 patients were reached.40% (n¼133) of the patients reached reported needsthat were addressed during these calls, ranging fromprescription problems, symptom management, andpatient education. No-show rates during our interven-tion period decreased to 19%. The most common rea-sons for no-shows included last minute patientcancellation and illness/hospitalizations.Conclusions and Implications. Our telephonicnursing intervention reduced no-show rates by 50%as compared to automated appointment reminderalerts. 40% of patients reached reported needs athome that were addressed, possibly reducing ED visitsand hospitalizations for pain and symptoms. This pre-liminary intervention can have implications forimproving quality of care and more efficient utiliza-tion of limited outpatient palliative care resources.

Got WiFi? Exploring the Feasibility ofTelevisits Among Vulnerable Patients(QI716)Anne Kinderman, MD, Zuckerberg San FranciscoGeneral Hospital/University of California, San Fran-cisco, San Francisco, CA. Heather Harris, MD, Zucker-berg San Francisco General/University of California,San Francisco, San Francisco, CA. Christopher

Strayhorn, MD MDiv, Zuckerberg San Francisco Gen-eral/University of California, San Francisco, San Fran-cisco, CA.

Objectives1. Appraise existing literature regarding the imple-

mentation of telehealth palliative care, withattention to the patient populations studied.

2. Review the methods, findings, and experiences ofa pilot telehealth initiative at an urban, safety netpalliative care clinic.

3. Reflect on lessons learned and formulate ap-proaches to overcoming barriers to completingtele-visits, for vulnerable patients.

Background. No-show rates for palliative care clinicvisits can be high (21-36% in our setting), frequentlydue to severe illness, limited transportation, or fatiguefrom multiple appointments. Telehealth visits mayexpand access to community-based palliative care forthese complex patients.Aim Statement. To explore the feasibility of tele-health visits with vulnerable patients in an urban,safety net palliative care clinic.Methods. We developed a brief technology access sur-vey to assess patients’ access to email and necessarytechnology, and experience with videoconferencing.The survey was professionally translated into Spanishand Chinese. All patients who completed in-personclinic visits between November 2017 and May 2018were eligible to participate. Patients who reported ac-cess to the necessary technology were offered trainingto complete televisits.Results. During the study period, 109 patientscompleted in-person visits. Patients were 26% Latino,24% Asian/Pacific Islander, 21% African American,and 20% Caucasian; 34% had Limited English Profi-ciency, and 10% were either marginally housed orhomeless. 89 patients (82%) completed the survey.60 patients (67%) reported access to a smartphone,tablet, or computer and were screened for eligibility.18 (20%) were deemed ineligible due to significantsensory impairment, cognitive impairment, or limitedtechnology experience. Of the 42 patients eligible fortelevisits, only 11 (26%) accepted training for televi-sits. The most common reasons patients declinedwere preference for in-person visits and lack of WiFiaccess. 5 patients were scheduled for televisits, and 4completed visits (4%). The no-show rate for televisitswas lower than the general clinic during the sametime period (20% vs. 26%).Conclusions and Implications. In our setting,likely 10-20% of patients have the access, capability,and interest to attempt televisits. More study isneeded to determine whether these patients no-show at lower rates for televisits compared to in-person clinic visits.



PC-PAICE Palliative CaredPromotingAccess & International Cancer Experiencein India Collaborative (QI717)Tayler Kiss-Lane, BS, Stanford Health Care, Palo Alto,CA. Karl Lorenz, MD MS MSHS, Stanford/VA PaloAlto Health Care System, Palo Alto, CA. Michelle De-Natale, MSc, Stanford Health Care, Palo Alto, CA.Jake Mickelsen, MBA, Stanford Health Care, Stanford,CA.

Objectives1. Describe the application of online methods to

teach Quality Improvement.2. Describe key principles of fostering successful in-

ternational collaboratives.Background. India has a burgeoning elderly popula-tion, and limited palliative care (PC). Using a LEAN-based quality improvement (QI) curriculum, US &Australian institutions (Stanford, Duke, UCSF, JohnsHopkins, UTS Sydney, Peter MacCallum) partneredwith Indian PC & cancer centers (AIIMS, CIPLA,Homi Baba, MNJ, Tata Memorial, Thrissur, Trivan-drum) 10/17-6/18 to improve care.Aims. We used PC-PAICE to train a cohort of Indianleaders to promote improvement in PC capacity andperformance. Participants were recruited from mem-bers of the Indian Association of Palliative Care, andfaculty of US and Australian sites with PC oncology ser-vices were recruited as site coaches. PC-PAICE 2018-9is transitioning to a US-Indian leadership model withStanford as resource hub and Indian maintenance ofcoaching, education, and team relationships.Methods. In addition to monthly online curricula,group, and site coaching, all teams met in person inFebruary 2018 at Delhi IAPCON for didactics, teambuilding, and problem solving. PC-PAICE curriculaencompassed baseline states, problem descriptions,outcomes measurement, intervention deployment,sustainability, and QI methods (e.g., run charts).Results. QI projects addressed earlier solid tumor(lung, head and neck, advanced cancer) PC refer-rals, documenting goals of care, hospital-hospice co-ordination, and improving community-based serviceallocation. 4/6 teams reached self-defined QI goalsfor the 1st cycle. On a 0-5 scale, 20/29 participantsfound project templates, key driver diagrams, in per-son workshop, monthly group WebEx, and sustain-ability plans to be most helpful; online videos,Pareto charts, internal progress scoring, Box, and a‘graduation template’ least helpful didactic features.One participant remarked, ‘‘I have always thought-here is a problem, this is the solution and jumpedright into it. PC-PAICE helped approach solutionsin a more open-minded approach through varioussteps and logic.’’

Interpretation. PC-PAICE demonstrated feasibility ofremote support to improve PC QI capacity, access andquality in India.

Online Specialty Education Impact onSocial Work Knowledge, Attitudes andSkills Regarding Palliative Care (QI718)Wendy Knowlton, DO HMDC, Sacred Heart MedicalCenter and Children’s Hospital, Spokane, WA.

Objectives1. Appraise tools to monitor distress and improve

communication.2. Support implementing goals of care (GOC)

conversations.3. Evaluate the impact of this education on work

engagement.Background. Social work is an integral part of health-care delivery especially for those with serious illness.Palliative care education can increase comfort andskills needed for healthcare social workers.Aim Statement. Examine impact of 8 month onlinecertificate program in palliative care on medical socialworkers.Methods. Forty-four social workers in the Providencesystem enrolled in online palliative care certificateprogram. Twenty-three (51%) were social workers for0-6 years and fourteen (32%) had 15+ years of experi-ence. A seventeen question, anonymous, emailed sur-vey was administered pre/ post course. Forty-threeresponded pre-course. Five participants withdrew.Twenty- eight responded post-course.Results. Social worker comfort in discussing GOCjumped from 25% pre to 75% post course and in-crease in having tools needed to monitor patientand family caregiver distress improved from 25% to67%. Work engagement improved as well as increasedsocial worker satisfaction with use of skills at work(37% pre and 57% post). Social worker perceptionof palliative care consultation as routine for patientswith medical complexity and not always associatedwith end of life care improved post course.Conclusions and Implications. Palliative care edu-cation is effective regardless of time as a social worker.Increased comfort with GOC conversations, improvedpatient, family and provider communication andincreased work engagement was noted after specialtytraining. Assessment tools allow monitoring of patientand family caregiver distress so interventions can beplanned, implemented and distress reassessed objec-tively. Improved comfort in GOC conversations allowssocial workers to operate at the top of licensureincreasing provider efficiency. Ease with GOC conver-sations increases delivery of congruent care. Socialworkers can increase provider efficiency and patient




safety through documentation of billable work and ac-curate care goals. Increased work engagement leads tosocial worker retention. Based on this descriptivestudy, healthcare systems are encouraged to exploreenrollment of social workers in such training to in-crease palliative care knowledge, attitudes and skills.

It Takes a Village: An InterdisciplinaryEffort to Improve the PAL-05 Measure ina Large Academic Medical Center (QI719)Grace La Torre, DO, Stony Brook Medicine, StonyBrook, NY. Susan Robbins, MS BSN, Stony Brook Hos-pital, Stony Brook, NY. Samantha Nagengast, MD,Stony Brook Medicine, Stony Brook, NY.

Objectives1. Review the Joint Commission PAL-05 measure

description and expectations.2. Discuss some of the barriers to meeting the PAL-

05 measure.3. Describe an interdisciplinary approach using a

step-wise method to improve PLA-05 measurecompliance.

Background. The National Consensus Project forquality palliative care (2013) guidelines recommendthat a patient’s or surrogate’s goals and treatmentpreferences are used as the foundation for the planof care. Despite recognition of the value of these con-versations by both physicians and patients, there isoften hesitancy to initiate these discussions. Accordingto the PEACE project the use of a Treatment Prefer-ences quality measure will promote discussionsregarding treatment preferences and goals betweenseriously-ill patients and their physicians. The JointCommission’s PAL-05 measure assures that documen-tation of the patient’s treatment preferences and goalsaccompanies the patient to the next level of care. JointCommission certified palliative care programs, howev-er, have found it challenging to develop a feasible andeffective approach to meet this measure.Aim Statement. Utilizing an interdisciplinaryapproach along with a step-wised method is crucialto the successful development and execution of thePAL-05 measure Initiative.Methods. In order to integrate the important PAL-05quality measure in our large academic medical center,we developed an interdisciplinary committee encom-passing representatives from the palliative care service,nursing leadership, social work, information technol-ogy, and quality measure. Several steps were taken,including monthly meetings scheduled over thecourse of six months, to discuss design, construction,and education of the PAL-05 measure initiative. Thisinitiative encompassed three phases. 1) design andconstruction; 2) education and training; 3) assessmentand impact.

Results. Data regarding compliance of the PAL-05measure was collected at the one month and threemonth mark. User feedback regarding the practicalityand effectiveness of this initiative was assessed via a sur-vey. Data collected showed marked increased in PAL-05 measure compliance, as well as the impact andfeasibility of this initiative.Conclusions and Implications. Our approach high-lights the importance of collaborating with other disci-plines, especially when dealing with the complexity ofcaring for seriously-ill and end-of-life patients.

Opioid Screening & Monitoring in thePalliative Care Clinic: An Attempt atStandardization (QI720)Tariq Lateef, MD, Mercy HealtheSt. Vincent MedicalCenter, Toledo, OH. Chad Hines, MBBS, MercyHealtheSt. Vincent Medical Center, Toledo, OH. JulieStausmire, MSN RN APRN CNS, Mercy HealtheSt.Vincent Medical Center, Toledo, OH.

Objectives1. List and recognize components of opioid

monitoring.2. Devise and implement his/her own screening

and monitoring plan for the Palliative CareClinic.

Background. The Centers for Disease Control (CDC)recommend Urine Drug Testing (UDT) for managingnon-palliative care patients on chronic opioid therapy(COT)1. There are no guidelines for the care of palli-ative care patients receiving (COT). Studies haveshown that in some cohorts roughly half of cancer pa-tients receiving palliative/supportive care haveabnormal UDT, suggesting that even palliative care pa-tients misuse/abuse opioids, and that the UDT mayserve as an important tool in uncovering thismisuse/abuse. Our Palliative Care Clinic (PCC)demonstrated inconsistent use of UDT testing andinaccurate ordering of the appropriate UDT for ourpatient population. Lack of a formal approved Mercyprotocol puts the Palliative Care Clinic at risk of beingdeemed as subjectively selecting random patients fordrug testing. This could be perceived as unfair or un-equal treatment of patients from different color, eco-nomic backgrounds or gender identity.Aim Statement.

1. Develop/implement a new policy for standardiza-tion of UDT in the PCC.

2. Increase provider compliance ordering the cor-rect UDT.

3. Increase patient/family education and compli-ance with opioid policy, risk assessment and useof opioid contracts.

Methods. We utilized the IHI Model for Improve-ment (AIM/Plan-Do-Study-Act).




Results. Cycle 1 investigated baseline clinical practiceas we had no existing policy for UDT in the PCC. Ourinstitution has 4 different UDT’s and we frequently or-dered the wrong test. Cycle 2 we developed and pre-sented a policy that was accepted by hospitaladministration. We provided a staff educational ses-sion for policy compliance, correct test codes, opioidcontract compliance, and physician follow-up ofUDT results. Cycle 3 we provided patient/family edu-cation and tracked 100% compliance with opioid con-tract use, correct UDT, policy compliance andphysician follow-up of UDT results. Cycle 4 is in pro-cess with initiation of opioid risk assessments andcontinued maintenance of previous interventions.Conclusions and Implications. Physician moni-toring of opioid therapy is essential from legal, regula-tory and professional mandates in the palliative caresetting.

The Implementation of a Palliative CareIntervention During InductionChemotherapy for Patients with AcuteMyeloid Leukemia (AML) at a CommunityCancer Institute (QI721)Annamaria Lattanzio-Hale, BSN RN OCN, LancasterGeneral Health, Lancaster, PA. Shanthi Sivendran,MD, Lancaster General Health-Penn Medicine, Lititz,PA. Caitlyn McNaughton, MA MHS PA, LancasterGeneral HealthePenn Medicine, Lancaster, PA.Shyam Balepur, MD, Lancaster General HealthePennMedicine, Lancaster, PA. Miriam Ryan, MSN CRNP,Hospice & Community Care, Lancaster, PA. Peter Ju-pin, MS, Lancaster General Health, Lancaster, PA.Wendy Wilson, LSW, Hospice & Community Care,Lancaster, PA. Heather Nelson-Przywara, LCSW, AnnB. Barshinger Cancer Institute, Lancaster, PA.

Objectives1. Recognize the benefits of upstream palliative

care in patients with hematologic malignancies.2. Describe multifaceted intervention to increase

palliative care utilization in patients with AMLundergoing induction chemotherapy.

3. Discuss future directions to assess uptake of palli-ative care utilization in patients with AML.

Background. Patients with hematologic malig-nancies are referred to palliative care less than pa-tients with solid tumor malignancies. Clinical trialsare underway at academic centers exploring earlyinpatient palliative care for patients newly diagnosedwith AML receiving induction chemotherapy. Feasi-bility of such interventions have not been studied ina community setting.Aim Statement. We structured a multi-faceted inter-vention for our community hematology and palliativeteam on the benefits of early palliative care in

hematologic malignancies with the aim to increaseutilization.Methods. In 2017, 24% of patients with AMLadmitted to Lancaster General Hospital for inductionchemotherapy received an inpatient palliative careconsultation. Needs assessment performed on the he-matology and palliative teams demonstrated a needfor integration of palliative care into clinical pathways,automatic triggers for consultation, a need forincreased education on AML, and creation of stan-dard elements in inpatient consultations.Results. In the spring of 2018, we integrated pallia-tive care into institutional AML clinical pathwayssuch that all patients receive inpatient palliativeconsultation within 72 hours of admission. A consulta-tion trigger was placed within the admission order setand is currently being incorporated into the chemo-therapy treatment protocol. An educational programon AML was created and presented by the hematologyteam to the inpatient palliative team. This interven-tion included creation of minimum standards forinpatient palliative consultation including providerassessment within 72 hours of referral, two visits perweek by a provider, and a minimum of one visit bythe interdisciplinary team during admission with afocus on symptom management, psychosocial assess-ments, and advanced care planning.Conclusions and Implications. We have demon-strated feasibility of creating standards for early pallia-tive care intervention in AML patients withcollaboration from the hematology and palliativeteams. Next steps will include assessing whether theseinterventions increase uptake of palliative care utiliza-tion and specific patient outcomes.

Increasing Outpatient Palliative CareReferrals in a Veteran’s Affairs HospitalOncology Clinic: A Quality ImprovementProject (QI722)Jessica Ma, MD, Duke University Hospital, Durham,NC. Amber Deptola, MD, Washington University inSt. Louis, St. Louis, MO. Eric Knoche, MD, VeteransAdministration, St. Louis, MO.

Objectives1. Describe a quality improvement project

regarding increasing outpatient palliative careutilization among advanced cancer patients.

2. Describe different PDSA cycles that can affectoutpatient palliative care referrals.

Background. The American Society of ClinicalOncology recently published guidelines recommend-ing palliative care (PC) to patients with advanced stagecancer. PC referral and involvement may provideimproved goals of care discussions and planning fortransitions to end of life care.




Aim Statement. To increase referrals from oncologyclinic to palliative care to 60% for patients withadvanced solid tumor malignancies.Methods. A PC screening tool was used to identify pa-tients with advanced solid tumor malignancies withother poor prognostic factors seen in oncology clinicfrom October 1 to November 17, 2017. Patients werefollowed over 6-months. Oncology clinicians, nurse co-ordinators, and social workers were educated in theuse of the screening tool. Additional PDSA cyclesincluded individual provider feedback on PC referralrates and integration of PC referrals to clinician pay-for-performance in 2018 fiscal year.Results. Among 559 unique patient encounters, 174patients met PC screening criteria. 33% of these pa-tients had been referred previously to PC, primarilyby oncology, other providers (ie GI, ENT, RadiationOncology), and inpatient medicine. At 6-month followup, 51% of patients had a PC referral, with new consultsplaced primarily by inpatient medicine, then oncologyproviders. 25% of patients had died and 26% were onhospice at 6 months. 42% of patients who died wereon home hospice and 84% were seen by palliativecare prior to death. Palliative care consultation was asso-ciated with hospice referral prior to death (p<0.01).Conclusions and Implications. Oncology patientswith advanced solid tumor malignancies are notcurrently referred to palliative care in congruencewith current guidelines. Outcomes data on the resultsof a screening tool to increase these referrals showedminimal improvement by oncology providers over 6-month period. Further long-term analysis is necessaryto evaluate the effect of pay-for-performance on PC re-ferrals for advanced cancer patients.

Supportive Cardiology QualityImprovement Project: IdentifyingSymptomatic, Advanced Heart FailurePatients for Palliative Care Consultation(QI723)Katie Marchington, MD, University Health Network,Toronto, ON, Canada. Warren Lewin, MD, TorontoWestern Hospital, University of Toronto, Toronto,ON, Canada. Adassa Wilson, MA BScN RN, UniversityHealth Network, Toronto, ON, Canada.

Objectives1. Reflect on one possible barrier faced by health-

care providers in documenting goals of care dis-cussions for in-patients with heart failure.

2. Describe the benefits and drawbacks of using thenumerical rating scale (NRS) for dyspnea as ascreening tool for symptomatic patients withheart failure.

Background. Heart failure is a life-limiting illnessand the leading cause of hospitalization in Canada.

The Toronto Western Hospital Palliative Care Consul-tation Service identified an opportunity for improve-ment when it noted only 9 in-patient cardiologyreferrals for palliative consultation were receivedfrom January to September 2017 despite 472 cardiol-ogy in-patient admissions.Aim Statement. The aim of this quality improvementproject was to increase the number of palliative consul-tations completed for cardiology in-patients with symp-tomatic, advanced heart failure by 50% compared tothe previous year, starting in February 2018. The projectalso aimed to document a numerical rating scale (NRS)score for dyspnea for 100% of patients at the time ofinitial consultation and last palliative care visit anddocument goals of care discussions for 100%of referredpatients by time of discharge.Methods. A process map identified an opportunityfor screening cardiology in-patients and screeningcriteria were developed: nurses identified patientswith a NRS for dyspnea of > 3/10; a palliative carephysician then attended cardiology interdisciplinaryteam rounds once weekly to identify symptomatic pa-tients with advanced heart failure and approachedthe cardiology physicians for referral.Results. Results from February to June showed an in-crease in total cardiology referrals for palliative consulta-tion from 3 (2017) to 19 (2018), with 16 referralsreceived at team rounds. Though all referrals receivedwere appropriate for palliative consultation, only 6 of 19referrals screened positive using screening criteria. Onlyten out of 19 patients were able to report dyspnea usingaNRS. Five ofnineteenpatients referredhaddocumenta-tion of goals of care discussions prior to consultation,compared to 18 out of 19 patients post-consultation.Conclusions and Implications. Refinement ofscreening criteria and criteria used to identify symp-tomatic patients is ongoing. This project may expandto other in-patient services and provided evidence tosupport teaching serious illness communication skillsto cardiology fellows.

Evidence-Based Triggers: IncorporatingPatient-Reported Outcomes (PROs) intoPalliative Care Referrals (QI724)Celine Marquez, MD, University of California, SanFrancisco Medical Center, San Francisco, CA. MichaelRabow, MD FAAHPM, University of California, SanFrancisco, San Francisco, CA. Yun Li, MD PhD, Uni-versity of California, San Francisco, San Francisco,CA. Laura Esserman, MD MBA, University of Califor-nia, San Francisco, San Francisco, CA.

Objectives1. Describe how to develop, scalable, transferable

and sustainable symptom management systemsto monitor and address common cancer




symptoms and trigger referral to palliative carewith PROs.

2. Describe how to test prototypes of an integratedPROs clinical assessment pathway and under-stand who are the stakeholders and how to betatest PRO measures.

Background. Specialty palliative care (SPC) hasbeen shown to improve quality of life, reduce un-necessary healthcare utilization, and decrease mor-tality in patients with advanced cancer. Despitecalls for universal palliative-oncologic co-manage-ment for people with metastatic disease, only a smallproportion of such patients are appropriately andpromptly referred.Aim Statement. Test hypothesis that improvingdocumentation, availability, tracking, and transpar-ency of PROs in oncology visits will increase SPCreferrals.Methods. 74 English-speaking patients seen at leastonce by oncology in a breast cancer clinic were invitedand agreed to participate in the study. Patientscompleted the electronic, validated PROMIS andPRO-CTCAE PRO questionnaires in the waitingroom. Results were presented to the oncologist duringthe encounter if PRO-CTCAE score exceeded 3 ("se-vere") in at least one domain. Data was also gatheredvia chart review and patient/provider interviews.Results. At baseline, 9 (12.2%) patients reported se-vere anxiety; 17 (23.0%) severe pain; and 27(32.1%) severe fatigue. At study entry, 25 (33.8%)already had been referred to SPC; 19 were seen(76% of those referred; 25.7% of all participants). 63(85.1%) had complete data. Among these, 13(20.6%) reported at least ‘‘severe’’ in > 2 or morecomponents of the CTCAE domains, 6 (46.2%) ofwhom had been previously referred to SPC. The 7 re-maining patients with severe symptoms never receiveda referral to SPC. Oncologists reported that referralswere limited by concern of negative impact to doc-tor-patient relationship and challenge of having thisconversation during a time-constrained oncologyencounter.Conclusions and Implications. While oncologistreferral has been a traditional mechanism throughwhich patients can access SPC services, the decisionto refer is complex and subjective. Referrals triggedby PRO have the potential benefit of accessing theproven survival benefits of PROs and rationalizing pa-tient identification for SPC.

An Innovative Tuck-In Program toImprove Service Delivery in HospicePatients (QI725)Laurl Matey, MSN CHPN, Hospice of Southern Maine,Scarborough, ME. Susan Mason, BSN RN, Hospice of

Southern Maine, Scarborough, ME. Daryl Cady, MBA,Hospice of Southern Maine, Scarborough, ME.

Objectives1. Demonstrate an innovative delivery of hospice

services to anticipate care needs in a ruralenvironment.

2. Manage resource delivery to anticipate patientstatus, current and future needs, to avoid servicefailures and improve continuity of care.

Background. Hospices face unique challenges,including broad service areas, and, in Maine, extremeweather conditions. Patients may express frustrationand anxiety if supply and medication needs are notwell anticipated, requiring urgent visits to providethem. Deploying weekend/overnight staff for visits un-associated with acute symptom management misdi-rects resources.Aim Statement. The quality initiative will result inanticipation and resolution of supplies, medications,and visit planning to improve continuity of care deliv-ery to hospice home care patients, and to decrease off-shift requests.Methods. Historical weekend/night triage data anal-ysis showed 45% of calls requested urgent medica-tions, supplies, or visit planning. A volunteer ledprogram uses a scripted Tuck-In questionnaire tocall patients several days before each weekend andanticipated weather event. Urgencies are prioritizedand forwarded to an RN for resolution. All calls are re-viewed each service day for quality assurance.Results. Immature data demonstrates an 18% declinein medication, supply, and visit planning requests.Acute symptom management data remain consistentwith pre-initiative data. Unanticipated benefits includecapture of additional clinical data useful in early inter-vention, and longitudinal data trending nurse perfor-mance in anticipating patient needs.Conclusions and Implications. Tuck-In programsfor hospices provide an additional evaluation point,review of service satisfaction, and early identificationof patient needs. Staff travel burden is reduced, andadditional nursing performance measures can bemonitored.

Integrating COMFORTTMSM at aComprehensive Cancer Center (QI726)Jennifer McClennon, MSN RN CNL, H. Lee MoffittCancer Center, Tampa, FL. Diane Portman, MDFAAHPM, H. Lee Moffitt Cancer Center, Tampa, FL.Sarah Thirlwell, MS MSN RN AOCN CHPN CHPCA,H. Lee Moffitt Cancer Center, Tampa, FL.

Objectives1. Describe use of COMFORT Communication Cur-

riculum to promote primary palliative careamong oncology nurses.




2. Recognize the benefits of role play for communi-cation based learning.

Background. While some nurses receive funda-mental communications skills training, opportunitiesexist to strengthen palliative communication skillsamong nurses in cancer care. At our NCI-designatedcomprehensive cancer center, we sought to adoptthe COMFORTTMSM Communication Curriculum tobroaden oncology nurses’ competencies to engage indifficult conversations in routine practice across allsettings. Developed through research, the curriculumincludes communication skills-building sessions andprovides participants with a communication toolkit.Aim Statement. To improve nurses’ patient centeredcommunication across the cancer trajectory.Methods. Participants in the NCI-funded COM-FORTTMSM Communication for Oncology Nurses professionaltrainingprogrampartneredwithourdepartments of Sup-portiveCareMedicine andNursingProfessionalDevelop-ment. A COMFORTTMSM Team was established todevelop and deliver a curriculum for our cancer center.The target audience included nurses and nursing/medical assistants throughout all clinical environ-ments. The program consists of seven 1.5-hour classes,including lecture and role-play to enhance skills. Eachclass dedicated time for debriefing and interactive dis-cussions to address barriers and concerns related tointegration of COMFORTTMSM techniques into nurses’busy practices. To foster engagement, the curriculumwas offered as part of Nursing Grand Rounds, as anelement of progression within the nursing Clinical Lad-der, and by linking COMFORTTMSM with our founda-tional nursing theory (Duffy’s Quality Caring Model).Results. To date, 74 individuals from various clinicalareas have participated in the COMFORTTMSM Curric-ulum, including registered nurses, advance practicenurses, care coordinators, clinical educators, and med-ical assistants. Results from the C-COPE survey indi-cate attendees’ comfort with palliative conversationsincreased following participation.Conclusions and Implications. We have successfullyadopted the COMFORTTMSM Communication Curric-ulum within our Nursing Grand Rounds format.Nurses from diverse clinical areas are equipped withCOMFORTTMSM techniques to communicate with pa-tients, families, and team members. Our COM-FORTTMSM Team plans to continue the currentprogram, explore inclusion in new staff orientation,and monitor impact on patient satisfaction.

Evaluation of Medication-Related QTcProlongation Risks in Patients ReceivingHospice Care (QI727)Bridget McCrate Protus, PharmD RPH BCGP CDP,Optum Hospice Pharmacy Services, Westerville, OH.

Amanda Lovell, PharmD BCGP, Optum Hospice Phar-macy Services, Westerville, OH.

Objectives1. Characterize the use of QTc prolonging medica-

tions in hospice patients.2. Recall the average QTc prolonging medication

burden is hospice patients.Background. Medications commonly used for symp-tom management in the hospice population arelinked with QTc interval prolongation, which maylead to torsades de pointes. Safety risk of QTc-prolongingmedications in this population may beunderestimated.Aim Statement. To develop a risk assessment scoringtool for QTc prolongation in hospice patients.Methods. A retrospective chart review of a nationalhospice pharmacy provider was completed. Decedentswith a cardiac-related primary hospice diagnosis andmedication claims profiled during between January1, 2018 and March 31, 2018 were included. Age, sex,and medications profiled were also collected. Chartsof decedent’s age $ 65 years were reviewed for medi-cations with a known or possible risk of QTc intervalprolongation. Independent patient risk factors andprofiled medications, based on CredibleMeds.org cat-egories of QTc interval prolongation risk, were scoredusing a modified RISQ-PATH tool. Independent pa-tient risk factors were also scored.Results. A total of 16,501 decedents were reviewedwith an average age of 87 years. The population is56.6% (n¼9,343) female. This subset of patientsscored 9 on RISQ-PATH based on independent riskfacts alone. Percentage of patients with QTc-prolong-ing drugs was 28.8% with 10.3% of the populationon a drug that CredibleMeds.org ranks as known riskof torsades. Additionally, 6.9% of patients were pre-scribed diuretics. After the initial review of populationdata, additional statistical analysis is in progress forfinal risk assessment tool development. Results ofthis analysis will be incorporated in final presentation.Conclusions and Implications. Our population re-view illustrates utilization of medications with QTcprolongation risk in a patients with several indepen-dent risk factors already present. Evaluating patientrisk for QTc interval prolongation will help prioritizesignificance of drug-drug interactions and informour discussions with interdisciplinary teams aboutthe risk vs benefits of these medications for symptommanagement.

Sleepless Nights: Trazodone Use andInsomnia Evaluation in the Palliative CareClinic (QI728)Juan Pagan-Ferrer, MD, University of Iowa Hospitaland Clinics, Iowa City, IA. Yuya Hagiwara, MD




MACM, University of Iowa, Iowa City, IA. KashelleLockman, PharmD MA, University of Iowa Collegeof Pharmacy, Iowa City, IA.

Objectives1. Discuss the evaluation of insomnia in the pallia-

tive care clinic of a large academic medicalcenter.

2. Discuss the use of trazodone in the palliative carepopulation and discuss the concerns regardingthe use of trazodone.

Background. Insomnia is a common symptom in thePalliative Care patient population with some esti-mates showing up to 70% of patients having symp-toms of Insomnia. However, patients continue todepend on pharmaceutical agents owing in part dueto lack of access to Cognitive Behavioral Therapyand longtime use of medications to treat insomnia.Although trazodone has limited efficacy forinsomnia, it is used roughly by 1% of the U.S. adultpopulation. We sought to better define within ourclinic population the likelihood patients were usingtrazodone and also the likelihood of recognition ofits use for insomnia.Aim Statement. evaluate the use and documentationof Insomnia and trazodone use in the outpatient clinicin the palliative care clinic of a large academic medicalcenter.Methods. This is a retrospective chart review assess-ing patients seen in the Palliative Care Clinic at a largeacademic center from June 2017 to November 2017.Data includes demographics, presence of trazodonein their medication list, and documentation of ESAS(date documented, and drowsiness score). Chartswere manually reviewed for any documented discus-sion of insomnia either in the problem list, discussion,or history. Descriptive statistics were used duringanalysis.Results. Fifteen (9.2%) of 163 patients had trazo-done in their medication list during the reviewperiod; 6(40%) were female with mean age of63.5. Discussion of insomnia was documented foronly 4(26.6%) patients. The mean ESAS drowsi-ness score was 5.03. The most common prescribeddose of trazodone was 50 mg (47%). We were un-able to identify thorough assessment or alternativenon-pharmacologic strategies within any of thesecharts.Conclusions and Implications. Discussion with pa-tients regarding the effectiveness of interventions tohelp treat insomnia are rare in palliative care patientsreceiving trazodone. Future interventions will focuson implementing better screening methods, patienteducation, and documentation regarding insomniaand its management.

Driving Organizational Change in SeriousIllness Communication: Successes andChallenges of Implementing the SeriousIllness Care Program (QI729)Joanna Paladino, MD, Ariadne Labs, Boston, MA. JaneKavanagh, BA, Ariadne Labs/Brigham and Women’sHospital, Boston, MA. Justin Sanders, MD MSC,Dana-Farber Cancer Institute, Boston, MA. ErikFromme, MD MSCR FAAHPM, Ariadne Labs, Boston,MA.

Objectives1. Describe different approaches to the use of clin-

ical tools, clinician training, and systems-changesto improve serious illness communicationthroughout a system.

2. Examine aggregate implementation data acrossthree health systems on structures and processesof implementing a primary palliative care pro-gram focused on communication.

3. Explore strategies to maximize clinician behaviorchange and factors that may influence uptake ofserious illness communication on the frontlines.

Background. Clinicians commonly miss opportu-nities to engage seriously ill patients in conversationsabout values and goals, or do so late in the illnesscourse. Achieving more, earlier, and better seriousillness communication requires system-level change.Aim Statement. Examine early learnings about suc-cesses and challenges of implementing a communica-tion quality-improvement initiative in three healthsystems.Methods. Three U.S. systems have adapted and im-plemented the Serious Illness Care Program (SICP),which includes tools, training, and systems-changes,in partnership with Ariadne Labs (AL) using a sys-tem-level implementation model. We combined andanalyzed structure and process data from the three sys-tems and routinely collected coaching notes to under-stand early successes and challenges.Results. From 2016-present, AL trained and coached24 champions (including 20 palliative care specialists)at the 3 systems. Champions launched SICP in cardiol-ogy, oncology, geriatrics, surgery, and primary care;Champions have trained 330 clinicians. EMRs havebeen modified to include an accessible template;1,852 patients have a documented conversation thusfar. Preliminary analysis of coaching notes revealed:1) Structures and capabilities (e.g. clinician training,EHR template) can be replicated, yet clinician uptakeof serious illness conversations varies and may dependon factors not captured in the model, e.g. attitudes to-ward palliative or end-of-life care; 2) Uncertaintiesabout the role of inter-professionals (e.g. nurses)and specialists vs. generalists in serious illness



communication may contribute to workflow chal-lenges; 3) Supportive coaching, leadership/peerengagement, and/or data-reporting are likely toenhance practice change but require time andresources.Conclusions and Implications. Successful adapta-tion and adoption of SICP structures and processesin three health systems suggests the promise of a sys-tems-level implementation model to improve seriousillness communication. More effective workflows thatactivate the care team and a better understanding ofthe mechanisms and contextual factors that supportpractice change are likely to enhance efforts.

Rapid Access Service for SymptomManagement: An Out-Patient PalliativeMedicine Clinic Initiative in a CancerInstitute (QI730)Armida Parala-Metz, MD, Atrium Health, Charlotte,NC.

Objectives1. Discuss the operation of a mobile out-patient

consult team.2. Discuss the challenges for the mobile out-patient

service.Background. The benefits of early introduction ofpalliative care in patients with advanced cancer areincreasingly recognized and integration of palliativecare in standard oncology care is recommended.Since oncology care is mostly in the out-patientsetting, access to out-patient palliative care service isimportant.Aim Statement. To share our experience with ourmobile out-patient palliative medicine consult team.Methods. In May 2017, the section of Palliative Med-icine under the Department of Supportive Oncologyat Levine Cancer Institute piloted a Rapid Access(RA) service for symptom management. This is a mo-bile out-patient service to complement the out-patientpalliative medicine clinic. The goals of the service are:1) provide immediate assistance to patients withpoorly controlled symptoms related to their cancerand/or treatments; 2) prevent unnecessary emergencyroom (ER) visits; and 3) facilitate early palliative med-icine integration in cancer care. The RA service is staf-fed by a nurse and a palliative medicine physician. Theteam sees urgent referrals for uncontrolled symptoms.Results. 183 patients were referred over 12 months.75% of the patients have solid tumor malignancies,15% have gynecologic cancers, and 9% have hemato-logic malignancies. Majority of patients referred wereseen on the same day or next day. The most commonreasons for referral are uncontrolled pain (83%) andGI symptoms (6%). Of the patients seen by the RAteam, only 4% needed to be sent to the ER.

Longitudinal follow-up was arranged in the out-pa-tient palliative medicine clinic.Conclusions and Implications. The RA access ser-vice increased out-patient palliative referrals of pa-tients with advanced cancer.

Evaluation of Goals of CareCommunication Training for MedicalOncologists (QI731)Sophia Paraschos, MD, Transitions LifeCare, Raleigh,NC. Joshua Dowd, MD, University of North CarolinaHospitals, Chapel Hill, NC. Susanna Thach, MD, Uni-versity of North Carolina School of Medicine, ChapelHill, NC. Gary Winzelberg, MD MPH, University ofNorth Carolina School of Medicine, Chapel Hill,NC. Kathryn Wessell, BA, University of North Carolinaat Chapel Hill, Chapel Hill, NC. Laura Hanson, MDMPH FAAHPM, University of North Carolina, ChapelHill, NC.

Objectives1. Develop succinct goals of care communication

skills training for oncologists.2. Design an evaluation tool for the above training

sessions.3. Recognize that such succinct training sessions

can be perceived as effective, relevant, and prac-tice-changing.

Background. Medical oncologists commonly havegoals of care (GOC) conversations with their patients,but many report insufficient formal communicationtraining. The lengthy nature of many trainingcurricula may limit participation.Aim Statement. Evaluate the effect of a 90-minuteGOC communication training for medical oncologyproviders on perceived barriers to GOC conversations.Methods. This training evaluation was conducted attheUniversity ofNorthCarolina atChapelHill. Recruit-ment targeted medical oncology providers specializingin lung, breast, and genitourinary cancers. Medicaloncology leadership participated in study planningand assisted with recruitment. Training consisted of a90-minute interactive primary palliative care skills ses-sion, offered four times to accommodate clinicians’schedules. Training consisted of elements from theAriadne Labs Serious Illness Conversation Guide, Vital-Talk, and a case-based exercise. Participants learned todocument advanced care planning (ACP) notes in theEMR. Participants evaluated the training on semi-struc-tured surveys using a 5-point Likert scale (1¼ not at all,5 ¼ very much) and open-ended questions assessingperceived communication barriers, quality and rele-vance of training, and expected effect on practice.Results. Seventeen of twenty eligible medical oncolo-gists (85%) and five of nine NPs (56%) attended thetraining and completed the evaluation. They




perceived the most significant barriers to GOC conver-sations to be ‘‘patients may be too stable to warrant aGOC conversation’’ (mean Likert scale score 3.2)and ‘‘patients may be upset by GOC conversations’’(3.0). Participants strongly agreed that training pro-vided skills applicable to clinical practice (4.8) andthat they would recommend the training (4.8). Inresponse to an open-ended question about effect onpractice, participants most often identified specificcommunication principles, including ‘‘Ask-Tell-Ask’’and ‘‘I wish’’ statements.Conclusions and Implications. We successfully re-cruited medical oncology providers to attend a 90-minute GOC communication training, which wasperceived positively and expected to change practice.To assess impact on practice we will review docu-mented GOC discussions before and after trainingfor participants.

Improving Advance Care Planning inResidency Through Annual Wellness Visits(QI732)Rushil Patel, MD, Memorial Sloan-Kettering CancerCenter, New York City, NY.

Objectives1. Describe the structure and outcomes of a project

to increase initiation of advance care planning byInternal Medicine Residents in Medicare AnnualWellness Visits.

2. Evaluate Internal Medicine residents’ perspec-tives on initiating advance care planning duringthe Medicare Annual Wellness Visit.

3. Assess the feasibility of advance care planning inthe Medicare Annual Wellness Visit.

Background. The Annual Wellness Visit (AWV)through Medicare offers an opportunity to initiateadvance care planning (ACP) in the outpatientsetting. In continuity clinics staffed by medical resi-dents, the AMV is also an opportunity for improvingtrainees’ skills for supporting ACP. We piloted anintervention to enhance residents’ initiation of ACPduring AWV.Aim Statement. Residents will initiate ACP with 1 pa-tient during an AWV.Methods. Among patients for whom they served asthe primary care physician, second and third year In-ternal Medicine residents at a university-affiliated com-munity hospital used a query of the electronic medicalrecord (EMR) to identify patients in Medicare with $1 high-risk medical condition (specified ICD-10 co-des), dependence for $ 1 activity of daily living(collected during nursing intake at last visit), and $1 hospitalization or emergency department visit inthe past year. Each resident selected one patient whomet all four criteria to address ACP within in an

AWV. At the visit, patients first reviewed Prepare ForYour Care (https://www.prepareforyourcare.org), aninteractive, online resource for ACP with a clinic pa-tient educator. The resident then discussed the pa-tient’s preparatory responses and health-relatedvalues, and documented them using an EMR tem-plate. A post-intervention questionnaire surveyed resi-dents’ views of the experience (1-7 scale, 7¼mostpositive) and suggestions for improvement. The pri-mary outcome was the number of residents who initi-ated ACP with 1 patient.Results. 10 of 12 residents initiated ACP. Most com-mon conditions were CHF requiring hospitalization,ESRD on hemodialysis, diabetes with severe complica-tions, and COPD on home oxygen. The average scores(n¼8) for overall experience, impact on patient care,and likelihood of incorporating ACP in future prac-tice were 6.0, 6.1, and 6.6, respectively. Common areasfor improvement included delays in scheduling visitsand visit duration.Conclusions and Implications. The AWV is an op-portunity to initiate ACP with high-risk patients whiletraining physicians in this important area of practice.Future improvements include streamlining clinicworkflow.

Palliative Medicine Education for InternalMedicine Residents (QI733)Gina Piscitello, MD, University of Chicago, Chicago,IL. Shannon Martin, MD MS, University of Chicago,Chicago, IL.

Objectives1. Explain the improvement that a lecture series in

palliative medicine can have in resident knowl-edge and skill in providing palliative care.

2. Describe an educational program in palliativemedicine that can be used to train internal med-icine providers.

Background. Caring for patients with palliative needsis an essential part of internal medicine training, how-ever this care is often not a focus of resident educa-tion. Less than half of accredited internal medicineprograms have required palliative care lectures intheir curriculum. Due to a shortage in palliative carespecialists, internal medicine physicians need basicpalliative training to help fill this need.Aim Statement. The purpose of this quality improve-ment project was to create and maintain a curriculumto improve resident knowledge and skill in caring forpatients with palliative care needs.Methods. Internal medicine residents at the Univer-sity of Chicago were surveyed to determine theirknowledge, attitudes, and skills regarding palliativecare pre and post-curriculum. The curriculum utilizedpresentations, workshops, and simulation over a one-




year period. Chi-squared analysis was used to comparepre and post-curriculum results.Results. Response rate was 37% pre-curriculum(n¼46/123) and 41% post-curriculum (n¼51/123).Based on the pre and post-curriculum surveys, 61%self-reported baseline competence in providing symp-tom management for patients with chronic disease orlife limiting illness, improving to 88% post-curriculum(p¼0.002). On initial survey, a minority reportedcompetence in providing patients with palliative re-sources (26%) which improved to 48% post-curricu-lum (p¼0.027). Reported knowledge in the role ofpalliative care consultants increased from 78% to90% post-curriculum (p¼0.105) and knowledge inthe educational training palliative care consultantsreceive increased from 35% to 57% (p¼0.029).Conclusions and Implications. This study discov-ered that at baseline, many internal medicine residentsreport lack of competence in providing palliative carefor patients. Following the one year curriculum,improvement in resident knowledge and skill wasobserved, however residents still report weaknesses inthese areas. Based on these results, this curriculumhas the potential to improve resident knowledge andskill in caring for patients with palliative needs.

You’ve Got This! Developing PrimaryPalliative Care Education Within a Safety-Net Health System (QI734)Timothy Poore, MD, University of California, SanFrancisco, San Francisco, CA. Catherine Deamant,MD FAAHPM, Cook County Health and Hospitals Sys-tem, Chicago, IL. Bridget Sumser, LMSW, Universityof California, San Francisco Medical Center, San Fran-cisco, CA. Anne Kinderman, MD, Zuckerberg SanFrancisco General Hospital/University of California,San Francisco, San Francisco, CA. Heather Harris,MD, Zuckerberg San Francisco General/Universityof California, San Francisco, San Francisco, CA.

Objectives1. Identify needs and opportunities to implement

primary palliative care education with a focuson vulnerable and underserved patient popula-tions served in safety-net settings.

2. Describe the impact of an interdisciplinary pri-mary palliative care education session for pro-viders caring for seriously ill, vulnerable andunderserved patients in safety-net setting.

Background. California recently passed a state billmandating access to palliative care services for Medi-Cal patients with serious illness. With a limited numberof palliative care specialists serving these often vulner-able and underserved patients with complex psychoso-cial needs, it is crucial to educate non-palliative caretrained front-line providers in primary palliative care.

Aim Statement. To increase access to palliative carethrough the development and implementation of ano-cost primary palliative care curriculum for vulner-able and underserved patients in the San Franciscosafety-net system.Methods. Curriculum development was based on acomprehensive needs assessment, including: inter-views with content experts, organizational leadersand key stakeholders; an environmental scan; a litera-ture review; and an online survey. We created a novelhalf-day training program with content focused ondefining palliative care and serious illness, differenti-ating palliative care from hospice care, and seriousillness communication skill training specific to vulner-able patient care. Curricular impact was gaugedthrough pre and post-surveys which assessed for confi-dence in participants’ understanding and ability toprovide palliative care (Likert scale from 1-10).Results. Four half-day education sessions were con-ducted with a total of 40 participants from the SanFrancisco Department of Health, including non-clin-ical case managers, social workers, nurses, nurse prac-titioners and physicians. Participants reportedsignificantly higher confidence in their ability todescribe palliative care to a patient (pre-5.09 to post-8.33), differentiate palliative care vs. hospice care(4.92 to 8.73), define serious illness (6.08 to 8.60),define illness trajectories (4.98 to 7.90), elicit patient’sillness understanding, prognostic awareness and goals(5.32 to 8.23) and describe advance care planning(6.03 to 8.23).Conclusions and Implications. A half-day courseintroducing basic palliative care concepts and commu-nication skills to non-palliative care trained inter-disci-plinary providers can improve confidence in providingpalliative care to patients in a safety-net setting.

Coaching Palliative Care Conversations:Evaluating the Impact on ResidentPreparedness and Goals of CareConversations (QI735)Rachel Rodenbach, MD, University of Pittsburgh,Pittsburgh, PA. Dio Kavalieratos, PhD, University ofPittsburgh, Pittsburgh, PA. Anoo Tamber, MD, BCCancer, Surrey, BC, Canada. Corey Tapper, MD, Uni-versity of Pittsburgh Medical Center, Pittsburgh, PA.Judith Resick, MSN MPH RN, University of Pittsburgh,Pittsburgh, PA. Robert Arnold, MD FACP FAAHPM,University of Pittsburgh, Pittsburgh, PA. Julie Childers,MD MS FAAHPM, University of Pittsburgh, Pittsburgh,PA.

Objectives1. Describe a novel coaching intervention to

improve palliative care skills of residentphysicians.




2. Identify positive outcomes in resident prepared-ness and goals of care documentation resultingfrom palliative care coaching sessions.

Background. Lectures alone are likely inadequate toprepare resident physicians for goals of care (GOC)discussions. Drawing on adult learning theories, wedeveloped a real-time, learner-centered coachingintervention to improve palliative care (PC) skills ofresidents.Aim Statement. Evaluate whether short interactivePC didactic and coaching sessions increased residentphysicians’ preparedness in discussing PC topics andcompletion of GOC discussions with hospitalizedpatients.Methods. Prospective quality improvement studyinvolving internal medical residents during theirfour-week hospitalist rotation at an urban medical cen-ter. Two PC physicians led the sessions. Brief didacticstrained each three-person resident group in a GOCcommunication roadmap and appropriate documen-tation. Informal coaching sessions centered on build-ing skills related to conducting GOC conversationsand addressing other PC topics residents raised. Resi-dents completed surveys addressing their level of pre-paredness on a 5-point scale (from 1 ¼ not wellprepared to 5 ¼ very well-prepared) in issues relatedto discussing GOC pre/post-rotation. We measuredGOC documentation in a defined patient populationconsidered ‘‘at-risk’’ by the health system based onage, comorbid conditions, and frequent hospitaliza-tions. We monitored data monthly and made smallcurricular adjustments during the year.Results. We trained 39 residents over 12 months inthrice-weekly coaching sessions lasting on average 16minutes. Residents’ level of preparedness increasedacross several GOC topics. The greatest increaseswere in eliciting patients’ fears for the end of life(pre/post 3.3/4.1, change +0.8) and helping patientstalk with their families about the future (pre/post3.4/4.2, change +0.8). Documented GOC discussionsin at-risk patients increased from 17.4% pre-hospitali-zation to 53.9% by the end of hospitalization. Residentphysicians rated coaching sessions as useful (4.5) andrelevant (4.3) to their training.Conclusions and Implications. Brief coaching ses-sions can integrate PC education into a busy clinicalservice and improve the likelihood that residents willfacilitate GOC discussions with hospitalized patients.

The Implementation of a ComprehensiveAdvanced Care Planning (ACP) Program(QI736)Alan Roth, DO FAAFP FAAHPM, Jamaica HospitalMedical Center, Jamaica, NY. Angelo Canedo, PhD,

Jamaica Hospital Medical Center, Richmond Hill,NY. Vandana Chauhan, PhD, Jamaica Hospital Medi-cal Center, New York, NY.

Objectives1. Explain the importance of advance care planning

(ACP) to our nation’s healthcare system and spe-cifically how it affects end-of life care.

2. Discuss our advance care planning hospital wideperformance improvement project.

Background. A recent national survey found thatabout 90% of people say that talking with their lovedones about end-of-life care is important but about27% have actually done so. Much of this care involvesextremely aggressive measures to preserve life, result-ing in higher costs and decreased patient/familysatisfactionAim Statement. To create an Advanced Care Plan-ning (ACP) system as an integral part of thehealth care routine and the community cultureby facilitating conversations that help individualsand those close to them come to a clear and fullerunderstanding of their values, goals and treatmentpreferences.Methods. Over a two year period, we conductedthree hospital wide trainings on having ACP dis-cussions, developed educational materialsincluding an informational video, improved work-flow in the electronic medical record (EMR) totrack discussions, education and documentation,and stressed use of billable codes for discussionto generate revenue.Results. Initial preliminary data showed 24.76% ofoutpatients and 36.95% of inpatients had an advancedirective in the EMR. This increased to 92% into thesecond year for outpatients. The improved EMR work-flow to track discussion and education revealed that,at start of second year, 69% of inpatients had beenasked about advance directives (‘‘discussion’’), 52%were provided information packet/shown educationalvideo (‘‘education’’), and 15% of inpatients had anadvance directive on file. Increased discussions arefurther revealed by steadily increasing revenue frombillable discussions (CPT codes 99497 & 99498) with$4,693 at end of first year to $9,998 three monthsinto second year.Conclusions and Implications. Training providers,distribution of educational materials to patients andimproving workflows in the EMR resulted in increasedACP billable discussions. Examination of inpatientdata suggests that increased discussions do not readilytranslate into documentation on file, suggesting cul-tural factors might affect actual completion ofdocuments.



Decreasing Barriers and IncreasingConfidence: Ambulatory Advance CarePlanning Internal Medicine ResidentCurriculum (QI737)Adi Shafir, MD, University of Pittsburgh Medical Cen-ter, Pittsburgh, PA. Dio Kavalieratos, PhD, Universityof Pittsburgh, Pittsburgh, PA. Jane Schell, MD, Univer-sity of Pittsburgh, Pittsburgh, PA. Julie Childers, MDMS FAAHPM, University of Pittsburgh, Pittsburgh, PA.

Objectives1. Employ REMAP as a framework for teaching

ambulatory ACP to learners.2. Illustrate innovative teaching methods including

use of drills and ‘homework’ to reinforcelearning.

3. Identify barriers and opportunities to ambulatoryACP completion by residents and physicians ingeneral.

Background. Patients are eager to participate inadvance care planning (ACP) discussions, but thereare numerous physician barriers to outpatient discus-sions. Education in overcoming these barriers arerare in residency programs.Aim Statement. Reduce perceived barriers to ACPand increase resident willingness and confidence tohave ACP conversations in the clinic.Methods. Second and third year internal medicineresidents participated in two 3-hour sessions duringan ambulatory care rotation. The first session pre-sented information about ACP; a conversation frame-work using the acronym REMAP (Raise the issue,respond to Emotion, Map patient values, Affirm thepatient, and propose a Plan); drills practicing usingREMAP, and documentation in the electronic medicalrecord. Residents were asked to discuss ACP with aclinic patient in between sessions, and write downhow the conversation went. During the second session,residents practiced ambulatory ACP skills using asimulated patient, and debriefed the homework. Apre-survey was completed by participants before thecurriculum, and a post-survey was completed immedi-ately after, ranking confidence and barriers on a 5-point Likert scale.Results. 54 residents completed the pre-survey, and50 completed the post-survey. Pre-post interventionmedians were compared using Wilcoxon-Mann-Whit-ney U tests due to non-normal data. After completionof the curriculum, residents felt more confidentbringing up ACP (p<0.001), discussing choosing asurrogate decision-maker (p<0.001), were morewilling to bring up ACP (p¼0.007), and felt it wasmore important to bring up ACP (p¼0.002). Notablythey no longer felt time was a barrier to discussingACP (p<0.001), and no longer felt uncomfortableinitiating the discussion (p¼0.049).

Conclusions and Implications. An ambulatory ACPcurriculum that includes a structured conversationframework and opportunities to practice with simu-lated patients is effective in improving resident confi-dence and willingness to complete ACP and helppatients identify a surrogate decision maker. Futureresearch will evaluate whether education leads toimprovement in completion of AD.

Redundancy or Value-Added? EthicsConsults in Hospitalized Patients withPalliative Medicine Involvement (QI738)JoAnna Slobodnjak, MD, Maine Medical Center, Port-land, ME. Frank Chessa, PhD, Maine Medical Center,Portland, ME. Rebecca Hutchinson, MD MPH, MaineMedical Center, Portland, ME.

Objectives1. Describe services typically provided by clinical

ethics services and the ways in which these over-lap with reasons for palliative medicine consults.

2. Discuss ways in which incorporation of ethicalprinciples in palliative medicine notes and rec-ommendations may provide additional reassur-ance to primary teams.

Background. Many academic medical centers haveboth palliative medicine (PM) and clinical ethicsteams; there is overlap in the services provided bythese teams.Aim Statement. We sought to better understandwhat services are provided by clinical ethics consulta-tions not currently offered by PM.Methods. We performed a retrospective chart reviewon all patients seen by both PM and clinical ethics dur-ing a single admission over a two-year period. Assess-ment and recommendations from clinical ethicsnotes were abstracted verbatim. Two members of thestudy team used MAXQDA software to independentlycode themes present in the clinical ethic notes; theymet to discuss and reach consensus on codes.Results. We identified 84 patients seen by both palli-ative medicine and clinical ethics during a singleadmission over a two-year period. The three most com-mon issues addressed by the ethics consult were goalsof care, medical surrogacy, and providing support.Specific issues relating to goals of care included: pa-tients refusing medical treatment or wanting to returnhome, code status discrepancies, and concerns aroundwithdrawing or continuing life-sustaining treatment.Surrogacy issues addressed included: assistance withidentifying an appropriate surrogate, concerns aboutthe surrogate, decision making when no surrogate isavailable, and guardianship. There was specific lan-guage regarding legal or ethical recommendations in32/84 patients (38%). We observed frequent explicituse of ethical principles in ethics notes, such as




autonomy and beneficence, which were not includedin PM notes.Conclusions and Implications. We noted that manyof the services provided by clinical ethics are similar tothose offered by PM including assistance with goals ofcare conversations and advice regarding surrogacy.However, use of language such as ‘‘ethically permis-sible’’ or ‘‘legally permissible,’’ accompanied by moralreasoning, may be delivering additional reassurance tomedical teams not currently provided by PM. PM clini-cians may be able to further assist primary teams by us-ing ethical reasoning in their assessments andrecommendations.

Development and Implementation of aPatient-Centered Tool for the Assessment ofan In-Patient Palliative Care Team (QI739)Ana Marija Sola, BA BS, University of California, SanFrancisco School of Medicine, San Francisco, CA. Ale-jandro Perez, BS, University of California, San Francis-co, San Francisco, CA. Sirisha Narayana, MD,University of California, San Francisco, San Francisco,CA. Teri Rose, BA, Braddock Family Foundation, Wal-nut Creek, CA. David O’Riordan, PhD MPH, Univer-sity of California, San Francisco, San Francisco, CA.Giovanni Elia, MD FAAHPM, University of California,San Francisco, San Francisco, CA.

Objectives1. Identify primary issues for patients’ satisfaction

with an inpatient Palliative Care team.2. Describe areas for improvement for the palliative

care team’s operation and composition as identi-fied by patients receiving care.

3. Recognize the use of Lean A3 process to improvethe administration and refine the content of asurvey tool.

Background. Patient feedback is an important partof evidence-based, high value care. We wanted todevelop a tool for more rigorous assessment of Pallia-tive Care Services (PCS) at our institution.Aim Statement. Design and implement a protocol,using a standard Lean A3 problem-solving approach,for collecting inpatient feedback on Palliative Care(PC) team performance.Methods. Eligible inpatients receiving PCS at ourinstitution were approached in person over a 9-monthperiod, to complete a semi-structured interviewregarding their experience of care. The survey toolincluded Likert scale-based and open-ended ques-tions. We examined characteristics of all patientsmeeting eligibility criteria and thematically reviewedresponses from patient interviews. Lean A3 methodswere applied to plan and improve the process.Results. Of the 74 eligible patients, 21 completed theinterview. Major themes included: Felt understood

(excellent/good: 95%); communicating plans (excel-lent/good: 80%), effectively respond to spiritual andreligious needs (excellent/good: 75%), team availabil-ity (always: 65%), controlling/alleviating symptoms(excellent/good: 80%), sharing information aboutillness (excellent/good: 70%), likelihood to recom-mend PCS (very likely: 90%). The open-ended ques-tions identified satisfaction with time spent with andclarification of issues by the PC team. Other commonthemes included the desire for increased culturalsensitivity and diversity of the PC team. We observedmixed responses about patients’ previous or currentunderstanding of PC and the PCS offered.Conclusions and Implications. Patient’s perceptionof team effort, active listening and strength of relation-ship with providers has a beneficial impact on the pa-tient’s experience of care. Areas for improvementwere identified as team availability and sharing of in-formation between providers and patients. Feedbackregarding team diversity and previous misunderstand-ings of PC highlighted the need for continuing publiceducation and re-assessment of the composition of thePC team at our institution. Lean A3 methods werehelpful in planning and improving the surveyprocess.

Caution! Unstable Patients Will CollapseWithout Warning: Improving AdvanceDirective Completion for Patients withChronic Obstructive Pulmonary Disease inan Urban, Safety-Net Hospital (QI740)Amanda Stephens, MD, National Jewish Health, Den-ver, CO. Kathleen Bickel, MD MPhil MS, Universityof Colorado School of Medicine, Aurora, CO. DanielHandel, MD, University of Colorado School of Medi-cine, Aurora, CO. Stefani Madison, MD, StephensonCancer Center, Oklahoma City, OK.

Objectives1. Recognize the unique challenges to advance care

planning discussions in patients with severe pul-monary disease.

2. Identify markers of poor prognosis in pulmonarydisease through a novel advance care planningtrigger.

3. Evaluate an intervention to increase advance careplanning in the outpatient setting with severepulmonary disease.

Background. Despite recommendations, advancedcare planning (ACP) occurs infrequently in patientswith COPD. A few studies describe rates of 11-15%,with scant information regarding methods to increaseACP in this population.Aim Statement. Over six months, to increaseadvance directive (AD) completion by 10% in patientswith COPD requiring outpatient subspecialty care.




Methods. Stakeholders (physicians and nurses) at anurban, safety-net pulmonary subspecialty clinicconvened, reporting three primary challenges inACP: discomfort discussing ACP in clinic, inability tolocate AD documents and identifying patients appro-priate for ACP. Consequently, a two-step interventionwas implemented over 8 months: 1) education ad-dressing ACP discomfort and 2) a novel remindernudge with COPD-specific ACP criteria plus re-struc-turing the clinic’s AD process. As pulmonary providerswere encouraged to complete patient ADs themselvesor refer to an outpatient palliative care specialist, ADcompletion of patients with COPD seen in eitherclinic was tracked, using statistical process control p-charts.Results. Before the intervention (June 2016-September 2017), the monthly AD completion rateamong COPD patients seen in pulmonary and pallia-tive subspecialty clinics was unstable by statisticalprocess control, with a mean of 25.4% (range 13%-39%). The average monthly number of patients withCOPD seen in both clinics was 106. With the educa-tion cycle (end October 2017), followed by thereminder nudge and AD paperwork optimization(March 2018), the new average AD completion ratewas 28.8% (range 6-42%). Special cause signals indi-cating significant process change were a shift(December 2017-June 2018) and a point outside the3-sigma upper control limit (June 2018).Conclusions and Implications. Although the proj-ect aim was not met, the combined intervention wasassociated with special cause improvement in ADcompletion. The process was not under statistical con-trol prior to intervention, and continued measure-ment is necessary to ensure sustained results.However, this study implies that with combined educa-tion, a nudge and re-organization, AD completion inthis population can improve, thus improving out-comes for patients and families.

Compassionate Technology: Palliative CareTelemedicine in the Rural Hospital Setting(QI741)Matt Stinson, PA, Calaway Young Cancer Center, Glen-wood Springs, CO. Elizabeth Belanger-Shugart, APRN,Valley View Hospital, Glenwood Springs, CO. Kather-ine Morrison, MD FAAHPM, University of Colorado,Aurora, CO. Maurice Scott, MD, The University of Col-orado School of Medicine, Denver, CO.

Objectives1. Recognize the challenges of providing palliative

care in rural hospitals.2. Describe the use of telemedicine to improve ac-

cess to palliative care in our organization.

3. Reflect on how a similar approach may be used inother rural medical organizations.

Background. Early hospital-based palliative care isassociated with significant improvement in patientquality of life and lower hospital costs. Although thereis a robust presence of palliative care services in urbancenters, there is a substantial disparity of care for seri-ously ill patients in rural areas. The medical literaturedemonstrates that telemedicine has been successful inthe hospital setting for treatment of disease and in thehome setting for palliative care. There is currently noavailable literature describing the use of telemedicineto address inpatient palliative care at rural hospitals.Aim Statement. The purpose of the program is todetermine the feasibility of utilizing telemedicinewith palliative care services for adult inpatients at a ru-ral community hospital.Methods. An interdisciplinary team was formed atValley View Hospital consisting of local providers, so-cial workers, chaplains, and physicians from the Uni-versity of Colorado Anschutz Medical Campus viateleconferencing. Palliative care consultations wereperformed with adult inpatients with a focus onadvance care planning, symptom management,communication and prognostication. At subsequentvisits, the patients were asked about the perceivedvalue of the service and acceptability of the teleconfer-encing component.Results. The program launched in January 2018, and19 patients were seen in the initial 4 months of the ser-vice. 95% of patients tolerated teleconferencing welland reported satisfaction with the service. There wasa 26% increase in completion and documentation ofadvance care plans following the visit. 30 patientswere identified as needing palliative care services butcould not be seen due to limited staff and time.Conclusions and Implications. Telemedicine maybe an option for rural healthcare facilities needinginpatient specialized palliative care services. Key com-ponents to program success include concurrent edu-cation for involved providers, adequate staffing, andsufficient technological support for telemedicineequipment and software.

Characterizing Life-Sustaining TreatmentDecisions of Seriously Ill Veterans DuringPilot Testing of the Veterans HealthAdministration’s Life-Sustaining TreatmentDecisions Initiative (QI742)Anne Walling, MD, University of California, Los An-geles, Los Angeles, CA. Karleen Giannitrapani, PhD,VA Health Services Research and Development Ser-vice/Stanford, Palo Alto, CA. Mary Beth Foglia, PhDMA RN, Veterans Health Administration, WashingtonD.C. Jill Lowery, PsyD, Veterans Health




Administration, Washington D.C. Lisa Lehmann, MD,Veterans Administration, Bedford, MA. Natalie Lo,OT, MPH, Center for Innovation to Implementation,Menlo Park, CA. Ariadna Garcia, MS, Stanford Univer-sity, Palo Alto, CA. David Bekelman, MD MPH, Univer-sity of Colorado, Denver, CO. Karl Lorenz, MD MSMSHS, Stanford/VA Palo Alto Health Care System,Palo Alto, CA.

Objectives1. Describe essential elements of the Life-Sustaining

Treatment Decisions Initiative.2. Apply lessons learned from this evaluation of the

Life-Sustaining Treatment Decisions Initiative toother implementation efforts related toimproving advance care planning.

Background. To ensure that seriously ill Veterans’values, goals and preferences for life-sustaining treat-ments are elicited, documented and honored, VA’sNational Center for Ethics in Health Care imple-mented the Life-Sustaining Treatment DecisionsInitiative (LSTDI).Aim Statement. We aimed to characterize goals ofcare conversations (GOCC) and LST decisions atfour VA pilot sites between 8/11/2014 and 11/14/2016.Methods. Data from the patient medical record waslinked to health factor (HF) data from the LSTDI tem-plate for initial GOCC. Descriptive statistics were per-formed for the following HF: Decision MakingCapacity (DMC), Consent, Goals of Care (GOC) andResuscitation status. We evaluated HF associationswith DMC and chi-square t-tests were used to evaluatecomparisons. We performed brief chart abstractionsfor rare instances of validity concerns.Results. 6664 Veterans had >¼ 1 GOCC and were onaverage 72 years old, 93% male, 87% white, 61% ur-ban. 35% of Veterans died. 15% with documentedGOCC lacked DMC and <1% lacked a decision-maker.GOC varied for the cohort and included (more thanone goal allowed): to be cured (8%), to prolong life(34%), to improve/maintain quality of life (62%), tobe comfortable (53%), to obtain support for family/caregiver (8%), to achieve life goals (2%), and other(11%). Most with an initial LSTDI note had a DNR or-der (59%). Veterans lacking DMC were more likely tohave comfort oriented goals (49% vs. 77%, p<0.01)and a DNR order (53% vs. 84%, p<0.01) comparedto those with DMC. Most cases examined via chartabstraction due to data validity concerns were implic-itly validated and identified opportunities to make ad-justments to the LSTDI template to improve workflow.Conclusions and Implications. LSTDI and goals ofcare documentation was successfully implemented atfour pilot sites. Lessons learned will inform ongoingimplementation across the VA nationally.

Integrating Creative Art Therapy withPalliative Care (QI743)Jeanie Youngwerth, MD FAAHPM, University of Colo-rado School of Medicine, Denver, CO. Jean Kutner,MD MSPH FAAHPM, University of Colorado Schoolof Medicine, Aurora, CO. Liz Somes, MD, ColoradoUniversity Denver Hospice and Palliative Med Fellow-ship, Denver, CO. Amy Jones, MA LPC, University ofColorado Hospital, Aurora, CO. Angela Wibben, MMMT-BC, University of Colorado Hospital, Aurora, CO.

Objectives1. Discuss the integration of a creative art therapy

program with a palliative care service.2. Describe program evaluation outcomes with

symptoms, and patient and family member per-ceptions regarding their experiences with crea-tive art therapy.

3. Demonstrate the emotionally therapeutic effectsof creative art therapy with patient examples.

Background. The University of Colorado Hospital(UCH) implemented a Creative Art Therapy (CAT)program in September 2016, offering CAT to hospital-ized patients as a component of comprehensive Pallia-tive Care services.Aim Statement. To evaluate short term effects of theCAT session on patient symptoms and patient/familyperceptions regarding their experience with CAT.Methods. CATwas made available to all adult (18 andolder) UCH Palliative Care Consult Service (PCCS)patients beginning in September 2016. The programevaluation occurred between 10/1/16-6/1/17. Pa-tients self-selected a CAT session with a music or arttherapist. The program evaluation consisted of aquantitative and qualitative component:

a. Self-report of 3 symptoms that we hypothesizedCATwould have a short-term effect on: pain, anx-iety, and well-being (0-10 scale) using the Ed-monton Symptom Assessment Scale (ESAS)prior to and following the CAT session; n¼12patients.

b. Semi-qualitative patient and family member in-terviews using 6 questions to acquire in-depth in-formation about perceptions of and experienceswith the CAT session, administered within 1 dayof the CAT session; n¼40 patients and familymembers.

Results. During the study period, there were 366 CATpatient encounters. 12 patients completed pre/post-CAT session ESAS. Symptom scores showed a trendin improvement on a 0-10 scale for pain (4.8 to 4.3;p¼0.410), anxiety (2.7 to 2.4; p¼0.699), and well-be-ing (5.8 to 4.8; p¼0.376) from pre-intervention topost-intervention. 40 patients and family memberscompleted semi-qualitative interviews. Qualitativeanalysis revealed the over-arching theme of improved



quality of life through emotionally therapeutic effects,with sub-themes of non-medical relationships, distrac-tion, family engagement, and personalized care.Conclusions and Implications. This pilot CAT-PCCS program evaluation supports the need toconfirm the association with positive trends in pa-tient-reported pain, anxiety, and well-being in furtherstudies with larger enrollment, and provides qualita-tive themes of patient and family member emotionallytherapeutic effects.

Caring About CancerdAdvance CarePlanning Group Visit Intervention (QI744)Jeanie Youngwerth, MD FAAHPM, University of Colo-rado School of Medicine, Denver, CO. Nancy Robert-son, MSN ANP-BC ACHPN, University of Colorado,Aurora, CO. Erin Nielsen, LCSW, The Holding Groupand University of Colorado Hospital, Denver, CO. Hill-ary Lum, MD PhD, University of Colorado School ofMedicine, Aurora, CO.

Objectives1. Implement the ENACT-GV intervention to per-

sons living with cancer.2. Determine the feasibility and acceptability of the

ENACT-GVs at the University of Colorado Hospi-tal Anschutz Cancer Pavilion.

3. Evaluate the impact of ENACT-GV on medicaldurable power of attorney documentation.

Background. Electronic health record (EHR) docu-mentation of a medical durable power of attorney(MDPOA) for persons living with cancer is only 25%at the University of Colorado Hospital Anschutz Can-cer Pavilion (UCH-CP). Engaging in Advance Careplanning Talks Group Visit (ENACT-GV) interventioncan improve ACP discussions and EHR documenta-tion in the primary care setting.Aim Statement. To increase MDPOA documentationfrom 25% to 50% in persons living with cancer withinsix months.Methods. Patients were recruited from UCH-CPGenitourinary Cancer Clinic by letters and phonecalls. Three ENACT-GV were conducted, each consist-ing of two 2-hour sessions, one month apart, facili-tated by a palliative care nurse practitioner andUCH-CP social worker. The groups consisted of #10participants with an option for patients to bring aguest. ACP documentation in the EHR (baseline andone week after the 2nd session) and the 4-item ACPEngagement Survey (baseline and immediately postsession) were assessed. Cycle 1 implemented ENACT-GVs at UCH-CP, cycle 2 optimized patient referrals,and cycle 3 enhanced group facilitation techniques.Results. Ten patients (<2% recruitment rate) wererecruited. Three patients returned for the 2nd session(30% retention rate). Seven guests attended ENACT-

GVs (all attended only the 1st session). At baseline,20% had a MDPOA in the EHR; one week after the2nd session, 100% had MDPOAs in the EHR(p<0.001). 17% of patients (n¼2205) who did notattend ENACT-GV had MPDOAs in the EHRcompared to 100% who attended ENACT-GV(p<0.001). 87% of participants believed ENACT-GVwas better than normal clinic visits for ACP discus-sions. 93% of participants would recommendENACT-GV to another person. 50% of patientspreferred one session over two sessions.Conclusions and Implications. ENACT-GV was anacceptable method to engage persons living with can-cer in ACP discussions and significantly improveddocumentation, however, recruitment was low.Limiting ENACT-GV to a single group session mayimprove feasibility.

Residential Homes for the Dying: AnUntapped Resource for Teaching Patientand Family-Centered End of Life CareBefore Professional School Training (S801)Catherine Adams, MD PhD, St. Peter’s ssHospital, Al-bany, NY. Kelly Melekis, PhD MSW, Skidmore College,Saratoga Springs, NY. Carol Weisse, PhD, Union Col-lege, Schenectady, NY.

Objectives1. Describe a clinical training program for under-

graduate health professions students to providebedside end of life care.

2. Explain the effect which participation in theCARE program had on undergraduate healthprofessions students.

Original Research Background. There is a lack ofclinical training opportunities in end-of-life care, espe-cially at the bedside. The Community Action,Research, and Education (CARE) program was devel-oped to help students interested in healthcare betterunderstand the challenges of providing care whenno cure is possible. Program participants train to serveas surrogate family members and spend 8 weeksproviding care to 2 hospice patients in residentialhomes for the dying in upstate New York. In additionto providing 24 hours of direct bedside care per week,students complete 10 online learning modules, eachof which emphasize different skills for providing endof life care. Students also meet with hospice and palli-ative care professionals and conduct agency-drivenresearch to benefit the home where they provide thiscare.Research Objectives. To determine whether theCARE program improves empathy and self-efficacy toprovide end of life care among participants.Methods. Eighteen undergraduate health profes-sions students from 4 different institutions completed




the 8-week summer program that included weekly di-dactic sessions with hospice and palliative care pro-viders. Empathy and Self-Efficacy to provide end oflife care were measured in 18 undergraduate healthprofessions’ students (7 men; 11 women) betweenthe ages of 19 and 27 before and after the program.Results. Paired t-tests revealed significant increases inperceived self-efficacy to provide end of life care(p < .001) and empathy (p < .05) among participantsfollowing completion of the program.Conclusion. Residential homes for the dying offer aunique patient care experience with time to practiceend of life care with instruction by, and observationof, more experienced caregivers.Implications for Research, Policy, orPractice. There are approximately 30 residentialhomes for the dying in upstate NY and 30 othersacross the U.S. in need of caregivers. This educationalinitiative represents an opportunity to improve com-munity-based end of life care and cultivate commu-nities of compassionate caregivers.

Characteristics of Hospices ProvidingHigh-Quality Care (S802)Rebecca Anhang Price, PhD, RAND Corporation, Ar-lington, VA. Anagha Tolpadi, MS, RAND Corporation,Santa Monica, CA. Joan Teno, MD MS, Oregon Healthand Science University, Portland, OR. Marc Elliott,PhD, Rand Corporation, Santa Monica, CA.

Objectives1. Identify hospice characteristics associated with

high performance on CAHPS Hospice Surveymeasures.

2. Identify hospice characteristics associated withhigh performance on Hospice Item Setmeasures.

3. Compare hospice characteristics associated withhigh performance on Hospice Item Set measuresto characteristics associated with high perfor-mance on CAHPS Hospice Survey measures.

Original Research Background. Newly availabledata from the Hospice Quality Reporting Programallow for examination of hospice characteristics thatare associated with high-quality hospice care.Research Objectives. Examine hospice characteris-tics associated with high performance on HospiceItem Set (HIS) and Consumer Assessment of Health-care Providers and Systems (CAHPS) Hospice Surveymeasures.Methods. We used 2015 hospice claims and 2016 Pro-vider of Services data to identify structural features ofhospices, characteristics of their patients, and theirprocesses of care. We used logistic regression modelsto assess the association between hospice characteris-tics and hospices’ being in the top quartile of 2015-

2017 performance for HIS measures, CAHPS mea-sures, or both.Results. Of the 2,746 hospices in our analysis, 5.6%were in the topquartileofbothHISandCAHPSmeasureperformance. Hospice characteristics associated withbeing in the top quartile for HIS included being in afor-profit chain, larger size (91+ patients per year), andhaving fewer than 40% of patients in a nursing home.Characteristics associated with being in the top quartilefor CAHPS included being a non-profit and non-chainhospice, smaller size (< 200 patients per year), andserving a rural area. Providing professional staff visitsin the last two days of life to a higher proportion of pa-tientswas associatedwithhospices’being in the topquar-tile of HIS and in the top quartile of CAHPS.Conclusion. Hospice characteristics associated withstrong performance on clinical process measuresdiffer from those associated with better patient andfamily experiences of care; however, some hospicesachieve high performance on both domains, suggest-ing that there is no inherent tradeoff between them.Implications for Research, Policy, orPractice. Variation in care quality by hospice charac-teristics suggests opportunities for improvement.

Teaching the Skill of Shared DecisionMaking Utilizing a Novel OnlineCurriculum: A Blinded RandomizedControlled Pilot Study (S803)Joshua Arenth, MD, University of North CarolinaHealthcare, Chapel Hill, NC. Kenneth Pituch, MD,University of Michigan, Ann Arbor, MI. Jessica Turn-bull, MD MA, Monroe Carell Jr. Children’s Hospital,Nashville, TN.

Objectives1. Identify the components of shared decision mak-

ing using a values-guided support approach.2. Identify language that identifies parental values

and how language to elicit these values can belearned via the educational intervention.

3. Identify common ways that decisions may beinadvertently and inappropriately framed duringconversations in high stakes situations.

Original Research Background. Competence inshared decision making for non-palliative care profes-sionals is becoming essential as children’s hospitals areincreasingly caring for more chronically and criticallyill children. The resources of Pediatric PalliativeCare teams are often stretched thin, and the medicalteam is often unable to ascertain families’ goals ofcare during rounds or bedside discussions, often dueto lack of communication training and skills.Research Objectives. To test the effectiveness of anonline module in improving the language of shareddecision making used by non-palliative care pediatric




providers who often participate in shared decisionmaking with patients and families.Methods. Pediatric subspecialty fellows were video-re-corded in a simulated patient encounter with parentsfacing a decision to either go forward with a life-extend-ing procedure or transition to a course of care aimed atcomfort. Conversations were evaluated with a validatedscoring tool for the degree of shared decision makingpresent on a scale of 0-11. The intervention groupthen received a brief online curriculum aimed at teach-ing the skill of shared decision making. Participantsfrom both groups then repeated the same simulationand were reassessed. Members of the control groupthen became a delayed-intervention group and also un-derwent the curriculum and a third simulation.Results. Regression analysis demonstrated the odds ofimproved performance in mean total score for interven-tion groups was 39.78 times greater than that of the con-trol group (95% CI [1.72 - 919.29]; P-value 0.022).Conclusion. Shared decision making is becomingmore and more important as children’s hospitals areincreasingly caring for more chronically and criticallyill children. Our data show that an easily accessibleeducational intervention in the form of an online mod-ule format is aneffectiveway of teaching thesebehaviors.Implications for Research, Policy, orPractice. Shared decision making behaviors in non-palliative care pediatric providers can be significantlyimproved by access to online educational modules.

Impact of Physician Attire on PalliativeCare Patients’ Perception of PhysicianCompassion and Professionalism: ARandomized Clinical Trial (RCT) (S804)Ahsan Azhar, MD FACP, The University of Texas MDAnderson Cancer Center, Houston, TX. KimbersonTanco, MD, The University of Texas MD AndersonCancer Center, Houston, TX. Ali Haider, MD, TheUniversity of Texas MD Anderson Cancer Center,Houston, TX. Janet Williams, MPH, The Universityof Texas MD Anderson Cancer Center, Houston, TX.Minjeong Park, PhD, The University of Texas MD An-derson Cancer Center, Houston, TX. Hilda Cantu, BS,The University of Texas MD Anderson Cancer Center,Houston, TX. Carolina Diaz, BSN, The University ofTexas MD Anderson Cancer Center, Houston, TX. Pet-ra Rantanen, BA, The University of Texas MD Ander-son Cancer Center, Houston, TX. Diane Liu, MS,The University of Texas MD Anderson Cancer Center,Houston, TX. Eduardo Bruera, MD FAAHPM, TheUniversity of Texas MD Anderson Cancer Center,Houston, TX.

Objectives1. List different elements of communications skills.

2. Discuss how attire can be a form of non-verbalcommunication.

Original Research Background. Environment is animportant component of communication skills. Physi-cians’ communication style including attire may influ-ence patient perceptions. Previous studies mostlybased on pictures of providers in different attires pro-vide conflicting evidence.Objectives. This RCT aimed to explore the effects ofa physician’s attire on patients’ perceptions. Hypothe-sis was that patients will perceive the physician withformal attire as more compassionate & professionalthan the physician wearing casual attire.Methods. 105 English speaking adult patients pre-senting as follow-ups to out-patient supportive carecenter, were randomized to watch 2 standardized, 3-minute video vignettes, with similar script, depictinga routine clinic encounter. In one video, physicianwas wearing formal attire with tie and buttoned upwhite coat, while in the other, physician was in casualattire without a tie or white coat. Actors and patientswere all blinded to the purpose of the study. Investiga-tors were blinded to the videos watched by the pa-tients. After viewing each video, patients completedvalidated questionnaires rating their perception ofphysician compassion (0 ¼ best, 50 ¼ worst), profes-sionalism (5 ¼ poor, 25 ¼ very good) & overall prefer-ence for the physician.Results. No significant differences seen betweenformal and casual attire for compassion [median (in-terquartile range), 25 (10, 31) vs 20 (8, 27); P¼0.31]and professionalism [17 (13, 21) vs 18 (14, 22);P¼0.42]. 30% (32) patients preferred formal, 31%(33) preferred casual attire and 38% (40) had no pref-erence. Subgroup analysis did not show statistically sig-nificant differences among age, sex, marital status &education level for compassion, professionalism &physician preference.Conclusions and Implications. Doctor’s attire didnot have an impact on patients’ perceptions of physi-cians’ level of compassion & professionalism and didnot influence their preferences for their doctor ortheir trust and confidence in the doctor’s ability toprovide care. More RCTs are needed to better under-stand the impact of different forms of attire on patientperceptions & preferences.

Advance Care Planning Education forPsychiatrists: A Novel Training Workshop(S805)Nicole Bates, MD, University of Pittsburgh MedicalCenter, Pittsburgh, PA. Jane Schell, MD, Universityof Pittsburgh, Pittsburgh, PA. Pierre Azzam, MD, Uni-versity of Pittsburgh Medical Center, Pittsburgh, PA.




Julie Childers, MD MS FAAHPM, University of Pitts-burgh, Pittsburgh, PA.

Objectives1. Identify medical and psychosocial factors that

support a greater need for advance care planningamong psychiatric patients.

2. Argue for the need for psychiatrists to receivetraining in advance care planning.

3. Evaluate the potential of a one-time, skills-basedtraining workshop to impact attitudes, comfort,skills, and knowledge of psychiatry residents inadvance care planning.

Original Research Background. Despite aperceived need, rates of advance directive completionamong psychiatric patients are low, and psychiatristsreceive little to no training in advance care planning.We identified advance care planning as an unmeteducational need in psychiatry and developed askills-based training workshop for psychiatry residents.

Research Objectives.� Identify areas of need for advance care planning(ACP) training for psychiatry residents at the Uni-versity of Pittsburgh Medical Center.

� Develop and implement a residency-wide educa-tional workshop to train psychiatry residents inmedical and mental health ACP.

� Assess residents’ attitudes, comfort, knowledge,and skills in engaging in medical and mentalhealth advance care planning before and afterthe workshop.

Methods. Based on results of a needs assessment, wedeveloped a three-hour ACP educational workshop forpsychiatry residents, including drill-based guided prac-tice in ACP communication skills and a case-based ro-leplay exercise to complete mental health advancedirectives. Psychiatry residents participating in thistraining completed pre- and post-workshop surveys as-sessing their attitudes, comfort, knowledge, and skillsin ACP domainsResults. Psychiatry residents completed this trainingand the pre-workshop (n ¼ 42) and post-workshop sur-veys (n ¼ 41). After the training, residents reported agreater responsibility to facilitate psychiatric ACP (p ¼0.03). They cited greater comfort in discussing end-of-life care (p ¼ 0.04), facilitating medical ACP(p ¼ 0.002), and facilitating psychiatric ACP(p <0.001). They reported being more able to addresskey elements of ACP with patients (p <0.001). Theyalso demonstrated statistically significant score increaseson objective measures of ACP knowledge. Finally, resi-dents reported high satisfaction with this training.Conclusion. A one-time educational workshop tar-geting produced improvements in psychiatry resi-dents’ attitudes, comfort, skills, and knowledge inACP.

Implications for Research, Policy, orPractice. This educational workshop is the firstknown ACP training for psychiatry residents. Thismodel proved effective at our institution and may beadapted to other psychiatry training settings.

The Landscape of Cardiac Palliative CarePractices in the United States (S806)Amy Beasley, DNP RN CHPN, University of Alabama,Tuscaloosa, AL. Dio Kavalieratos, PhD, University ofPittsburgh, Pittsburgh, PA. Marie Bakitas, DNSc NP-C FAAN, University of Alabama at Birmingham, Bir-mingham, AL.

Objectives1. Describe clinical characteristics of U.S. cardiac

palliative care programs.2. Differentiate between U.S. cardiac palliative care

programs challenges and successes.Original Research Background. Patients withadvanced cardiac disease (CD), and their caregivers,may benefit from early palliative care (PC) services.There is limited information on the number and na-ture of U.S. cardiac PC programs.Research Objectives. To describe operational andclinical characteristics of U.S. cardiac PC programs.Methods. We developed, pilot tested, and dissemi-nated an Internet survey to a convenience snowballsample of U.S. cardiac PC programs. The surveyincluded closed- and open-ended questions on prac-tice type, CD, service utilization, staffing, practicecharacteristics, referrals, funding sources, services,and challenges.Results. Seven (70%) cardiac PC programscompleted the survey. Most programs identified ascomprehensive (in-, out-, home services), outpatient(12.5%), inpatient (25%), or combined (25%). Pro-grams varied in duration of existence, from onemonth-12 years. All programs served HF patients,while others serve heart transplant (75%), COPD(75%), pulmonary arterial hypertension (62.5%),and interstitial lung disease (62.5%). Most outpatientprograms (71.4%) saw less than 50 new patientsannually, whereas one practice served 400. Most com-mon referral reasons were discussion of goals, man-agement of emotional symptoms, and preparednessplanning for ventricular assist devices. The mostcommonly managed symptoms were fatigue, dyspnea,depression, and anxiety. PC co-management (57.1%)and consult only PC (57.1%) were the most commonpractice models. All of the cardiac PC practices hadexpanded since opening. Qualitative follow up inter-views are planned to provide a deeper programunderstanding.Conclusion. Though cardiac PC programs are intheir infancy, there has been significant growth and



need for expansion. Understanding current cardiacpalliative care services, potential funding sources,and future needs is a high clinical/research priority.Implications for Research, Policy, orPractice. Study results provide an initial picture ofcardiac PC specialty programs which will be furtherexpanded based on qualitative interviews.

A Palliative Care Patient Navigator andCounseling Intervention for Latinos withStage III/IV Cancer (S807)David Bekelman, MD MPH, University of Colorado,Denver, CO. Regina Fink, PhD AOCN CHPN FAAN,University of Colorado Anschutz Medical Campus,Aurora, CO. Evelinn Borrayo, PhD, University of Col-orado Denver, Denver, CO. Danielle Kline, MS, Uni-versity of Colorado Denver, Aurora, CO. TimothySannes, PhD, University of Colorado Anschutz,Aurora, CO. Stacy Fischer, MD, University of ColoradoSchool of Medicine, Aurora, CO.

Objectives1. Describe a method for culturally tailoring a coun-

seling intervention.2. List 3 elements of feasibility when pilot testing a

new intervention.Original Research Background. Latinos withadvanced cancer are unlikely to access palliative careand can have high distress. To address this, we adaptedand combined two successful interventions, a patientnavigator and a counseling intervention. In priorwork, the patient navigator intervention increasedadvance care planning for Latinos, and the counselingintervention reduced depression in a general popula-tion using behavioral activation and interpersonalpsychotherapy.Research Objectives. (1) Adapt the content andformat of the counseling intervention to ensure cul-tural and literacy appropriateness with Latinos; (2)Determine the feasibility of the combined patientnavigator and counseling intervention, includingvideo counseling visits.Methods. (1) Community participatory actionapproach. We collaborated with Latino patient naviga-tors (n¼ 5) and a Latina psychologist with multicultur-alism expertise to revise the counseling treatmentmanual and patient materials. (2) Pilot test in Latinopatients with stage III/IV cancer who screen positivefor high distress or depressive or anxiety disorder.Results. 1) The counseling intervention underwentmajor changes. The written patient materials wereadapted to a 5th-6th grade reading level. Thirteen stor-ies of adjusting to illness were culturally tailored in aniterative process using paired navigator/study teammembers. Stories were transformed into video scriptsusing a similar process, with multiple revisions to

increase cultural tailoring and adhere to core coun-seling components. Videos were subsequently pro-duced. 2) 14 of 23 eligible Latinos with stage 3/4cancer enrolled. Participants were distressed (meanbaseline NCCN distress 6.2/10, SD 1.4; PHQ8 10.6,SD 6.2; GAD7 9.4, SD 6.5). Intervention visits areongoing and final pilot data will be presented.Conclusion. A patient navigator and counselingintervention was adapted for use with Latinopopulations.Implications for Research, Policy, orPractice. Developmental studies that culturally tailorestablished interventions to specific populationsrequire time and funding. The community participa-tory action approach we used could be applied toother interventions and populations.

Behind the Scenes: The CareCoordination/Non-Billable TimeAssociated with Outpatient PediatricPalliative Oncology (S808)Katharine Brock, MD MS, Emory University, Chil-dren’s Healthcare of Atlanta, Atlanta, GA. Kristen Al-len, MPH, Children’s Healthcare of Atlanta, Atlanta,GA. Karen Wasilewski-Masker, MD, Emory/Children’sHealthcare of AtlantaeAFLAC Cancer Center, Atlanta,GA. Jeffrey Klick, MD, Children’s Healthcare of Atlan-ta, Atlanta, GA.

Objectives1. Describe the unique aspects of a pediatric pallia-

tive oncology clinic and the time spent in carecoordination.

2. Identify the demographic and disease-based fac-tors that contribute to increased care coordina-tion time.

3. Formulate talking points to advocate for appro-priate care coordination time in your outpatientclinic model.

Original Research Background. Integrated pediat-ric palliative oncology (PPO) outpatient models areemerging to assist oncologists and patients with longi-tudinal support, symptom management, and care co-ordination. Considerable time is devoted to carecoordination, but the scope, time per patient, and ra-tio of non-billable to billable (NB:B) minutes is un-known. This information is crucial to designing newPPO outpatient clinics in order to understand andadvocate for appropriate personnel, physician time,and resources.Research Objectives. To determine the trends andratio of NB:B minutes for PPO clinic patients.Methods. All encounters were tracked from June2017 through April 2018 for a single-institution 1-dayper week PPO clinic. Administrative minutes andPPO inpatient time were excluded. Billable and non-




billable (e.g. care coordination) minutes were re-corded. Descriptive statistics were conducted. Theoverall ratio of NB:B minutes and ratios by diagnosistype and vital status were calculated. One-way ANOVAand chi-square tests were used to assess differences inthe NB:B ratios.Results. Out of 98 patients, PPO had billable visitson 54 (55%) and assisted without billing in thecare of 44 (45%). Twenty-four (25%) patients aredeceased; vital status did not differ by diagnosistype (p¼0.29). Patients had solid tumors (ST; 42,43%), brain tumors (BT; 33, 34%), leukemia/lym-phoma (L/L; 21, 21%), and other diagnoses (2,2%). Overall NB:B ratio was 1.03. NB:B ratiosdiffered among diagnoses (p<0.0001), with L/Lthe highest at 2.5 compared to ST (0.9), BT (0.8)and other (0.5). Deceased patients had a higher ra-tio of NB:B minutes than alive patients (p<0.0001;1.9 vs 0.8).Conclusion. Care coordination in PPO clinic is time-intensive and grows with clinic volume. For patientswith L/L and those who were deceased, non-billableminutes outpaced billable clinical minutes.Implications for Research, Policy, orPractice. When devising a PPO outpatient program,this NB:B ratio should be accounted for in physiciantime, and personnel devoted to patient and familyassistance.

The Gang’s All Here: All-InclusiveInterprofessional Education in a Palliativeand Hospice Center (S809)Kathleen Broglio, DNP ACHPN ANP-BC CPE FPCN,Dartmouth-Hitchcock Medical Center, Lebanon, NH.Peter DiMilia, MPH, The Dartmouth Institute,Lebanon, NH. Jillian Miller, MSN, Dartmouth-Hitch-cock Medical Center, Lebanon, NH. Ruth Thomson,DO MBA HMD FACOI FAAHPM, Dartmouth-Hitch-cock Medical Center, Lebanon, NH. Melissa Garland,BSN RN, Dartmouth-Hitchcock Medical Center,Lebanon, NH. Martha Bruce, PhD MPH, Dartmouth-Hitchcock Medical Center, Lebanon, NH.

Objectives1. Describe the interprofessional education process

that included non-clinical staff.2. Discuss tools utilized to measure readiness for

interprofessional learning, team collaboration,burnout and knowledge.

3. Describe outcomes of the interprofessionaleducational intervention.

Original Research Background. Nursing assistants(NAs), housekeepers, dietary staff, volunteer coordi-nators, and clerical staff are integral team members

in inpatient palliative and hospice settings, but maynot be included in interprofessional education.Research Objectives. The primary aim of this studywas to evaluate the impact of interprofessional educa-tion on collaborative work practices. Secondary aimsincluded: knowledge changes, job burnout, andretention.Methods. Staff in a new palliative and hospice carecenter participated in a 12-session interprofessionaleducation program. Participants’ preparationincluded discipline specific reading materials. Sessionformat focused on group exercises to maximize inter-action among disciplines. All participants were evalu-ated pre/post education and at 3-month follow-upusing the following tools: Readiness for Interprofes-sional Learning Survey (RIPLS), Assessment of Inter-professional Team Collaboration Survey (AITCS),and Maslach Burnout Inventory (MBI). Registerednurses (RNs) and NAs also completed knowledge sur-veys. RNs also completed the End of Life ProfessionalCaregivers Survey (EPCS).Results. RNs (n¼ 15), NAs (n¼ 4), housekeepers (n¼2), cooks (n¼ 3), a volunteer coordinator (n ¼ 1), andclerical staff (n¼ 3) participated in the education. A sig-nificant increase in participant AITCS and EPCS scoreswas observed post education and sustained through3-month follow-up (AITCS 16.4 percent increase at3-month, p<0.0001; EPCS 17.7 percent increase at3-month, p<0.0001). RNs (15.7 total point increase95% CI (6.3, 25.1)) and NAs (4.3 total point increase95%CI(0.3, 8.2)) both demonstrated significantlyimproved scores on knowledge surveys and, and after 3months, reported significantly reduced emotionalexhaustion(20.0, 95%CI(14.9, 25.2)) anddepersonaliza-tion (6.0, 95%CI (3.2, 8.9)).Conclusion. Our interprofessional education inter-vention has improved overall team function, whichmay ultimately affect patient outcomes and experi-ence with care.Implications for Research, Policy, orPractice. The education’s sustained impact will beevaluated at 6 months follow-up. Interprofessional ed-ucation will be studied among diverse groups withinthe broader medical center and region.

General Practitioners’ Barriers andFacilitators to Opioid Prescription inMedellin, Colombia (S810)Daniela Castano, MD, Universidad Pontificia Bolivari-ana, Medellin, Colombia. Andrea Echavarria Barboza,MD, Universidad del Rosario, Bogota, Colombia. AnaMaria Arango, MD, Medicina, Medellin, Colombia.Carolina Jaramillo, MD, Pontifical Bolivarian




University, Medellin, Colombia. Alicia Krikorian, PhD,Universidad Pontificia Bolivariana, Medellin,Colombia.

Objectives1. Identify the attitudes and specific practices of

general practitioners regarding opioidprescription.

2. Recognize general practitioners’ perceived bar-riers and facilitators of opioid prescription.

3. Describe how general practitioners’ knowledgeof opioids influence opioid prescription.

Original Research Background. Understanding keyattitudes, barriers and facilitators of opioid prescrip-tion can help overcome the current undertreatmentof pain in Low and Middle-Income Countries(LMICs).Research Objectives. Identify barriers and facilita-tors for opioid prescription by GPs in Medellin,Colombia.Methods. Descriptive-quantitative, cross-sectionalstudy. A 53-item questionnaire was designed toassess: 1) attitudes (concerns and confidencewhen prescribing opioids), 2) practices related toopioid prescription, 3) perceived barriers and fa-cilitators of opioid prescription; and 4) knowledgeof opioids. We surveyed GPs who graduated frommedical schools in Medellin, and GPs currentlypracticing in ten institutions in Medellin,Colombia. 179 participants completed the ques-tionnaire. Descriptive and correlational analyseswere conducted.Results. The mean age was 33.3 years old (SD 10.43),53.3% were female, 47.8% had been practicing forover 5 years, 51.66% did not receive training in painmanagement during medical school and only 2.8%received training after medical school. Regardingknowledge, 49.5% responded accurately. Commonconcerns were managing adverse effects and potentialopioid abuse. Participants felt less confident to pre-scribe opioids to pediatric, pregnant patients, andthose with history of substance abuse. Common bar-riers to opioid prescription were restriction to access,high costs, and insufficient training. Adequate opioidprescription practices were significantly and positivelycorrelated with confidence when prescribing opioids(0.466; p<0.01), knowledge of opioids (0.422;p<0.01), age (0.233; p<0.01), and training in painmanagement (0.308; p<0.01); and significantly andnegatively correlated with perceived barriers (-0.332;p<0.01).Conclusion. Adequate practices regarding opioid useare related to training, knowledge and confidencewhen prescribing opioids. However, GPs in ourcontext have insufficient knowledge about pain con-trol and opioid use.

Implications for Research, Policy, orPractice. These results indicate the urgent need to in-crease pain management training for current andfuture GPs.

Efficacy Variables in Cancer VersusNoncancer Patients Treated withMethylnaltrexone or Placebo: An Analysisof 2 Placebo-Controlled Studies (S811)Bruce Chamberlain, MD HMDC FACP FAAHPM, Palli-ative Consulting, Davenport, IA. Michelle Rhiner,DNP MSN ACHPN RN-BC GNP-BC, Loma Linda Uni-versity Medical Center, Loma Linda, CA. Neal Slatkin,MD, University of California Riverside School of Med-icine, Riverside, CA. Nancy Stambler, DrPH, ProgenicsPharmaceuticals Inc., New York, NY. Robert Israel,MD, Bausch Health Companies, Laval, Quebec,Canada.

Objectives1. Describe that methylnaltrexone is effective in

treating opioid-induced constipation in patientswith advanced illness with and without active can-cer regardless of baseline opioid requirements.

2. Articulate that treatment with methylnaltrexoneimproves laxation response, has a fast onset oflaxation in patients with or without cancer, andreduces the need for rescue laxatives versusplacebo.

3. Describe that methylnaltrexone taken everyother day or as needed is effective in patientswith advanced illness.

Original Research Background. Methylnaltrexone(MNTX) inhibits opioid peripheral adverse effects.Research Objectives. Post-hoc analysis of pooleddata from randomized, double-blind studies andopen-label extensions of adults with advanced illnessand opioid-induced constipation.Methods. Patients received SC MNTX 0.15 mg/kg orplacebo (study 302) and SC MNTX 8 mg (38e<62kg), 12 mg ($62 kg), or placebo (study 4000) everyother day for 2 weeks and MNTX (same doses asneeded) during the first 2 weeks of open-label exten-sions. Double-blind populations were stratified bythose with/without cancer. Endpoints includedrescue-free bowel movements (RFBM) within 4 hoursafter each dose for $2 of the first 4 doses; time torescue-free laxation; rescue laxatives use; and $3RFBMs/week with $1 RFBM/week increase in $3 of4 weeks.Results. Median baseline opioid use (mg/day) wasgreater in cancer (187.9 placebo [n¼114]; 180.0MNTX [n¼116]) versus non-cancer patients (80.0 pla-cebo [n¼71]; 120.0 MNTX [n¼62]). MNTX signifi-cantly (P<0.0001) improved the proportion ofcancer (56.9%) and non-cancer (58.1%) patients



with an RFBM within 4 hours after each dose for$2 ofthe first 4 doses versus placebo (5.3% cancer; 11.3%non-cancer). Median time to laxation was significantly(P#0.0002) shorter in cancer (0.96 hours) and non-cancer (1.25 hours) patients 24 hours after the firstdose versus placebo ($23 hours). Rescue laxativeswere used by 39.7% of cancer and 30.6% of non-can-cer MNTX patients versus 51.8% and 39.4% of pla-cebo patients. Of 108 open-label extension double-blind MNTX patients, 79 (73.1%) achieved $3RFBMs/week with $1 RFBM/week increase in $3 of4 weeks versus 48 (46.6%) of 103 double-blind placebopatients (data from double-blind and 2 weeks of open-label).Conclusion. MNTX improved laxation with a fasteronset and reduced rescue laxative use.Implications for Research, Policy, orPractice. These data support the efficacy of MNTXin cancer/non-cancer patients.

Treatment with Methylnaltrexone inPatients with Opioid-Induced Constipationwith or Without Active Cancer (S812)Bruce Chamberlain, MD HMDC FACP FAAHPM, Palli-ative Consulting, Davenport, IA. Michelle Rhiner,DNP MSN ACHPN RN-BC GNP-BC, Loma Linda Uni-versity Medical Center, Loma Linda, CA. Neal Slatkin,MD, University of California Riverside School of Med-icine, Riverside, CA. Nancy Stambler, DrPH, ProgenicsPharmaceuticals Inc., New York, NY. Robert Israel,MD, Bausch Health Companies, Laval, Quebec,Canada.

Objectives1. Describe that methylnaltrexone is equally effec-

tive in treating opioid-induced constipation inpatients with advanced illness with and withoutactive cancer.

2. Articulate that treatment with methylnaltrexonedoes not increase pain scores in patients treatedwith opioids for pain due to advanced illnesses in-dependent of cause.

Original Research Background. Subcutaneous(SC) methylnaltrexone is approved for opioid-induced constipation (OIC) in adults with chronicnon-cancer pain and OIC in adults with advancedillness or with active cancer who require opioid dosageescalation for palliative care.Research Objectives. Post hoc analysis of pooleddata from 3 randomized studies of patients withadvanced illness and OIC.Methods. Patients received single doses of SC MNTX0.15 or 0.30 mg/kg or placebo (study 301); SC MNTX0.15 mg/kg or placebo every other day for 2 weeks(study 302); and SC MNTX 8 or 12 mg in patients38e<62 or $62 kg, respectively, or placebo every

other day for 2 weeks (study 4000). Data were stratifiedby those with/without cancer. Efficacy endpointsincluded laxation #4 hours and rescue-free laxation(RFL) #24 hours after the first dose; time to RFL;and pain scores.Results. Median baseline opioid use was higher incancer (MNTX: 190 mg/d, n¼198; placebo: 200mg/d, n¼157) versus non-cancer patients (MNTX:120.0 mg/d, n¼82; placebo: 80.0 mg/d, n¼80).MNTX significantly increased the percentage of pa-tients with a laxation response #4 hours and RFL#24 hours after the first dose in cancer (MNTX:61.1% and 71.2% vs placebo: 15.3% and 41.4%,respectively; P<0.0001) and non-cancer patients(MNTX: 62.2% and 74.4% vs placebo 17.5% and37.5%, respectively; P<0.0001). MNTX significantlyreduced the median time to RFL at 4 hours in cancer(MNTX: 1.1 h, placebo: >4 h; P#0.0001) and non-cancer patients (MNTX: 1.1 h, placebo: >4 h;P#0.0001). Mean changes in pain scores were similar(cancer patients, MNTX: �0.4 vs placebo: �0.2; non-cancer patients, MNTX: �0.4 vs placebo: �0.4).Conclusion. MNTX increased laxation responses andimproved clinical signs of constipation in OIC patientswith/without cancer.Implications for Research, Policy, orPractice. MNTX patients continued opioid treatmentwith a reduction in constipation symptoms.

Physician Use of Empathy During ClinicalPractice (S813)Amber Comer, JD PhD, Indiana University, Indianapo-lis, IN. Nina Ustymchuk, BS, Indiana University/Pur-due University Indianapolis, Indianapolis, IN. LynnD’Cruz, PT MPA, Indiana University/Purdue Univer-sity Indianapolis, Indianapolis, IN. Stephanie Bartlett,PT MS, Indiana University/Purdue University Indian-apolis, Indianapolis, IN. Lyle Fettig, MD, EskenaziHealth, Indianapolis, IN.

Objectives1. Discuss mixed methods research outcomes of

how and when physicians use empathy when in-teracting with their patients during clinicalpractice.

2. Demonstrate the importance of empathy and itsuse during clinical practice.

Original Research Background. The use ofempathy during clinical practice is paramount todelivering quality patient care and is importantfor understanding patient concerns at both thecognitive and affective levels. Physician use ofempathy is associated with better patient and fam-ily experiences, higher patient satisfaction,increased patient compliance, and trust.Conversely a lack of empathy may adversely impact




a patient’s treatment course or impede communi-cation about prognosis.Research Objectives. To determine how and whenphysicians use empathy when interacting with theirpatients.Methods. A cross-sectional survey of 76 physiciansworking in a large urban hospital was conducted inAugust of 2017. Physicians were asked a series of ques-tions with Likert scale responses as well as asked torespond to open-ended questions.Results. All physicians self-report that they always(42.1%) or usually (57.9%) use empathic statementswhen engagingwith patients. 98.6%ofphysicians believethat their colleagues always (19.2%) or usually (79.5%)use empathic statements when communicating with pa-tients. Almost half of physicians indicated that usingthe words ‘‘I understand’’ denotes an empathic state-ment. 68% of physicians report that they would not liketo receive more training or assistance about how andwhen to use empathy in the health care setting.Conclusion. Although almost all physicians reportthat they and their colleagues use empathic state-ments while engaging with patients, almost half ofphysicians surveyed identified that telling a patient‘‘I understand’’ is an expression of empathy; however,using the words ‘‘I understand’’ is traditionally notconsidered to be an expression of empathy. Althoughalmost half of all physicians could not correctly iteratewords of empathy, the majority of physicians reportthat they are not interested in receiving more trainingabout how and when to use empathy.Implications for Research, Policy, orPractice. This study emphasizes the need for physi-cian education on both the importance and applica-tion of empathy during clinical practice.

‘‘No Really, Dr. Surgeon, I Am NOT Hereto Kill Your Patient; Let’s Collaborate!’’:The Development of Surgical OncologyEducation Resources for Palliative CareProviders in the Perioperative Setting(S814)Allyson Cook, MD, Stanford University, Palo Alto, CA.Rebecca Aslakson, MD PhD FAAHPM FCCP, StanfordUniversity School of Medicine, Stanford, CA.

Objectives1. Describe the process used to develop surgeon

and palliative care clinician approved learningmaterials focusing on diagnosis and surgicalintervention to be used in the perioperativesetting by palliative care clinicians in a multi-cen-ter study of perioperative palliative care.

2. Present the obstacles to consensus and how thematerials were finalized, approved, and opera-tionalized for use in a multi-center study by allstakeholders.

Original Research Background. Evidence supportssignificant cultural barriers to palliative care in surgi-cal settings. Development of palliative care interven-tions targeting surgical patients and family membersmay benefit from close partnership between surgicaland palliative care clinicians.Research Objectives. To facilitate close engagementbetween surgical, oncologic, and palliative care clini-cians to agree on key, and sometimes controversial, in-formation concerning surgical oncologic diagnosesand operations for the use of palliative care cliniciansproviding palliative care in the perioperative settingfor patients with upper GI cancers preparing forsurgery.Methods. Perioperative palliative care goals, bar-riers, and approaches were discussed among patient,family member, surgeon, oncologist, anesthesiolo-gist, and palliative care clinician team members ata stakeholder summit. Based on these discussions,surgeon, palliative care physician and anesthesiolo-gist team members developed one-page informationsheets that were iteratively revised through adetailed feedback process involving front-line pallia-tive care clinicians, oncologists, anesthesiologists,and surgeons.Results. Sheets were developed for the seven diagno-ses of pancreatic adenocarcinoma, pancreatic neuro-endocrine tumors, hepatocellular carcinoma, gastriccarcinoma, cholangiocarcinoma, and esophagealadenocarcinoma and the five surgeries of pancreato-duodenectomy, distal pancreatectomy, hepatectomy,gastrectomy, and esophagectomy. Sheet contentinvolving prognostic information was controversialwith regards to framing and content discussed,requiring multiple iterations. After five iterativedrafts, the final content was approved by all stake-holders, including surgeons, and was operationalizedfor use in a multi-center randomized controlled trialof perioperative palliative care.Conclusion. Close engagement between stake-holders can facilitate acceptance and utilization ofpalliative care, even in settings where significant cul-tural barriers exist and when content includes poten-tially contentious issues.Implications for Research, Policy, orPractice. Barriers to palliative care can be addressedand overcome through close, open and iterative feed-back between key stakeholders.



Provider Perceptions of ImplementingHome-Based Palliative Care as a CoveredHealth Benefit (S815)Alexis Coulourides Kogan, PhD MSG, University ofSouthern California, Los Angeles, CA.

Objectives1. Describe the provider-level barriers and facilita-

tors to implementing a home-based palliativecare program within a community-based hospiceagency.

2. Describe several strategies for implementing ahome-based palliative care program within acommunity-based hospice agency.

Original Research Background. Despite repeatedstudies demonstrating that home-based palliativecare (HBPC) can improve patient outcomes whiledecreasing costs of care, replication of this model infee-for-service has been stymied by a lack of reimburse-ment structure. To overcome this barrier, a large Cali-fornia-based health insurer has begun to reimbursecontracting medical group providers for HBPC. Littleis known about the provider-level experience of devel-oping and implementing a reimbursable HBPC pro-gram as part of a community-based hospiceorganization.Research Objectives. The purpose of this qualitativestudy was to explore the provider-level impact of im-plementing HBPC as a covered health benefit.Methods. Focus groups were conducted among threeinterdisciplinary HBPC teams from community-basedhospice organizations implementing the HBPC pro-gram and serving geographically and culturally diversepatient populations.Results. Participants consisted of physicians (8%),nurse practitioners (23%), registered nurses (23%),social workers (30%), chaplains (8%), and patient co-ordinators/liaisons (8%). The majority of partici-pating providers were Caucasian (85%) females(85%). Qualitative data were analyzed using groundedtheory and results revealed three independentthemes: 1) Referrals to the HBPC program; 2) Organi-zational factors; and 3) Reimbursement for HBPC.Findings highlight barriers and facilitators to imple-menting HBPC such as: impact of the organization’sreputation in the community, the dynamic and ‘‘team-iness’’ of the HBPC team, having a site champion, andissues associated with working in a siloed medical sys-tem. Participants across all study sites also discussedseveral challenges with patient referrals that focusedon a lack of knowledge of palliative care (both physi-cians and patients) and physicians’ lack of communi-cation with the patient being referred to HBPC.Conclusion. Findings highlight barriers and facilita-tors to implementing HBPC and a set of implementa-tion strategies has been developed.

Implications for Research, Policy, orPractice. Findings hold implications for practiceand widespread replication of the HBPC model asother health insurers and CMS look for effectiveways to support their members with serious illness.

Access to Pediatric Palliative Care in theNeonatal Intensive Care Unit by Minoritiesand Rural-Dwellers in the Deep South:Patterns of Care from 2009-2017 (S816)Erin Currie, RN, University of Alabama at Birming-ham, Birmingham, AL. Joanne Wolfe, MD MPHFAAHPM, Dana-Farber Cancer Institute, Boston, MA.Renee Boss, MD, Johns Hopkins University School ofMedicine, Baltimore, MD. Deborah Ejem, PhD, Uni-versity of Alabama at Birmingham, Birmingham, AL.Sam Perna, DO, University of Alabama at Birming-ham, Birmingham, AL. James Dionne-Odom, PhDRN ACHPN, University of Alabama at Birmingham,Birmingham, AL. Susan Buckingham, MD, Universityof Alabama at Birmingham, Birmingham, AL.Kathleen McKillip, MD, Creighton University, Omaha,NE. Marie Bakitas, DNSc NP-C FAAN, University ofAlabama at Birmingham School of Nursing, Birming-ham, AL.

Objectives1. Describe patterns of pediatric palliative care and

care disparities in neonatal intensive care unit(NICU) patients and families hospitalized inthe Deep South (Alabama, Louisiana, andMississippi).

2. Describe implications for practice related to cur-rent trends of pediatric palliative care use in theNICU and implications for future research.

Original Research Background. Pediatric mortalityis the highest in the first year of life. In 2013, 23,446infants died in the U.S. Access to timely pediatric palli-ative care (PPC) services are limited for seriously ill in-fants and their families in the U. S. Deep South.Patterns of PPC in the neonatal intensive care unit(NICU) and the extent of PPC disparities in theDeep South are unknown.Research Objectives. Examine racial and geograph-ical differences in pediatric palliative care (PPC)consultation for seriously ill infants in the neonatalintensive care unit.Methods. This was a retrospective medical record re-view of infant decedents who received PPC while hos-pitalized in a level III NICU at an academic children’shospital in Alabama from 2009-2017. Demographiccharacteristics, timing of palliative care consultation,hospice enrollment, final resuscitation status, circum-stances at time of death, and interventions receivedin the last 48 hours of life. Data were analyzed inSPSS using descriptive statistics and t-tests.




Results. The percentage of infants receiving PPC inthe NICU increased over time from 7% in 2009 to38% in 2017. Infant decedents (N¼140) who receivedPPC in the NICU were mostly Caucasian (58%) andAfrican American (39%), receiving Medicaid (84%),and had genetic (53%) and prematurity (34%) diag-noses. There were no statistically significant differ-ences between racial or urban versus rural groups inthe timing of PPC consultation during the NICUadmission. Infants who lived over 1 hour away receivedPPC significantly later than infants living less than 1hour away from the NICU (p¼ 0.03).Conclusion. There were no racial or rurality differ-ences in PPC timing during hospitalization; however,traveling over an hour to the hospital was associatedwith a delay in receiving PPC.Implications for Research, Policy, orPractice. Interventions tailored to reduce disparitiesin timely PPC in the Deep South may need to accountfor families living great distances from their hospital-ized infant.

‘‘Why Would You Choose Death?’’: HeartFailure Patient Attitudes RegardingPalliative Care (S817)Brett Curtis, BS, University of Pittsburgh School ofMedicine, Pittsburgh, PA. Dio Kavalieratos, PhD, Uni-versity of Pittsburgh, Pittsburgh, PA.

Objectives1. Discuss relationships among attitudes toward

palliative care, advanced care planning and caresatisfaction among patients with heart failure.

2. Identify implications for PC implementation inHF care.

Original Research Background. Patient-level fac-tors potentially influencing perceived need for pallia-tive care (PC) in heart failure (HF) remain unclear.Research Objectives. Explore HF patients’ attitudestoward PC, including self-defined triggers for specialtyPC.Methods. Semi-structured interviews exploring pallia-tive needs, the extent to which those were met withincurrent HF management, and preferences regardingPC initiation. Two investigators independentlyanalyzed transcript data using thematic analysis. TheKansas City Cardiomyopathy Questionnaire (KCCQ)was administered to measure symptom burden.Results. 28 patients recruited from a quaternary carehospital were interviewed. The average participant was63 years old, male and Caucasian with 3.4 symptomsand KCCQ score of 39. 71% (n¼20) had advanced dis-ease (NYHA III/IV). After being read a definition ofPC expressing its role in symptom control and qual-ity-of-life across the illness trajectory, most viewed itfavorably. However, participants also expressed

preferences to delay specialty PC involvement untiltheir disease became terminal. Other themes include:(1) exhaustion of treatment options, and loss of abilityto perform activities-of-daily-living as triggers for spe-cialty PC involvement; (2) lack of relationship betweensymptom burden and advance care planning activities;(3) general satisfaction with HF management despiteidentifying gaps (e.g. social services management) intreatment.Conclusion. Our results suggest HF patients, despitepositively viewing PC as an option for symptom controlacross HF’s disease course, prefer to utilize PC solelyfor end-of-life care.Implications for Research, Policy, orPractice. Efforts are needed to negate patient reluc-tance to PC across the illness trajectory, as patientsmay believe PC is reserved exclusively for terminalcare.

Spirituality and Religiosity and Burnoutin Latin-American Palliative Care HealthCare Professionals (LAPC) (S818)Marvin Delgado Guay, MD, The University of TexasMD Anderson Cancer Center, Houston, TX. MariaMargarita Reyes Donoso Universidad Catolica deChile, Santiago, Chile. Jos�e Mario L�opez Saca HospitalNacional Dr. Juan Jos�e Fern�andez Zacamil, San Salva-dor, El Salvador. Miriam Elisa Riveros Rios, MD, Uni-versidad Nacional de Asuncion, Asuncion, Paraguay.Tania Pastrana, MD, RWTH Aachen University, Aa-chen, Germany.

Objectives1. Identify demographic factors related to spiritual-

ity and religiosity in PC health care providersfrom Latin America.

2. Identify factors related to burnout in PC healthcare providers and the relationship with spiritual-ity and religiosity.

Original Research Background. Spirituality(S) andreligiosity(R) are common with Latino cultural values.These elements are essential in delivering QualityPalliative Care (PC). There is limited literatureregarding Latin American clinicians’ spiritual and reli-gious characteristics, or how these commitmentsshape their clinical engagement and presence ofburnout.Research Objectives. To describe the frequency, in-tensity and importance of self-reported S and R andburnout on the clinical practice of LAPC.Methods. From 6/1, to 12/31, 2017, a cross-sectionalstudy using an anonymous and voluntary Online Sur-vey was provided to active members of PC-Latin Amer-ican-Association. We collected and analyzed dataregarding demographics, personal and professionalrole of S and R and burnout.




Results. 221/353 members from 20 Latin AmericanCountries participated, Response rate 63%. Medianage 47(SD+/-12), 75% were women. 40% were Cath-olic. 58% were physicians, 19% nurses, 12% psychol-ogy, and other12%. The median time of working inPC was 9 years (+/-7). LAPC considered themselvesspiritual (median: 8/10, range 0-10) and religious(5, 0-10). LAPC considered S/R very important intheir lives (9/10, 0-10 and 6/10, 0-10), respectively.LAPC reported that S/R was a source of strengthand comfort (9/10, SD+/-2), helped them to copewith their problems (8/10, SD+/-3), and helpedthem to keep their quality of life in a stressful workenvironment (8/10, SD+/-4), significant in thosebelonging to a church community (p¼0.000), timeworking in PC (p¼0.01), age (p¼0.03). 190/221(86%) reported strongly/somewhat agreed withthe statement: ‘‘I feel called to take care of patientswho are dying’’. 31/221(14%) reported being Burnedout. No significant difference among gender, profes-sion, age, years in profession or in PC, or importanceof spirituality and religion.Conclusion. Most of LAPC considered themselvesspiritual and religious. Low percentage of LAPC re-ported presence of burnout. The ‘‘call’’ and feelingenergized caring patients dying are associated withless burnout. More research is needed.Implications for Research, Policy, orPractice. Spirituality and Religiosity helping todecrease burnout in Latin American PC.

Validation of the Palliative PerformanceScale (PPS) to Predict Survival of OlderAdults Admitted to the Hospital from theEmergency Department (S819)Sarah DeWitt, MD, Medical University of South Car-olina, Charleston, SC. Jonas Te Paske, MD, EliteWomen’s Health, Fredericksburg, VA. Shana Sem-mens, MD, Banner University Medical Center, Tucson,AZ. Robin Hicks, DO, University of Pittsburgh MedicalCenter Pinnacle Health, Harrisburg, PA. Leigh Vaugh-an, MD, Medical University of South Carolina,Charleston, SC.

Objectives1. Identify those patients admitted from the emer-

gency department who have a high 6 month mor-tality rate and could benefit from early Palliativeintervention.

2. Recognize that patients who have low baselinefunctional status will have a lower 6 month sur-vival when admitted to the hospital from theemergency department.

Original Research Background. Emergency clini-cians have a role in early prognostication. Babcocket al. (2016) validated the PPS among adult patients

admitted to the hospital through the ED. Theydemonstrated the content validity that PPS of 0-30and 40-60 predict 6 month survival of 14% and 48%,respectively. However, their study was limited by amodest sample size at a single hospital of a predomi-nantly white affluent patient population. We hypothe-size that decreasing baseline PPS score will correlatewith decreasing survival and that PPS will be a predic-tor of mortality. If the PPS can discern those at highrisk of death it may provide a method to identify thosepatients who might benefit from a goals-of-care con-versation prior to hospital admission.Research Objectives. To evaluate the construct val-idity of the Palliative Performance Scale (PPS) as ameasure that can stratify the 6 month survival of olderadults admitted to an urban university hospital fromthe emergency department.Methods. Adults >55 years admitted from the EDwere interviewed by investigators on day of admission.Baseline PPS assessed and on admission and followedup at 6 months.Results. One hundred and forty five participantswere enrolled, 129 participants accounted for and16 were lost at the end of 6 month follow-up. Survivalat 6 months as follows: 86% survival of those withinitial PPS of 70-100 (13/95 died), 58% survival ofthose with initial PPS of 40-60 (18/43 died), and28% survival of those with initial enrollment PPS of10-30 (5/7 died) (Chi-squared statistic ¼ 21.15,p¼.000026).Conclusion. Palliative performance scale is validatedtool for predicting mortality at 6 months and can beused to screen patients admitted from the ED whocould benefit from Palliative care consult.Implications for Research, Policy, or Practice. P-CaRES tool in combination with PPS could helpdefine this population further and additional studiesmay be of benefit.

How Do Internal Medicine ResidentsPerceive Direct Observation for Educationin Goals-of-Care Communication? (S820)Laura Dingfield, MD, University of PennsylvaniaHealth System, Philadelphia, PA. Corrie Stankiewicz,MD MSEd, University of Pennsylvania, Philadelphia,PA. Rachel Miller, MD, Perelman School of Medicine,University of Pennsylvania, Philadelphia, PA. ClaireBocage, BA, University of Pennsylvania, Philadelphia,PA. Whitney Eriksen, PhD, University of Pennsylvania,Philadelphia, PA.

Objectives1. Describe resident perceptions of direct observa-

tion of goals-of-care communication.2. Describe barriers to direct observation of goals-

of-care communication.




Original Research Background. Effective communi-cation with patients and families is a core competencyof residency education. While internal medicine resi-dents frequently hold goals-of-care discussions with pa-tients and families, they report inadequate training ingoals-of-care communication (GOCC). Direct observa-tion presents opportunities for formative feedbackand competency-based assessment of resident GOCC.However, resident perceptions of direct observation asa teaching modality for enhancing goals-of-carecommunication have not been described.Research Objectives. To describe resident perspec-tives on use of direct observation as a method forimproving GOCC among residents.Methods. Fifteen semi-structured interviews were con-ducted and recorded. Recordings were transcribed, de-identified, and thematically analyzed in grounded the-ory framework using NVivo 11. A codebook represent-ing salient themes was created. Two reviewersestablished strong post-inter-rater reliability, k ¼.98 with 3 (20%) of the interviews.Results. Residents broadly shared negative attitudestowards direct observation as a learning tool, such asfeeling uncomfortable or anxious or that the experi-ence was artificial (53%). However, they reported awillingness to be observed, noting that the observationexercise is valuable in their development of GOCCskills (53%). The majority of residents describedGOCC as ‘‘higher stakes’’ than other patient-providercommunication (60%). Careful preparation and struc-turing of the direct observation encounter helpedmitigate resident unease about direct observation ofGOCC. Residents expressed preference for directobservation by faculty with expertise in GOCC(53%). The primary barrier to direct observation ofGOCC was time conflicts (60%).Conclusion. Residents described the experience ofdirect observation for GOCC as an uncomfortable,yet useful exercise due to the higher stakes nature ofthese discussions. The challenges to conducting directobservations did not stem from resident unwilling-ness, but rather logistics.Implications for Research, Policy, orPractice. Direct observation of GOCC should beencouraged, and procedures should be put in placeto set expectations and allow for time for observationand feedback surrounding resident GOCC.

How Family Caregivers Assist withUpstream Healthcare Decision-Making byCommunity-Dwelling Persons withAdvanced Cancer: A Qualitative Study(S821)James Dionne-Odom, PhD RN ACHPN, University ofAlabama at Birmingham, Birmingham, AL. Deborah

Ejem, PhD, University of Alabama at Birmingham, Bir-mingham, AL. Rachel Wells, MSN RN CNL, Universityof Alabama at Birmingham, Birmingham, AL. AmberBarnato, MD, Geisel School of Medicine at Dart-mouth, Lebanon, NH. Richard Taylor, DNP CRNPANP-BC, University of Alabama at Birmingham, Bir-mingham, AL. Gabrielle Rocque, MD, University ofAlabama at Birmingham, Birmingham, AL. YaseminTurkman, PhD MPH CRNP, University of Alabama atBirmingham, Birmingham, AL. Thomas Ramsey,PhD, University of Alabama at Birmingham, Birming-ham, AL. Matthew Kenny, MPH, University of Alaba-ma at Birmingham, Birmingham, AL. NataliyaIvankova, PhD MPH, University of Alabama at Bir-mingham, Birmingham, AL. Marie Bakitas, DNScNP-C FAAN, University of Alabama at BirminghamSchool of Nursing, Birmingham, AL. Michelle Martin,PhD, University of Tennessee Health Science Center,Memphis, TN.

Objectives1. Describe the roles of family caregivers in patients’

healthcare decision-making in the context ofadvanced cancer.

2. Describe two implications for outpatient andcommunity-based early oncology palliative careconcerning enhancing decision support for fam-ily caregivers.

Original Research Background. In the palliativecare context, the family caregiver role in patients’healthcare decision-making has focused on being asurrogate decision-maker at end-of-life. Less is knownabout family caregiver’s role in supporting upstreampatient decision-making in advanced cancer.Research Objectives. Describe how family membersassist community-dwelling relatives with advanced can-cer with current and prospective healthcare decisions.Methods. Qualitative descriptive study consisting ofone-on-one, semi-structured interviews with personswith metastatic cancer and their family caregivers.We elicited family members’ perspectives on howthey assist their relatives with any current and prospec-tive healthcare decisions. Transcribed interviews wereanalyzed using a thematic analysis approach. Co-investigators reviewed and refined themes.Results. Caregivers (n¼20) averaged 56 years of ageand were mostly female (95%), White (85%), andthe patient’s partner/spouse (70%). Patients (n¼18)averaged 58 years of age and were mostly male(67%) in ‘‘fair’’ or ‘‘poor’’ health (50%) with genito-urinary (33%), lung (17%), and hematologic (17%)cancers. Themes describing family member roles insupporting patients’ decision-making were: 1) seekinginformation about the cancer, its trajectory, anddifferent treatments options; 2) identifying treatmentand disease decision points, including decisions about



seeking emergent care; 3) ensuring family membershave a common understanding of the patient’s planof care; 4) initiating and facilitating conversationswith patients about coping, values, beliefs, and ‘‘whatif’’ scenarios about current and potential future healthstates and treatments; 5) implementing choices (e.g.,providing transportation) and addressing ‘‘spillover’’decisions (e.g., work arrangements) resulting frommedical treatment choices; and 6) making upstreamhealthcare decisions on behalf of patients whopreferred to have decisions made by their familycaregivers.Conclusion. These data highlight a previously unre-ported and understudied set of critical decision part-nering roles that cancer family caregivers play inpatient healthcare decision-making.Implications for Research, Policy, orPractice. Optimizing these roles may represent noveltargets for early palliative care decision support inter-ventions for family caregivers.

A Codified Process for MultidisciplinaryTeam Consensus Around the Terminationof Life Sustaining Treatments (LST) inFrance: An Interview Study (S822)Elizabeth Dzeng, MD PhD MPH, University of Califor-nia, San Francisco, San Francisco, CA. Nancy Kentish-Barnes, PhD, Saint Louis Hospital, Paris, AB. J RandallCurtis, MD MPH, University of Washington, Seattle,WA. Anne-Sophie Debue, MSc, Assistance PubliqueeHopitaux de Paris, Paris, France.

Objectives1. Describe three ways that the LAT process used at

the Paris hospitals in this study enabled nursesand other allied health professionals to feel em-powered to co-create consensus surroundingLST decision-making.

2. Demonstrate how a LAT-type process might beapplied to an American context to improveconsensus, ethical decision-making, and nurse/staff empowerment.

Original Research Background. In 2005, Frenchlawmakers passed the Claeys-Leonetti (C-L) Law,which prohibits futile care and authorized with-holding or withdrawing (WD/WH) of futile treat-ments. This law allowed patients to refuse futiletreatments and physicians to WD/WH LST whereappropriate, and provided a framework for mandatorymultidisciplinary team consensus around LST.Research Objectives. The objective of this study wasto understand how the C-L law influences multidisci-plinary team dynamics, clinician empowerment, andthe intensity of end-of-life care.Methods. Semi-structured in-depth interviews wereconducted with 13 physicians and 6 nurses (with

additional interviews underway) at two hospitals inParis. Participants were purposively sampled byseniority and profession/specialty to provide a rangeof perspectives and contribute to understandingemerging patterns and themes. Transcripts wereanalyzed using thematic analysis.Results. Decisions to WH/WD LST are achievedthrough consensus of the entire multidisciplinaryteam including physicians, nurses, and at times otherallied health professionals. Meetings to decide upontermination of LSTs (r�eunion de Limitation et Arretdes Th�erapeutiques Actives (LAT)) are an importantpart of the process and can be called by any teammember. Treatment decisions generally do not pro-ceed until every team member is in agreement. Thisprocedure improved nurse and junior physicianempowerment, although the degree to which individ-uals felt comfortable/empowered to speak up wasvariable.Conclusion. Since the passage of the C-L Law, deci-sion-making practices at two Parisian hospitals haveadapted procedures that provide time and space toachieve consensus amongst the entire interdisci-plinary team. These procedures encourage consensusand ethical decision-making around WD/WH of LSTs.More research is needed to determine how best toimplement interdisciplinary consensus and the impacton quality of decision-making.Implications for Research, Policy, or Practice. Anintervention using the LAT procedures could improvemultidisciplinary team consensus and improve nurseand junior physician empowerment around end-of-life decision-making in the United States.

Defining Palliative Opportunities inPediatric Patients with Bone and SoftTissue Tumors (S823)Jonathan Ebelhar, MD, Emory University, Atlanta, GA.Kristen Allen, MPH, Children’s Healthcare of Atlanta,Atlanta, GA. Karen Wasilewski-Masker, MD, Emory/Children’s Healthcare of AtlantaeAFLAC Cancer Cen-ter, Atlanta, GA. Katharine Brock, MD MS, Emory Uni-versity, Children’s Healthcare of Atlanta, Atlanta, GA.

Objectives1. Define and recognize palliative opportunities

that occur during a patient’s disease course,including how many and when these occur in apatient’s course with a bone/soft tissue tumor.

2. Describe the palliative opportunities that pre-ceded a palliative care consultation and thetiming of palliative care consultation duringthese patients’ illness.

Original Research Background. Pediatric patientswith solid tumors have many opportunities forincreased primary or specialty palliative care (PC).




However, how many, when they occur, and if they arecorrelated with other factors are unknown.Research Objectives. To define palliative opportu-nities within pediatric cancer, and explore how theseoccur in patients with solid tumor.Methods. A priori, nine palliative opportunity cate-gories were defined (disease progression and relapse,hospital admission for symptoms or social concerns,intensive care or marrow transplant admission, phase1 trial or hospice enrollment, DNR status). A single-center retrospective review was conducted on patientsaged 0-17 years at diagnosis with bone/soft tissue tu-mors who died from 1/1/12-11/30/17. Demo-graphic, disease, and treatment data was collected,and descriptive statistics were performed. Timing ofopportunities was evaluated over quartiles from diag-nosis to death.Results. Patients (n¼60) had a mean of 9 (SD¼4)palliative opportunities. Number or type of opportu-nities did not differ by demographics or diagnosis.PC consulted on 18 patients (30%) a median of 14.0months (IQR 25.0) after diagnosis, and 2.6 months(IQR 11.5) prior to death. Likelihood of PC consultdid not differ by diagnosis or total opportunities.The opportunities that preceded PC consult were pro-gression/relapse (9/18), escalated hospital level ofcare (4/18), symptom admission (3/18), and end-of-life (EOL) concerns (2/18). Hospice was involvedfor 72% of patients. The majority of opportunitiesoccurred in the last quartile of the disease course (me-dian 5.0, IQR 5.0).Conclusion. Patients with solid tumors incur manyevents warranting psychosocial or palliative support,which increase toward the EOL. Mean reported op-portunities is likely a minimum due to stringent collec-tion methods. No palliative opportunity ordemographic variable was associated with PCconsultation.Implications for Research, Policy, orPractice. Additional work is needed to further refinewhat qualifies as a palliative opportunity, how to fullycapture opportunities, and how those may differacross different cancers.

Identifying Late-Stage Cancer and ChronicKidney Disease Patients for PalliativeCare Research and Practice: ComputablePhenotypes and Natural LanguageProcessing (S824)Natalie Ernecoff, MPH, University of North Carolinaat Chapel Hill, Chapel Hill, NC. Kathryn Wessell,BA, University of North Carolina at Chapel Hill,Chapel Hill, NC. Laura Hanson, MD MPH FAAHPM,University of North Carolina at Chapel Hill, ChapelHill, NC. Adam Lee, MBA, University of North

Carolina at Chapel Hill, Chapel Hill, NC. Stacie Duset-zina, PhD, Vanderbilt University School of Medicine,Nashville, TN. Christopher Shea, PhD, University ofNorth Carolina at Chapel Hill, Chapel Hill, NC. Mor-ris Weinberger, PhD, University of North Carolina atChapel Hill, Chapel Hill, NC. Antonia Bennett, PhD,University of North Carolina at Chapel Hill, ChapelHill, NC.

Objectives1. Describe the utility of computable phenotypes

and natural language processing in a palliativecare population.

2. Describe the research and practice implicationsof systematic identification of palliative carepatients.

Original Research Background. Systematic identifi-cation of seriously-ill patients allows palliative care re-searchers and clinicians to test new models of caredelivery. Algorithms based on clinical indicatorsdincluding natural language processingdcan aid insuch identification.Research Objectives. To develop electronic healthrecord (EHR) phenotypes to identify patients withStage 4 solid-tumor cancer (CA) or Stages 4-5 chronickidney disease (CKD).Methods. We developed two computable EHR phe-notypes to retrospectively identify decedents whohad been admitted to an academic medical center inthe last six months of life with CA or CKD, respectively.Each search included International Classification ofDiseases (ICD) 9 and 10 codes and a date of death11/07/17-12/31/17 (CA) or 11/26/17-12/31/17(CKD). Additionally, the cancer search included natu-ral language processing (NLP) searches of notes forindicators of stage 4 CA (e.g., ‘‘stage IV,’’ ‘‘metasta-tic’’); the CKD search included glomerular filtrationrate (GFR) <30. For each EHR phenotype, we calcu-lated the sensitivity, positive predictive value (PPV),and false discovery rate (FDR). Results of the pheno-types were compared to manual chart review for indi-cators of late-stage disease among patients admitted tothe Oncology and Nephrology inpatient services,respectively.Results. The EHR phenotype identified 116 CA pa-tients, of whom 84 had Stage 4 CA, and 65 CKD pa-tients, of whom 23 had Stage 4-5 CKD. The EHRphenotype for Stage 4 cancer had a sensitivity of98.8%, PPV of 79.2%, and a FDR of 20.8% whencompared to the assessment of the primary oncologyservices. The EHR phenotype for Stage 4-5 CKD hada sensitivity of 100%, PPV of 47.9%, and a FDR of52.1% when compared to the assessment of the pri-mary nephrology service.Conclusion. EHR phenotypes can efficiently identifypatients with late-stage disease for palliative care.



Implications for Research, Policy, orPractice. EHR phenotypes may shape identificationof seriously-ill patients at high risk of having palliativecare needs for both research and clinical purposes.

Integrating Palliative Care Social Workersinto Sub-Acute Settings: Feasibility of theALIGN Intervention Trial (S825)Stacy Fischer, MD, University of Colorado School ofMedicine, Denver, CO. Matthew Golub, BS BA, Uni-versity of Colorado Denver, Denver, CO. Angela Plata,MS, Kaiser Permanente, Denver, CO. Jessica Retrum,PhD, MSSW, Metropolitan State University of Denver,Denver, CO. Erin Nielsen, LCSW, The Holding Groupand University of CO Hospital, Denver, CO. Dana Lah-off, LCSW, The Holding Group, Denver, CO. BreeOwens, LCSW, The Holding Group, Denver, CO. Lau-rel Tropeano, LCSW, The Holding Group, Denver,CO. Ellie Jensen, DO, Colorado Permanente MedicalGroup, Denver, CO. Leslie Wright, MA, Kaiser Perma-nente Colorado, Aurora, CO. Wendolyn Gozansky,MD MPH, Colorado Permanente Medical Group,Denver, CO.

Objectives1. Describe the justification for methodological

approach in this pragmatic trial design.2. Discuss the justification for the ALIGN interven-

tion and preliminary results of the interventionin the SNF setting.

Original Research Background. Sub-acute rehabili-tation (SNF), intended for short stay care transitions,cares for 1/3 of older adults in the last six monthsof life with low penetration of hospice and palliativecare. ALIGN (Assessing & Listening to IndividualGoals and Needs) is a palliative care social workerled intervention aimed to improve quality of life(QOL), goals of care (GOC) alignment, and providesupport to patients and caregivers.Research Objectives. Determine the feasibility ofconducting a trial of ALIGN in older persons andtheir caregivers admitted to SNF and conduct explor-atory analysis of ALIGN vs usual care on patient goalsof care alignment (curative, life-prolonging/rehabili-tative, comfort), QOL (FACT-G), and caregiver reac-tion assessment (CRA) and burden (Zarit).Methods. To conduct a pilot pragmatic randomizedstep wedge design of the ALIGN intervention versususual care in three SNFs with 120 older adults andcaregivers (optional) admitted with advanced medicalillness (LACE score >7) to determine feasibility andpreliminary efficacy.Results. To date, 362 SNF patients met illness criteriaand the team was able to approach 127 patients withinrequired 72 hours of admission. Enrollment rate forpatients ¼ 68%, caregivers ¼ 36%. Caregivers are

often not available to participate in-person, alternateapproaches are being tested. Baseline GOC alignment¼ 50%. Baseline FACT-G ¼ 70.0 (+16.9) demon-strating compromised QOL, CRA Self Esteem scale(1-5) ¼ 1.8 (+0.5) showing some benefit finding, andZarit ¼ 14.5 (+9.2) showing moderate burden.Conclusion. A pragmatic trial of the ALIGN interven-tion is feasible and needed based on low GOC align-ment, low QOL, and moderate caregiver burden.Implications for Research, Policy, orPractice. The current model for SNF does notaddress the palliative care needs of patients, ALIGNhas potential to be an effective, scalable, reproducibleintervention for to improve palliative care outcomeswithin sub-acute settings.

The Most Common Reflections of theDying (S826)Mary Gergis, PhD MSN BSN RN, Towson University,Towson, MD. Michael Ent, PhD, Towson University,Towson, MD.

Objectives1. Discuss the importance of learning about the

most common end-of-life reflections.2. Identify the most common reflections of the

dying.Original Research Background. Reflecting onone’s life can be therapeutic for dying people. Knowl-edge about common end-of-life reflection s may in-crease caregivers’ ability to identify appropriateoccasions to initiate or facilitate therapeutic lifereview.Research Objectives. Identifying the most commonend-of-life reflections.Methods. One hundred twenty-three nurses whowere HPNA members responded to an anonymoussurvey. Participants were asked to list the mostcommonly expressed reflections that patients haveabout their lives. A coding scheme was determinedbased on themes identified in participants’ responses.Two independent raters coded all the responses. Theoverall inter-rater agreement (Cohen’s kappa) wask¼ .823. After the data were independently coded,the two raters consulted to resolve discrepancies andgenerate a final set of codes.Results. The top five end-of-life reflections themesemerged from the data were; concern for lovedones, regret, spirituality, legacy, and lack of accep-tance/readiness. Nurses were more likely to reportthat patients were concerned about their families(50.41%) than their own morbidity/mortality(36.59%), c2(1) ¼ 9.39, p ¼ .002 and were more likelyto report that patients expressed concerns about theirloved ones as opposed to gratitude for them (10.57%),c2(1) ¼ 78.06, p < .001.




Conclusion. Concern for loved ones was identified asthe most common theme of end-of-life reflections.Implications for Research, Policy, orPractice. Health care professionals should extendtheir efforts to alleviate their patients’ concernsrelated to relationships with loved ones. Futureresearch could focus on determining the specificcauses of concern for loved ones among the dying pa-tients and developing measures to ameliorate suchconcerns.

Increasing Awareness of Palliative Care inthe Latino Community (S827)Joy Goebel, PhD RN FPCN, California State UniversityLong Beach, Long Beach, CA. Mara Bird, PhD, Centerfor Latino Community Health, Long Beach, CA. Lor-ene Morris, MPS, California State University Institutefor Palliative Care, San Marcos, CA. Erika Bonilla,MPH, California State University Long BeachResearch Foundation, Long Beach, CA.

Objectives1. Discuss challenges and opportunities to

improving awareness of palliative care in under-served communities.

2. Identify potential community partners to reachunderserved populations the provider’s areas ofpractice.

Original Research Background. Research indicatessignificant disparities in the use of palliative care (PC)in underserved populations. Although 17% of Ameri-ca’s population is Latino, and Latinos account for 48%of California’s population, only 6.9% of individualsreceiving PC are Latino. Thus, culturally and linguisti-cally sensitive approaches to improve awareness of PCare imperative to improving health outcomes amongLatinos.Research Objectives. To evaluate the feasibility ofapplying the promotores model to improving aware-ness of PC in the Latino Community. To describethe implementation and evaluation of a PC promo-tores project in an urban setting.Methods. Promotores are respected Latino commu-nity members who provide health information to theirlocal communities. Familias en Acci�on (a Latino advo-cacy and educational organization) and the Centerfor Latino Community Health at California State Uni-versity, Long Beach partnered to provide 3- to 8-hourface to face trainings to promotores on PC. Each pro-motora agreed to teach 10 additional people. Pre andpost surveys and 6-month follow-up telephone inter-views evaluated program effectiveness.Results. Sixty-one promotores participated; 57completed the 6-month follow-up (93% retentionrate). The mean age for the promotoras was 47.3 years(range 19-68, standard deviation [SD] 10.01); mean

length of US residency was 26.74 years (range 6-64,SD 11.18); 56 (91.8%) were female; and 52 (85%) re-ported their country of origin as Mexico. Initially, 57(94%) promotores reported no knowledge of PC. Atthe completion of the training, 60 (98%) reportedstrongly agreed or agree with the statement, ‘‘I amready to share this information with my community.’’Promotores provided the training in various settingsincluding schools, churches, and senior centers to1,950 community members, exceeding project Objec-tives by threefold.Conclusion. This project provides evidence that pro-motores are extremely effective in disseminating PCinformation throughout their communities.Implications for Research, Policy, orPractice. Rigorous methodologies are needed to pro-vide empiric evidence of specific healthcare outcomesfrom PC promotores projects.

‘‘Supportive Care Nurses’’ An InnovativeInpatient Primary Palliative CareConsultation Service at an AcademicMedical Center (S828)Yuya Hagiwara, MD MACM, University of Iowa, IowaCity, IA. Ann Struve, DNP ARNP, University of Iowa,Iowa City, IA. Christine Harlander, MSN, Universityof Iowa Hospitals and Clinics, Iowa City, IA. Ian Row-land University of Iowa Hospitals and Clinics, IowaCity, IA. Suzanne Streitz, BA, University of Iowa Hospi-tals and Clinics, Iowa City, IA. Timothy Thomsen, MDFACS, University of Iowa, Iowa City, IA.

Objectives1. Describe a nurse-led primary palliative care

consultation service to provide early palliativesupport for patients at a tertiary academic medi-cal center.

2. Determine acceptability and perceived effective-ness of a nurse-led primary palliative care consul-tation approach.

Original Research Background. The ability to pro-vide primary palliative care is a skill set that is underu-tilized by hospitalists. Many hospitalized patients lackaccess to palliative care consultation. To fill this gap,a primary palliative care consultation service couldbe beneficial to patients, families and hospitalists inaddition to specialty level palliative care consultservices.Research Objectives.

� To develop a nurse-led inpatient primary pallia-tive care consultation service.

� To evaluate the acceptability and perceived effec-tiveness of a nurse-led primary palliative careconsultation approach.

Methods. A single-site, single-arm pilot study was con-ducted at the University of Iowa Hospital and Clinics.




Two selected registered nurses, called the supportivecare nurses (SCN), received specialized training in pri-mary palliative care over a 6 month period. We incor-porated multiple teaching strategies includinginteractive didactic sessions, reading assignments,role-playing and shadowing of the interdisciplinarypalliative care team. The project team, including twopalliative care physician and two nurse practitioners,held daily supervision meetings with the SCN. Afterthe training period, the SCN worked primarily withthe general medicine hospitalist team. Acceptabilityof the pilot program was assessed 6 months after initi-ation of the pilot program. Providers on the hospitalistteam were surveyed regarding their experience withthe supportive care nurses.Results. The SCN received two hundred sixty oneconsults over the initial 6 month pilot period. Four-teen physician and nurse practitioner providerscompleted the survey. Satisfaction with the servicewas high (79%). The SCN assisted the hospitalistteam with: symptom management (86%); bettercommunication (86%) with and understanding(79%) of patient and families; completing advancedcare planning (79%); and discharge planning (71%).Conclusion. Incorporating a nurse-led primary palli-ative care consultation service to facilitate early pallia-tive care is acceptable and effective.Implications for Research, Policy, orPractice. Further research is required to examinethe impact of a primary palliative care consult serviceon patient and family centered outcomes.

Improving Goals of Care Discussion:Innovative Curriculum Development forInternal Medicine Residents (S829)Yuya Hagiwara, MD MACM, University of Iowa, IowaCity, IA. Juan Pagan-Ferrer, MD, University of IowaHospitals and Clinics, Iowa City, IA. William Clark,MD, University of Iowa Hospitals and Clinics, IowaCity, IA.

Objectives1. Describe the development of a communication

workshop for internal medicine trainees focusedon goals of care and code status discussions.

2. Recognize that brief educational interventionscan improve palliative care communication skills.

Original Research Background. Palliative carecommunication skills are applicable to all physiciansregardless of specialty. All internists need to under-stand and develop skills to facilitate conversationsthat ascertain patient preferences regarding life-sus-taining therapies and goals of care (GOC), particu-larly in the context of serious illness. Internalmedicine and palliative care educators both recognizethe need to further develop curriculum structures that

address these communication skills. This project willassess an initiative focused on enhancing these skills.Research Objectives. Assess the impact of acommunication workshop for internal medicinetrainees focused on goals of care and code statusdiscussions.Methods. The learners consisted of internal medi-cine residents (n¼67) at the University of Iowa duringthe 2017-2018 academic year. We developed and im-plemented a 100-minute interactive GOC communica-tion workshop to introduce learners to the knowledgeand skills of negotiating GOC and code status (CD).Multiple teaching strategies were used including pre-assigned reading material, video clips demonstratingeffective GOC discussion, didactics, and facilitatedrole-playing led by interdisciplinary faculty. The evalu-ation included: 1) a standard session evaluation surveyto examine learner reaction to the training; and 2) aretrospective pre/post self-assessment survey on confi-dence in discussing GOC.Results. 31 residents completed the survey. Analysisof paired evaluations showed significant improve-ments in residents’ confidence in conducting 1)GOC discussions (p¼0.0012); 2) CD discussion(p¼0.03) and 3) end-of-life discussions (p¼0.01).87% found the curriculum to be useful; 97% reportedthat palliative care training is essential.Conclusion. The GOC communication workshop wassuccessfully implemented and increased internal med-icine residents’ confidence in having difficult discus-sions. We need to further implement suchcurriculum in other residency-training programs.Implications for Research, Policy, orPractice. Further research is needed to understandthe amount of training required to help learners skill-fully incorporate GOC into their discussions withpatients.

Control Groups in RCTs ofPsychoeducational Palliative CareInterventions: A Systematic Review (S830)Carrie Henry, MSN CNM, Emory University, Atlanta,GA. Hyejin Kim, PhD, Emory University School ofNursing, Atlanta, GA. Eliza Park, MD, University ofNorth Carolina, Chapel Hill, NC. Sandra Ward, PhDRN, University of Wisconsin-Madison, Madison, WI.Mi-Kyung Song, PhD RN FAAN, Emory University, At-lanta, GA.

Objectives1. Explain the purpose of a control group that is

neither usual care nor no-treatment.2. Describe the current practice using control

groups that are neither usual care nor no-treat-ment in RCTs of psychoeducational palliativecare interventions.




3. Compare the explanations of the active interven-tion and the more-than-usual-care control in trialreports of RCTs of psychoeducational palliativecare interventions.

Background and Objective. In randomizedcontrolled trials (RCTs) an active intervention’s effi-cacy is determined relative to a control condition;thus the control condition’s design is as important asthe intervention’s. In RCTs of psychoeducational palli-ative care interventions, researchers often use a con-trol condition that is neither usual care nor no-treatment, but little is known about the extent ofresearch practices in designing such control condi-tions. Therefore, this study examined: 1) the currentresearch practice using control conditions that areneither usual care nor no-treatment controls in RCTsof psychoeducational palliative care interventions,and 2) the rationale and completeness of the descrip-tion of those control conditions in trail reports.Study Identification. PubMed, Embase, PsychINFO,and Web of Science were searched. The final sampleincluded nine trial reports.Data Extraction and Synthesis. For quality assess-ment, a modified Delphi list was used. The descrip-tions of both the active interventions and thecontrol conditions were assessed using a modifiedversion of Schulz et al’s Intervention Taxonomychecklist.Results. Four trials used attention controls designedto be equivalent to the structure of the active interven-tions. An additional four trials used a control condi-tion with some aspects of attention control, butneither the amount nor the intensity of attentionwere similar to the intervention. Only three trial re-ports stated the rationale for choice of control condi-tions. Most reports contained delivery mode,materials, duration, frequency, and sequence of boththe intervention and control conditions; but nonedescribed the qualifications or training required todeliver the control condition. Only one reportmentioned the fidelity monitoring method, and noreport included fidelity monitoring data.Conclusions and Implications for Practice, Policy,and Research. Our review of RCTs in psychoeduca-tional palliative care interventions calls for re-searchers’ attention to appropriate selection, design,conduct and report of control conditions.

Systematic Advance Care Planning andPotentially Avoidable Hospitalizations ofNursing Home Residents (S831)Susan Hickman, PhD, Indiana University School ofNursing, Indianapolis, IN. Kathleen Unroe, MD, Indi-ana University, Indianapolis, IN. Mary Ersek, RNFPCN, University of Pennsylvania, Philadelphia, PA.

Wanzhu Tu, PhD, Indiana University, Indianapolis,IN. Monica Ott, MD, Indiana University, Indianapolis,IN. Greg Sachs, MD FACP, Indiana University Schoolof Medicine, Indianapolis, IN.

Objectives1. Describe the systematic implementation of per-

son-centered advance care planning in nursinghomes using the Respecting Choices Last Stepsfacilitation model.

2. Discuss the associations between who is offeredACP, resident characteristics, and hospitalizationrates.

Original Research Background. Prior research sug-gests associations between nursing home (NH) resi-dents’ preferences for comfort-focused care andlower rates of hospitalization.Research Objectives. To compare hospitalizationrates among residents following intensive efforts to sys-tematically offer ACP.Methods. The sample consisted of 1,468 NH resi-dents enrolled in a multi-component demonstrationproject designed to reduce potentially avoidablehospitalizations between January 2015 e June2016. Embedded project nurses certified inRespecting Choices Last Steps facilitated ACP. Hos-pitalizations were tracked using Minimum Data Set3.0 data and judged as avoidable or unavoidable byproject nurses.Results. Comparisons were made between residentsbased on ACP status: 1) ACP indicting comfort care/DNH (n ¼ 497, 33%); 2) ACP with other preferences(e.g., code status only: n ¼ 771, 52%); and 3) no ACP(n¼ 218, 15%). Compared to the comfort care/DNHgroup, the overall hospitalization rate was 1.47 timeshigher for patients having other ACP preferences(p¼.005) and almost 2 times higher for those withno ACP (p¼.0003). Compared to the comfort care/DNH group, avoidable hospitalizations were 2.48times higher than for those with no ACP (p¼.0005).However, when adjusted for covariates includinggender, age, hospice, functional status, and cognition,there were no differences between the three groups.Conclusion. In this large, non-randomized study, theassociation between lower hospitalization rates andACP status were no longer significant once fullyadjusted for resident characteristics. Isolating the ef-fects of ACP may be challenging due to study design(multi-modal intervention, non-randomized) and thenature of the population. Although ACP was identi-fied as a key factor in reducing hospitalizations in aqualitative evaluation by stakeholders, it serve as areflection rather than a determinant of culturechange.Implications for Research, Policy, orPractice. Additional studies, including prospective



trials, are needed to determine the impact of ACP onnursing home resident outcomes.

Intensity in End-of-Life Care and HospiceUtilization for Patients Dying with HeartFailure (S832)Rebecca Hutchinson, MD MPH, Maine Medical Cen-ter, Portland, ME. Adam Black, BA, Maine MedicalCenter, Portland, ME. Kathleen Fairfield, MD, MaineMedical Center, Portland, ME.

Objectives1. Describe results from this analysis which found

that patients dying with advanced heart failurehad high rates of healthcare utilization.

2. Recognize patient characteristics that are associ-ated with receipt of more aggressive care for pa-tients dying with heart failure.

Original Research Background. Little is knownabout end-of-life (EOL) care received for patientsdying with congestive heart failure (CHF). One priorMedicare analysis found CHF patients received aggres-sive care at EOL and had low rates of hospiceutilization.Research Objectives. We used claims data to under-stand EOL healthcare utilization for patients dyingwith CHF.Methods. We analyzed a multi-payer database to iden-tify 11,685 patients 35 or older who died with CHF inMaine between 2013 and 2016. The primary outcomewas healthcare utilization during EOL, defined as 180days prior to death, including emergency department(ED) visits, hospitalizations, ICU admissions and hos-pice utilization. Patient characteristics analyzedincluded age, gender, medical comorbidities, areadeprivation index (ADI) and rurality.Results. Among 11,685 patients, 49% had $2 hospi-talizations, 73% had $2 ED visits, and 30% had anICU stay during EOL; 62% were admitted to hospice.Younger patients, men and patients with $2 other co-morbidities (compared with none) were more likely toreceive aggressive care on all measures and less likelyto utilize hospice. Rural patients were more likely tohave an ICU admission (34% in most rural vs 24%in urban, p<0.0001) but less likely to have$2 ED visitsor receive hospice. Patients residing in areas with high-er ADI were less likely to enroll in hospice (54% ofthose from the most deprived vs 69% in the least, p-value <0.0001) and more likely to have an ICU stay(33% in most deprived vs 27% in least deprived, p-value <0.0001).Conclusion. Healthcare utilization is high for pa-tients dying with advanced CHF, and we observed vari-ation across patient groups and according to ruralityand ADI.

Implications for Research, Policy, orPractice. Decreasing undue aggressiveness at EOLfor patients with advanced CHF may require betterrecognition of EOL, education of providers, patientsand families, and increased access to palliative care,especially in rural and socioeconomically deprivedareas.

Predictors of Depression and Anxiety inFamily Members Three Months After ChildAdmission to a Pediatric ICUs (S833)Masayuki Iwata, MSN RN PhN, Keio University, Chita-Gun, Aichi, Japan. Soojeong Han, RN AGNP-BC, Uni-versity of Washington, Seattle, WA. Helene Starks, PhDMPH, University of Washington, Seattle, WA. RossHays, MD, Children’s Hospital and Regional Med Cen-ter, Seattle, WA. Ardith Doorenbos, PhD RN FAAN,University of Washington, Seattle, WA.

Objectives1. Describe the difficulties family members’ experi-

ence when their child is in the ICU.2. Describe factors that are associated with family

members’ depression and anxiety.Original Research Background. Critically ill andinjured children and infants are admitted to chil-dren’s intensive care units (ICUs). Children’s familymembers are known to suffer from severe stress andto be a high-risk population for acute and chronic psy-chological problems. Depression, anxiety, and acuteand post-traumatic stress among family members dur-ing and after a child’s admission affect not only well-being but also social functioning and productivity.Research Objectives. We aimed to identify associa-tions between demographic and psychosocial variablesat early stages of a child’s ICU admission and depres-sion and anxiety in family members at approximately3 months after admission. We also explored predictivemodels for depression and anxiety at 3 months afterhospitalization.Methods. 380 family members of 220 children re-ported demographic and psychosocial status atapproximately 1 week after ICU admission (baseline),at discharge from the ICU, and at 3 months after thechild’s ICU admission. Clinical data were extractedfrom the children’s medical records. We used linearregression models and stepwise linear regression foranalyses.Results. We found gender (female) and child mortal-ity were strongly associated with family members’depression and anxiety at 3 months. Worse psycholog-ical health status at baseline, represented by reporteddepression, anxiety, and acute stress symptoms, wasassociated with more severe depression and anxietyat 3 months. Also, better social support at baseline




was associated with better psychological status (e.g.,lower depression and anxiety) at 3 months.Conclusion. We suggest a need to screen familymembers with validated scales and intervening withthose at high risk of depression and anxiety at 3months.Implications for Research, Policy, orPractice. Our findings suggest that a family’s ownfunctioning and ability to cope with stress may havea stronger impact on family members’ psychologicalhealth, we also suggest that health care providersempower family members to maintain and improvetheir own family’s well-being or resilience.

Cancer Patients’ and HealthcareProviders’ Perceptions About Supportiveand Integrative Oncology Services (S834)Cherry Jiang, BA, Case Western Reserve UniversitySchool of Medicine, Cleveland, OH. Olivia LarbiCase Western Reserve University, Cleveland, OH. De-nise Feyes, MS, Case Western Reserve University,Cleveland, OH. Hasina Momotaz, MSC, Case WesternReserve University, Cleveland, OH. Ming Li, PhD,Case Western Reserve University, Cleveland, OH. Ka-therine Daunov, APRN, University Hospitals ClevelandMedical Center, Cleveland, OH. Barbara Daly, RNFAAN, Case Western Reserve University, Cleveland,OH. Susan Mazanec, PhD RN, Case Western ReserveUniversity, Cleveland, OH. Samuel Rodgers-Melnick,MT-BC, University Hospitals, Cleveland, OH. SarahRolfe, RDN CSO LD, University Hospitals SeidmanCancer Center, Cleveland, OH. Nancy Tamburro,LISW-S OSW-C, University Hospitals Cleveland Medi-cal Center, Cleveland, OH. Richard Lee, MD, CaseWestern Reserve University, Cleveland, OH.

Objectives1. Describe cancer patients’ and healthcare pro-

viders’ perceptions of supportive and integrativeoncology services.

2. Compare cancer patients’ and healthcare pro-viders’ perceptions of supportive and integrativeoncology services.

Original Research Background. Supportive andintegrative oncology services improve quality oflife for cancer patients and are increasinglypopular.Research Objectives. To characterize and comparethe perceptions of supportive and integrativeoncology services among cancer patients and health-care providers.Methods. A cross-sectional survey was administeredat Seidman Cancer Center (SCC), an NCI desig-nated Comprehensive Cancer Center, to providersand patients in the spring of 2018. We inquiredabout familiarity, perceived importance, and

frequency of use, accessibility and barriers of 19 sup-portive and integrative oncology services. Data anal-ysis included the Chi-square test and Spearman’srank correlation (r).Results. A total of 585 surveys were obtained (421 pa-tients and 164 healthcare providers). Patients weregenerally over 60 (58.2%), female (57.4%), Caucasian(64.2%) with most at >1 year from starting treatment(59.9%). Healthcare providers were physicians(38.7%), RN partners (38.1%), and advanced practiceproviders (APPs) (23.2%). Most were female (74.3%),Caucasian (80%) and worked at SCC for >5 years(56.4%). Providers were more familiar with palliativecare (71.7%) and felt it was more important (92%)than patients did (25.2% and 43.6%, p<0.001). Pa-tients who were in treatment for a longer length oftime were more familiar with social work, palliativecare and psychiatry (r¼ 0.17, 0.14, 0.20; p<0.01).Most providers (>85%) of all types regarded palliativecare, social work and diet & nutrition services asimportant. The most common barrier for both pa-tients and providers was being unaware of the services(41.6% and 67.1%).Conclusion. Overall, healthcare providers were morefamiliar and considered most services to be moreimportant than patients with many supportive andintegrative oncology services. Being unaware of theservices was a common barrier.Implications for Research, Policy, orPractice. Interventions are needed to improve the pa-tients’ and providers’ awareness of supportive andintegrative oncology services and communication ofthe importance of these services.

Acute Care Utilization at End-of-Life inSickle Cell Disease: Highlighting the Needfor a Palliative Approach to Sickle CellDisease (S835)Emily Johnston, MD, University of Alabama at Bir-mingham, Birmingham, AL. Oyebimpe Adesina, MDMS, University of Washington, Seattle, WA. HeatherAmato, MPH, Public Health Institute, Richmond,CA. Susan Paulukonis, MA MPH, Tracking California,Richmond, CA. Smita Bhatia, MD MPH, University ofAlabama at Birmingham, Birmingham, AL.

Objectives1. Describe the acute nature and young age of

deaths of patients with Sickle Cell Disease.2. Consider what a palliative approach to care of pa-

tients with Sickle Cell Disease entails.Original Research Background. Despite recent ad-vances, people with sickle cell disease (SCD) continueto have a life expectancy <50y. Therefore, understand-ing end-of-life care in SCD is critically important, butremains understudied.




Research Objectives. To determine the location ofdeath and acute care utilization by people with SCDat end-of-life.Methods. The study utilized the California SickleCell Data Collection Program database, which com-bines data from administrative sources, vital records,and Medicaid claims. We examined people with SCDwho died between 2006 and 2015 (cases) and exam-ined their utilization of the hospital, emergencydepartment (ED), and intensive care unit in theirlast year of life compared to living controls withSCD matched 1:1 based on age, year, insurance, andincome.Results. The 485 cases with SCD died at a mean ageof 44y (SD: 16y). Most people with SCD died in thehospital (63%) after short admissions (mean 3.4d)and the ED (15%). In the last year of life, peoplewith SCD were admitted for an average of 42d (SD:49d) over 5 inpatient admissions. Utilization patternswere stable throughout the year and comparable forcases and controls until the month before deathwhen the cases had a sharp increase in utilizationwith the exception of 1) a slow increase in the lengthof hospital admissions for cases (2.6 days 12 monthsbefore death to 5.7 days the month before death)and 2) more ED visits for young adult (22-39y) casescompared with children and older adult SCD casesor young adult SCD controls.Conclusion. People with SCD are dying acutely andat a young age e with most dying in the hospital (aftershort visits) and in the ED.Implications for Research, Policy, or Practice. InSCD, a palliative approach to care should be extendedbeyond managing chronic pain and psychosocial chal-lenges to include advanced directives and living willsat a young age.

Periprocedural Code Status Discussions forInpatients Undergoing PercutaneousGastrostomy Tube Placement (S836)Rebecca Kalman, MD, Maine Medical Center, Port-land, ME. Rebecca Hutchinson, MD MPH, MaineMedical Center, Portland, ME.

Objectives1. Describe recommendations for ‘‘Do-Not-Resusci-

tate’’ orders in the periprocedural period.2. Discuss current low rate of documented peripro-

cedural code status conversations for patients un-dergoing inpatient G-tube placement.

Original Research Background. Gastrostomy tube(G-tube) placement is a common procedure per-formed for patients with life-limiting diseases. Patientsmay present for G-tube with a ‘‘Do-Not-Resuscitate’’(DNR) order. Despite multiple national societies rec-ommending periprocedural conversations for patients

with a DNR status, it is not clear how often these con-versations occur.Research Objectives. We sought to evaluate the fre-quency of documented code status conversations forinpatients who are DNR at the time of G-tube place-ment at an academic medical center. We also exploredfactors associated with the presence of a documentedconversation.Methods. We performed a retrospective chart reviewfor adult in patients undergoing G-tube placement be-tween May 2016 and May 2017. We abstracted demo-graphic information, type of G-tube inserted, codestatus, indication for G-tube and mortality data. Forpatients with a code status other than ‘‘Full’’ at timeof G-tube, notes five days pre- and post-procedurewere reviewed for documentation of a code statusdiscussion.Results. We identified 254 adult inpatients who un-derwent G-tube placement during the one-year studyperiod. 101/254 patients (44%) were 66 or older,62% were male and more than half had the highestseverity of illness. The most common indication forG-tube was dysphagia/aspiration (23% of patients).Thirty-three (13%) had code status other than ‘‘Full’’at the time of procedure. Of those, 9 (27%) had docu-mented code status discussion. Patients for whomanesthesia was involved were significantly more likelyto have a documented code status discussion (89%of patients with an anesthesia consult vs. 33% of pa-tients without; p¼.0057).Conclusion. The majority of patients with code statusother than ‘‘Full’’ at the time of procedure did nothave documented discussions in the chart despiteclear recommendations from major medical societies.Implications for Research, Policy, orPractice. Future work should include investigationinto interventions to improve the rate of code statusconversations as well as ensuring these are appropri-ately documented.

A Pilot Study of Hospice AdmissionPredictors of Hyperactive TerminalDelirium (S837)Jeannette Kates, PhD RN CRNP, Thomas JeffersonUniversity College of Nursing, Philadelphia, PA.

Objectives1. Identify types of terminal delirium and its rele-

vance to hospice patient care.2. Describe the relationships between hospice

admission data and hyperactive terminaldelirium.

3. Recognize opportunities for future researchabout terminal delirium.

Original Research Background. Terminal deliriumis a common occurrence at the end of life. It is a




poor prognostic indicator that is distressing to pa-tients, caregivers, and providers. The pathogenesis ofterminal delirium is not well understood.Research Objectives. The purpose of this study wasto explore the relationships between hospice admissiondata and the incidence of hyperactive terminaldelirium. Specifically, the study explored whether phys-ical, psychosocial, and/or spiritual data collected at hos-pice admission was associated with terminal delirium.Methods. This was a retrospective cohort studydesign utilizing chart review. Admission data werecollected from the Hospice Item Set (shortness ofbreath, pain, scheduled and PRN opioids, bowelregimen); psychosocial assessment (psychiatric diag-nosis, significant fears); and spiritual assessment(active spirituality). Chi-squared test of independence,Fisher’s exact test, and ANOVA tested for relationshipsbetween independent variables and the dependentvariable of hyperactive terminal delirium.Results. A sample size of 148 deceased hospice pa-tients was included in this study. The independent var-iables of psychiatric diagnosis, significant fears, activespirituality, shortness of breath, pain, and scheduledopioid were not found to have significant relation-ships with hyperactive terminal delirium (p > .05).The relationship between PRN opioid and terminaldelirium was found to be significant [c2 (1, N ¼148) ¼ 4.587, p < .05]. Risk ratios indicated that therisk of terminal delirium, within the category of PRNopioid prescription, increased by 58.5% relative tothe group of non-PRN opioid prescription. Logisticregression analysis indicated that while PRN opioidprescription (B ¼ .806, p ¼ .034) is a significant pre-dictor variable, although when combined with theequation constant, it did not provide a worthwhile pre-dictive model (Nagelkerke R-Square ¼ .041).Conclusion. There were no clear relationships be-tween physical, psychosocial, and spiritual admissiondata and hyperactive terminal delirium.Implications for research. This study provides pre-liminary data needed to inform future research aboutterminal delirium predictors.

Impact of Race and Ethnicity onEnd-of-Life Experiences for Childrenwith Cancer (S838)Erica Kaye, MD, St. Jude Children’s Research Hospital,Memphis, TN. Courtney Gushue, DO, NationwideChildren’s Hospital, Columbus, OH. Samantha De-marsh, BS, Ohio University Heritage College of Oste-opathic Medicine, Athens, OH. Jennifer Snaman,MD, Dana-Farber Cancer Institute, Boston, MA. Lind-say Blazin, MD, St. Jude Children’s Research Hospital,Memphis, TN. Liza Johnson, MD MBE MPH, St. JudeChildren’s Research Hospital, Memphis, TN. Deena

Levine, MD, St Jude Children’s Research Hospital,Memphis, TN. Justin Baker, MD FAAP FAAHPM, St.Jude Children’s Hospital, Memphis, TN.

Objectives1. Recognize data regarding racial and ethnic dis-

parities in the provision of end-of-life (EOL)care within the adult oncology literature, andhow these data contrast with the limited litera-ture on EOL disparities in the context of pediat-ric oncology.

2. Examine the impact of race and ethnicity onEOL variables for pediatric palliative oncologypatients in this study.

3. Discuss potential hypotheses to explain why theimpact of race/ethnicity on EOL variables maydiffer between the medical and pediatriconcology literature.

Original Research Background. Racial and ethnicdisparities in the provision of end-of-life (EOL) careare well described in the medical oncology literature.However, the impact of racial and ethnic disparities atEOL in the context of pediatric oncology remainspoorly understood.Research Objectives. To investigate associations be-tween race/ethnicity and EOL experiences for chil-dren with cancer.Methods. A retrospective cohort study was conductedon 321 children with cancer enrolled on a palliativecare service who died between 2011 and 2015 usinga comprehensive standardized data extraction toolcomprising a broad spectrum of EOL metrics.Results. Black patients were more likely to receivecardiopulmonary resuscitation as compared to Whitepatients (correlation coefficient¼1.413, confidenceinterval¼0.359e2.467, p¼0.009); however, theremainder of variables related to treatment and EOLcare did not correlate significantly with race. Hispanicpatients were less likely to receive cancer-directed ther-apy within 28 days prior to death (correlation coeffi-cient ¼-0.708, confidence interval ¼ -1.398e0.018,p¼0.044) as compared to White patients, yet theywere more likely to report a goal of cure over comfortas compared to Non-Hispanic patients (correlation co-efficient ¼1.129, confidence interval¼0.042e2.216,p¼0.042). The remainder of EOL variables were notfound to be significantly correlated with ethnicity.Conclusion. In contrast with data from the medicaloncology literature, neither race nor ethnicity corre-lated with most EOL variables for pediatric palliativeoncology patients treated at a large urban pediatriccancer center. Multicenter investigation is needed toascertain the impact of racial/ethnic disparities onEOL experiences of children with cancer.Implications for Research, Policy, orPractice. These data suggest that race and ethnicity



have minimal impact on EOL experiences for chil-dren with cancer treated at this study site. These find-ings differ significantly from the medical oncologydata, highlighting the critical need for further investi-gation of associations between race/ethnicity andEOL care for children with cancer across diverse treat-ment centers.

Barriers to Provision of Palliative andHospice Care to Children and Families inthe Community: A Population-Level Surveyof Hospice Nurses (S839)Erica Kaye, MD, St. Jude Children’s Research Hospital,Memphis, TN. Ashley Kiefer, MD, The University ofTennessee Health Science Center Hospice and Pallia-tive Medicine Fellowship, Memphis, TN. Jason Rey-nolds, MD, Cook Children’s Hospital, Fort Worth,TX. Kristina Zalud, BS, Creighton University Schoolof Medicine, Omaha, NE. Justin Baker, MD FAAPFAAHPM, St. Jude Children’s Hospital, Memphis, TN.

Objectives1. Discuss the challenges related to access of quality

palliative and hospice care to children and fam-ilies in the community.

2. Describe the self-reported training and experi-ence levels of community-based hospice nurseswith provision of care to children and familiesacross the domains of symptom management,end-of-life care, goals of care, family-centeredcare, and bereavement.

3. Describe the self-reported comfort of commu-nity-based hospice nurses with provision of careto children and families across the domains ofsymptom management, end-of-life care, goals ofcare, family-centered care, and bereavement.

Original Research Background. Approximately500,000 children in the United States suffer fromlife-limiting illnesses annually, many of whom are hos-pice eligible. Unfortunately, most children enrolled inhospice agencies receive services in the absence ofspecialized pediatric programs.Research Objectives. To determine the levels ofexpertise and comfort of hospice nurses who providecare to children and families in the community.Methods. A cross-sectional survey was developed, pi-lot-tested, and widely distributed to hospice nursesacross a tristate region. Survey items assessed nurseexperience and comfort across the domains of symp-tom management, end-of-life care, goals of care, fam-ily-centered care, and bereavement.Results. A total of 71 hospices that provide services tochildren participated, from which 551 respondentscompleted surveys. The majority of nurses reportedno training in pediatric palliative or hospice care(89.8%), with approximately half reporting < 5 years

of hospice experience (53.7%) and no experienceproviding care to pediatric patients (49.4%). Thosewith pediatric hospice experience reported limited op-portunities to maintain or build their skills, with themajority providing care to children several times ayear or less (85.7%). Nurses reported feeling some-what or very uncomfortable providing services to chil-dren during the illness trajectory and at the end of lifeacross all domains.Conclusion. Children with serious illness who receivecare from local hospices often interface with nurseswho lack training, experience, and comfort in the pro-vision of palliative and hospice care to pediatricpatients.Implications for Research, Policy, orPractice. These findings should inform future devel-opment and investigation of educational resources,training programs, and child- and family-centered pol-icies to improve the delivery of palliative and hospicecare to children in the community.

Evolution of an InterprofessionalPalliative Care Fellowship and Integrationof a PGY-2 Pain Management andPalliative Care Pharmacy Resident (S840)Benjamin Kematick, PharmD, Dana-Farber CancerInstitute, Boston, MA. Bridget Scullion, PharmDBCOP, Dana-Farber Cancer Institute, Arlington, MA.Jane deLima Thomas, MD FAAHPM, Dana-FarberCancer Institute, Boston, MA. Shih-Ning Liaw, MDFAAP, Dana-Farber Cancer Institute, Boston, MA.Juliet Jacobsen, MD DPH, Massachusetts General Hos-pital, Boston, MA.

Objectives1. Describe the results of a survey of interprofes-

sional fellows in regard to their educational expe-rience after completion of a fellowship.

2. Discuss comments from fellows and faculty in re-gard to their experience participating in an inter-professional education.

Original Research Background. Clinical practice inpalliative care lends itself to a deeply integrated teamdynamic. Early exposure of multiple disciplines inpalliative care training may enhance the function ofthe team in delivering patient care. The Harvard Inter-professional Palliative Care Fellowship (HIPCF) is theumbrella program for several specialty-specific pro-grams, including physician, nursing and social work.In 2016, a Palliative Care Pharmacy Resident wasadded to the interprofessional fellowship, which wasunique nationally.Research Objectives. Describe the interprofessionaleducational experience of one class of fellows (with apharmacy fellow) before and after the completion ofthe fellowship.




Describe the interprofessional educational experienceof fellows in an interprofessional fellowship prior tothe inclusion of a pharmacy resident.Describe fellowship faculty experience with teachingan interprofessional fellowship.Methods. Surveys were administer to 3 cohorts offellowship participants, fellows from the 2017 AY (in-clusive of a pharmacist), fellows from 2012-2016 AY,and fellowship faculty. The 2017 fellows were adminis-tered a survey assessing interprofessional educationcompetencies in a post-then-pre fashion. Past fellowswere administered a survey in a post fellowshipfashion. Faculty were asked about their subjectiveexperience in administering the fellowship. The on-line survey was completed over 12 weeks fromDecember 2017-March 2018.Results. 2017 fellows assessed a significant improve-ment in 3 out of 4 interprofessional education domains;Values and Ethics, Roles and Responsibilities and,Communication (+0.592 p¼0.006, +0.935 p¼0.011,+0.932 p¼0.039, respectively). All four domains trendedtoward improvement in self assessed competencies. Pastfellows similarly showed high levels of self-assessed com-petency after completion of their fellowships.Conclusion. The Harvard Interprofessional PalliativeCare Fellowship is a unique post-graduate educationopportunity nationwide. The four professions whohave the opportunity to participate receive a valuableinterprofessional educational experience in additionto a robust clinical curriculum.Implications for Research, Policy, or Practice. Asmore fellowships move to an interprofessional educa-tion model, data regarding interprofessional educa-tion is warranted.

Improving Conversations About ICDManagement: A Training Session forAdvanced Heart Failure Clinicians (S842)Ian Kwok, MD, Icahn School of Medicine at Mount Si-nai, New York, NY. Harriet Mather, MD MSc, IcahnSchool of Medicine at Mount Sinai, New York, NY.Karen McKendrick, MPH, Icahn School of Medicineat Mount Sinai, New York, NY. Laura Gelfman, MDMPH, Icahn School of Medicine at Mount Sinai,New York, NY. Nathan Goldstein, MD FAAHPM,Mount Sinai Hospital, New York, NY.

Objectives1. Describe how patients and clinicians conceptu-

alize the role of ICD deactivation in the settingof advance heart failure.

2. Practice and demonstrate specific examples ofeffective communication techniques related toICD management and goals of care in the settingof advanced heart failure.

Original Research Background. Implantable cardi-overter-defibrillator (ICD) devices reduce the inci-dence of sudden cardiac death for high-risk patients,but can also cause pain and anxiety at the end oflife. However, conversations about ICD deactivationare difficult and occur infrequently.Research Objectives. The study is a sub analysis ofdata from educational sessions for clinicians whocare for patients with advanced heart failure aspart of a 6-center, randomized-controlled trial. Pri-mary outcomes compared pre- and post-workshopscores assessing confidence and skill in communica-tion about advance care planning, ICD deactivationdiscussions, and use of empathy in conversationswith patients.Methods. Clinicians at intervention sites partici-pated in a 90-minute communication training work-shop focusing on goals of care communicationincluding ICD deactivation, while clinicians at con-trol sites received a lecture introducing the study.We first compared pre-workshop scores betweencontrol and intervention groups, then comparedpre- and post-workshop scores for the interventiongroup, based on self-ratings on a 5-point Likertscale.Results. 82 heart failure clinicians were enrolled.Pre-workshop scores showed high levels of confi-dence (4.22, SD¼0.76) and skill (3.96, SD 0.76)in advance care planning in the setting of ICDmanagement in both control and interventiongroups, while comparisons of pre- and post-work-shop scores in the intervention group showed de-creases in confidence (-1.10) and skill (-0.12).None of these results reached statisticalsignificance.Conclusion. Heart failure clinicians had high baselineself-perceptions of confidence and skill in advance careplanning in the setting of ICD management, with nodifferences between intervention and control sites.There were no significant changes seen immediately af-ter the communication training intervention.Implications for Research, Policy, orPractice. The results suggest that heart failure clini-cians may overestimate their ability to engage inadvance care planning discussions involving ICD deac-tivation, which may act as a barrier to the propagationof primary palliative care in this setting. Furtherresearch is needed to improve conversations aboutICD management.

ICU Bereaved Family Members’Adjustment During the First Year (S843)In Seo La, MSN RN, University of Maryland School ofNursing, Baltimore, MD. Debra Wiegand, PhD RN




CHPN FAAN FPCN, University of Maryland, Balti-more, MD.

Objectives1. Identify how bereavement outcomes of family

members change during the first year after a pa-tient dies in the ICU setting.

2. Identify factors influencing prolonged griefsymptoms over time.

Research Background. Nearly one in five deaths oc-curs in the ICU. Family members participate in theend-of-life decision making process for their criticallyill family member and are considered to be at riskfor psychological problems. However, little is knownabout how they adjust over time.Research Objectives. Based on the Double ABCXmodel of family adaptation, the purpose of this studywas to examine the prevalence and changes in symp-toms of prolonged grief, post-traumatic stress, depres-sion, stress, and anxiety.Methods. A longitudinal design was used to conductthe investigation using three time points (1 to 3, 6,and 12 months after death). Family members were sur-veyed using validated instruments. Data were analyzedusing descriptive statistics and linear mixed modeling.Results. Participants (n ¼ 30) were 60% female, 60%Caucasian, and 57% spouses. At 1-3 months after loss,60% had symptoms of post-traumatic stress disorderand then 30.4% had symptoms 6 months after loss. At1-3 months after loss, symptoms of depression, anxiety,and stress were found in 40%, 30%, and 26.7% of familymembers, respectively. At 6 months, symptoms ofdepression (17.4%), anxiety (13.3%), and stress (13%)were observed. In linear mixed models, symptoms ofprolonged grief and post-traumatic stress, depression,and anxiety improved over time. However, they didnot report a significant reduction of stress over time.There was also a negative association between social sup-port and prolonged grief symptoms over time, indi-cating that the association decreased as time increased.Conclusion. Bereavement outcomes reduced overthe first year after loss, but family members tend tobe at risk for psychological problems in the earlymonths after bereavement. They may be best helpedif early support can be provided by relatives, friends,and health care providers.Implications. Further larger longitudinal studies andearly supportive intervention studies need to be con-ducted to prevent the negative impacts of bereave-ment on family members’ psychological outcomes.

Agreement Between Two Brief Tools toAssess Pain in the Palliative CareSetting (S844)Maya Ladenheim, BA, University of California, SanFrancisco School of Medicine, San Francisco, CA.

David O’Riordan, PhD MPH, University of California,San Francisco, San Francisco, CA. Giovanni Elia, MDFAAHPM, University of California, San Francisco,San Francisco, CA. Steven Pantilat, MD FACPFAAHPM, University of California, San Francisco,San Francisco, CA.

Objectives1. Describe the most frequently used single-item

tools to assess pain severity, their strengths andlimitations.

2. Examine the agreement between two single-itempain assessment tools commonly used during apalliative care consultation.

3. Demonstrate the agreement between patientreport of pain improvement.

Original Research Background. The numeric rat-ing scale (NRS) and the verbal rating scale (VRS)are commonly used, unidimensional tools for assess-ing pain among patients referred for palliative care(PC).Research Objectives. Examine the agreement be-tween the NRS and VRS for the assessment of painamong seriously ill patients, and assess change inpain scores from initial assessment to follow-up 24-hours later.Methods. Patients receiving inpatient PC who re-ported pain ‘right now’ were assessed using the NRS(0¼‘no pain’ to 10¼‘worst’) and the VRS (none,mild, moderate, severe), and had a follow-up assess-ment 24-hours later. Patients were also asked if theirpain had improved over the past 24-hours. Improve-ment was defined as change of 2+ points for theNRS and 1+ category for the VRS.Results. Thirty-four patients provided baseline andfollow-up pain assessments. At baseline, the meanNRS pain level was 4.3/10. Using the VRS, 40.6% re-ported mild pain, 34.4% moderate, and 25.0% se-vere. There was excellent agreement(r¼0.8,p<0.0001) between these measures. Atfollow-up, the mean NRS score was 4.4/10. Usingthe VRS, 6.2% reported no pain, 39.3% mild,33.3% moderate, and 21.2% severe. There was alsoexcellent agreement between the two measures atfollow up (r¼0.7,p<0.0001).Patient report of pain improvement at follow-up, iden-tified that 42.8% improved. Using the NRS, 18.2%improved, and 34.4% improved with the VRS. Theagreement between patient-reported improvementand the NRS change score was slight(k¼0.09,p¼0.5), and the VRS change score was fair(k¼0.3,p¼0.03).Conclusion. There was excellent agreement betweenthe NRS and VRS. However, perceptions of improve-ment did not align with those identified using theNRS or VRS.



Implications for Research, Policy, orPractice. Pain is a subjective and complex symptom.Assessing change in pain using unidimensional toolsmay not fully capture the patient experience andmore detailed measures may be needed.

Health Care Utilization and Intensity atEnd of Life is High Amongst Adults WhoRelapse Following AllogeneicHematopoietic Cell Transplantation (S845)Jessica Langston, MD MPH, Stanford University, PaloAlto, CA. Vandana Sundaram, MPH, Stanford Univer-sity, Palo Alto, CA. Vyjeyanthi Periyakoil, MD, StanfordUniversity School of Medicine, Stanford, CA. LoriMuffly, MD, Stanford University, Stanford, CA.

Objectives1. Recognize specific challenges for palliative and

end-of-life care in the acute leukemia relapsepopulation.

2. Devise novel interventions to improve end-of-lifecare in this patient population.

Original Research Background. Relapse is the lead-ing cause of death for patients with acute leukemia(AL) and myelodysplastic syndrome (MDS) who un-dergo allogeneic hematopoietic cell transplantation(HCT).Research Objectives. We describe survival, intensityof healthcare utilization, and characteristics associatedwith high resource utilization at EOL.Methods. Adult patients with AL/MDS who under-went HCT at a large regional referral center with sub-sequent relapse between 2005 and 2015 were includedin this retrospective study. We created a compositescore for EOL healthcare utilization intensity sum-ming the presence of any of the following criteria:death in hospital, use of chemotherapy, emergencydepartment (ED), hospitalization, intensive care unit(ICU), intubation, cardiopulmonary resuscitation, orhemodialysis in the last month of life. Higher scoresindicate more intense healthcare use at EOL. Multi-variable linear regression analysis was used to deter-mine variables associated with EOL healthcareutilization intensity.Results. 154 patients were included. 140 (91%)died within two years of relapse with median(IQR) survival after relapse for those who died of5 months (1-9). Overall inpatient healthcare utiliza-tion in this cohort was high with 44% visiting theED at least once, 92% hospitalized (16% >¼ 5times), and 38% using ICU. Utilization was higheven among those receiving no additional disease-directed therapy. For those patients who died, themedian (range) intensity score for EOL healthcareutilization was 2 (0-8). Most (70%) had a marker

of high-intensity healthcare use at EOL or died inhospital. In multivariable analysis, post-relapsechemotherapy plus cell therapy (donor lymphocyteinfusion and/or repeat HCT) (estimate (95% CI):1.41 (0.45-2.37)) compared to no treatment wasassociated with more intense EOL healthcare use;no other variables met significance.Conclusion. Inpatient healthcare utilizationfollowing post-HCT relapse is high despite knownpoor prognosis, including at EOL.Implications for Research, Policy, orPractice. Interventions are needed to minimizenon-beneficial treatments and promote goal-concor-dant EOL care in this seriously ill patientpopulation.

Development of New UndergraduatePalliative Care Knowledge Measure (S846)Megan Lippe, PhD RN, University of Alabama, Tusca-loosa, AL. Andra Davis, PhD MN BSN, WashingtonState University, Vancouver, WA.

Objectives1. Describe the process for developing a new pallia-

tive care knowledge measure.2. Explore possible ways to utilize the new knowl-

edge measure to evaluate current students’ palli-ative care knowledge.

Original Research Background. Palliative carenursing education has long been guided by the Endof Life Nursing Education Consortium (ELNEC) cur-riculum, originally released in 2001. However, no mea-sure to evaluate student’s knowledge exists thatappropriately reflects current palliative care bestpractice.Research Objectives. The purpose of the presenta-tion is to describe the development and psychometricevaluation of a new knowledge measure to be used forthe evaluation of undergraduate nursing student’spalliative care knowledge.Methods. I) Creation of the new knowledge mea-sure was guided by relevant research literature in in-strument and scale development. The knowledgemeasure, titled the Undergraduate Nursing PalliativeCare Knowledge Survey (UNPCKS), was developedin four systematic steps: 1) item generation froma team of seven palliative care and nursingeducation experts; 2) pilot test of UNPCKS; 3) in-strument revision with experts; and 4) psychometrictesting.Results. The final version of the UNPCKS is a 27-item, multiple-choice instrument that evaluates un-dergraduate nursing students’ palliative care knowl-edge. Students at three universities (n¼262)completed the UNPCKS for psychometric testing.




Exploratory factor analysis revealed two primary fac-tors: palliative care principles and nursing-specific re-sponsibilities. Content validity was established by theexpert panel.Conclusion. The UNPCKS is a psychometricallystrong contemporary measure that can be utilized toevaluate students’ palliative care knowledge.Implications for Research, Policy, orPractice. Future testing of the efficacy of the measureto evaluate changes in knowledge corresponding withpalliative care or ELNEC-Undergraduate educationare in progress. However, the instrument can be inte-grated into existing education programs to evaluatestudents’ palliative care knowledge.

Enhancing Healthcare Students’ PerceivedCompetence to Care for Dying Patients: AnInterprofessional Simulation (S847)Megan Lippe, PhD RN, University of Alabama, Tusca-loosa, AL. Anne Halli-Tierney, MD, The University ofAlabama, Tuscaloosa, AL. Robert McKinney, PhDMSW, The University of Alabama, Tuscaloosa, AL.

Objectives1. Describe the three steps of the interprofessional

withdrawal of life-sustaining measures simulation.2. Identify at least one course at the attendee’s

home institution within which the simulationcan be integrated.

Original Research Background. Within the contextof an aging society in which individuals are increas-ingly diagnosed with multiple chronic conditions,there is a critical need for effective interprofessionaleducational interventions to prepare healthcare pro-vider students to care for dying patients.Research Objectives. The purpose of this study wasto evaluate the impact of an interprofessional with-drawal-of-life-sustaining measures simulation on medi-cal residents’, nursing students’, and social workstudents’ perceived competence to care for dyingpatients.Methods. Medical residents (n¼8), social work stu-dents (n¼8), and pre-licensure nursing students(n¼57) were divided into small interprofessionalteams to engage in a three-stage simulation. In eachstage, teams engaged in pre-briefing, performance,and debriefing. First, team members communicatedwith family members regarding the need for a trache-ostomy insertion. Second, nursing students assessedthe client following the development of septic shockand communicated findings to the physician. Third,the team members communicated with the familyand removed the patient from life-sustaining interven-tions. Perceived competence to care for dying patientswas evaluated prior to and immediately following thesimulation.

Results. The majority of participants had no priorhospice/palliative care patient experience and hadnot recently experienced the loss of a loved one orcared for a loved one who died. Most students hadreceived hospice/palliative care education. Internalconsistency reliability of the new measure was high(Cronbach’s a¼ .957). Mixed ANOVA results demon-strated significant improvements in perceived compe-tence overall (p<.001) without a noted interactioneffect or difference based upon healthcare profession.Conclusion. The interprofessional withdrawal-of-life-sustaining measures simulation significantly enhancedperceived competence for healthcare students. Rec-ommendations and implications will be discussed.Implications for Research, Policy, orPractice. The simulation was an effective, dynamicmechanism to educate students from medicine,nursing, and social work regarding communicationand end-of-life care.

Patient-Nurse Discordance in Goals ofCare at End-of-Life (S848)Amy Lipson, PhD MPH MSW, Case Western ReserveUniversity, Cleveland, OH. Sara Douglas, PhD, CaseWestern Reserve University, Cleveland, OH.

Objectives1. Describe the demographics of patients with

advanced cancer and oncology nurses in theoutpatient setting.

2. Identify patterns in discord in goals of care be-tween the patient and nurse.

Original Research Background. Providing high-quality care to patients with advanced cancer relieson patients receiving treatment that aligns with theirown goals of care (GoC). Goal concordance is depen-dent on communication between patients and theirhealth care providers. While some research has beenconducted on patient-oncologist discordance inGoC, little is known about the role of nurses in thiscontext.Objective. The purpose of this study was to deter-mine concordance in GoC at end-of-life between pa-tients with advanced cancer and nurses.Methods. Using a sub-sample from a longitudinal,descriptive design study, data were collected on sub-jects with a diagnosis of stage 4 cancer and theirnurses. Subjects were asked, ‘‘Regarding your care/the care of this patient, what is most important toyou right now?’’ Anchors on the instrument wereQOL (0) and survival (100)dwith a value of 50 indi-cating equal weight on both domains. Discordancewas defined as a > 40 point difference on the VAS.Results. Results from the 167 PTs diagnosed withadvanced cancers who died during the study periodare presented. Mean age for PTs was 64.0 (SD¼10.3,




range¼36-88) with 51.5% being female. The majorityof the RNs were female (84.6%) with a mean age of48.9 (SD¼7.0, range¼28-61) and 13.3 years of experi-ence in oncology nursing. In 37% of the PT-RN dyads,there was discord in GoC at the last interview prior todeath. There was a statistically significant difference(p¼.000) as the PTs reported goals more survival-focused (M¼47.4, SD¼31.3) than the RNs (M¼29.9,SD¼21.4).Conclusion and Implications for Practice, Policy,and Research. These data show some discord be-tween PT and RN at end-of-life and highlight theneed for eliciting patient GoC throughout the illnesstrajectory. With the continuity of the PT-RN dyad inthe outpatient setting, RNs are in the unique positionto discuss GoC to ensure that PTs receive carecongruent with their own preferences.

#Palliativecare and #Hospice: AComprehensive Instagram Analysis (S849)LaKeesha Lockett, DO, The University of TexasHealth Science Center, San Antonio, TX. ConnieCheng, MS, The University of Texas Health ScienceCenter, San Antonio, TX. Laura Reilly-Sanchez HealthCareers High School, San Antonio, TX. Sandra San-chez-Reilly, MD FAAHPM, UT Health Science Centerand South Texas Veterans Health care System, San An-tonio, TX. Jeanette Ross, MD AGSF FAAHPM, TheUniversity of Texas Health Science Center, San Anto-nio, TX.

Objectives1. Assess Instagram posts related to #hospice and

#palliative care.2. Identify trends in #Hospice and #Palliativecare

Instagram postings.3. Compare trends between #hospice and #pallia-

tive care.Background. Image-sharing social media platformssuch as Instagram have grown in popularity and arerapidly becoming a media for personal, business andhealth-related information sharing. Instagram hasover 1 billion active monthly users, with 64% of usersranging from 18 to 34 years old. The content of Insta-gram images of palliative care (PC) and hospice havenot been investigated.Objective. To assess Instagram posts related to #hos-pice and #palliative care, illustrating general popula-tion knowledge and opinions of hospice andpalliative care and comparing trends.Methods. Instagram-public data was retrieved for 30days (images, video, metadata) for the two hashtagscontaining PC and hospice. Images were screened; du-plicates removed and excluded non-English content.A customized coding scheme to determine major con-tent themes and images was developed. Themes were

characterized by user (hospice/PC/personal); posttype (picture, video); description (selfie, infographic),content (advertisement, fundraiser, education, ani-mals); purpose (personal, work, business, motivationaland healthcare-related). Content was evaluated toexplore trends comparing hospice vs. PC.Results. #hospice (n¼2916) and #palliativecare(n¼1075), 389 posts were included after screening;292 #Hospice and 97 #palliativecare images selectedfor analysis. Most images were pictures of others andmost images were posted by individual users. Commoncontent themes included fundraising (26.99%; 12.5%PC vs. 31.85% hospice; p¼0.002), advertisement24.68%, p¼NS), family/friends (14.65%; 28.13% PCvs. 10.27% hospice; p<0.001), education (11.05%;p¼NS), and animals (8.23%; 2.08% PC vs. 10.27% hos-pice; p¼0.038). If the user had a hospice/PC back-ground, they will be more likely to post educationalmaterials (p¼0.0388).Conclusions. Instagram #hospice and #palliativecareare available and widely used. #hospice is used moreoften in general and when identifying common cate-gories such as fundraising, posting about the story ofloved one or pets. Hospice and PC organizationshave started to use posts as a means of education.Future studies should analyzed how Instagram canbe utilized as a means for educating the population,worldwide promotion and empowering patients andcaregivers.

Management of Fatigue at the End of Life:A Systematic Review (S850)Susan Lowey, PhD RN CHPN FPCN, State Universityof New York College at Brockport, Brockport, NY.

Objectives1. Identify the most common and effective treat-

ment modalities for fatigue management at theend of life.

2. Describe implications from this review of currentevidence on directions for future researchfocused on symptom management of fatigue.

Background and Objectives. Fatigue is a commonand distressing symptom associated with patientswith progressive terminal illnesses who are nearingthe end of life. The cause of fatigue is multifacetedand its presentation is exclusively subjective, whichcan both contribute to difficulty with treatment ap-proaches. The purpose of this systematic review wasto examine and describe the state of evidence sur-rounding the most current and effective treatmentsfor fatigue in patients nearing the end of life.Study Identification. Medical databases weresearched using keywords including: fatigue, fatiguemanagement, terminal illness, end-of-life care, and palliativecare. Articles published between 2000 and 2018




focused on the management or treatment for fatiguein end-of-life, hospice or palliative care were included.Review articles or those that did not meet minimumquality score were excluded. Twenty-seven articleswere included in the analysis.Data Extraction and Synthesis. Study purpose, sam-ple, design and results were extracted from eacharticle. Using the matrix method, commonalitieswere further analyzed both within and across thesample.Results. 85% of the articles were quantitative andfocused on describing presentation of the symptomand/or treatment. Ten of the papers focused on symp-tom clusters with the diagnosis of cancer to be themost common illness antecedent associated with fa-tigue. Both pharmacological and non-pharmacolog-ical approaches were identified but few randomizedcontrol trials focusing exclusively on fatigue manage-ment were found.Conclusion and Implications for Practice, Policy,and Research. Research aimed at addressing bestpractices for fatigue management at the end of lifeare lacking, particularly those involving randomizedtrials. Few papers focusing exclusively on fatiguewere found, and this symptom is most widely includedin papers that examine symptom clusters, most oftenin cancer patients. Studies focused on examiningand comparing fatigue treatment in cancer and non-cancer illnesses should be considered.

Pediatric Cardiothoracic Surgeons andPalliative Care: A National Survey Study(S851)Rachna May, MD, University of Oklahoma Health Sci-ences Center, Oklahoma City, OK. Emily Balkin, MD,University of California, San Francisco, San Francisco,CA. Jess Thompson, MD, University of Oklahoma,Oklahoma City, OK. Elizabeth Blume, MD, BostonChildren’s Hospital, Boston, MA. Satish Rajagopal,MD, University of California, San Francisco, San Fran-cisco, CA.

Objectives1. Outline pediatric cardiothoracic surgeons’ un-

derstanding and use of pediatric palliative care.2. Describe the barriers perceived by pediatric

cardiothoracic surgeons to palliative careconsultation.

Original Research Background. Among childrenwith heart disease who die in the hospital, the majorityoccur within the first year of life in an intensive caresetting. Discrepancies in understanding of prognosisand perceptions of suffering between parents and phy-sicians caring for children with heart disease havebeen identified. While pediatric cardiology and

palliative care (PC) provider attitudes towards PChave been described, no studies have assessed the per-spectives of pediatric cardiothoracic surgeons (CTS).Research Objectives.

� Evaluate pediatric CTS understanding and use ofPC.

� Describe barriers perceived by pediatric CTS toPC consultation.

Methods. A survey was distributed electronically topediatric CTS in the United States. Subjects wereexcluded if they did not have a pediatric componentto their practice. The survey consisted of 22 primarilyclosed-ended questions with multiple choice answers.Results. Of the 220 CTS who were mailed the survey,36 opened the survey and 5 did not meet inclusioncriteria (n¼31). Median years of practice was 23.5and 87.1% were male. Almost all (90%) reportedthat they had experience consulting PC. While 68%felt PC consultation was initiated at the appropriatetime, 29% felt it occurred too late. When asked theappropriate timing for PC consultation in patientswith hypoplastic left heart syndrome, 45% selected‘‘at time of prenatal diagnosis’’ and 30% selected‘‘when surgical and transcatheter options have beenexhausted.’’ Common barriers to PC involvementincluded the perception of ‘‘giving up’’ (40%) andconcern for undermining parental hope (36%).Conclusion. While a majority of pediatric CTS arefamiliar with PC, there is variation in perception ofappropriate timing of consultation and significant bar-riers to consultation still exist.Implications for Research, Policy, orPractice. This is a preliminary study to better under-stand the attitudes of pediatric CTS regarding PC. Re-sults will help inform future research and educationalefforts for physicians caring for children withadvanced heart disease.

Patterns of Anti-Cancer Therapy Use inthe Last 14 Days of Life in a CommunityCancer Institute (S852)Caitlyn McNaughton, MA MHS PA, Lancaster GeneralHealth/Penn Medicine, Lancaster, PA. Selah Lynch,MS, University of Pennsylvania, Philadelphia, PA. Kris-tina Newport, MD FAAHPM, Penn State Health andCollege of Medicine, Hershey, PA. Sarah Svetec, BA,Haverford College, Haverford, PA. Riley Wong, BAS,University of Pennsylvania, Philadelphia, NY. RachelHolliday, BS, Lancaster General Health/Penn Medi-cine, Lancaster, PA. Shanthi Sivendran, MD, PennMedicine at Lancaster General Health, Lancaster, PA.

Objectives1. Define the costs of providing aggressive end of

life care.




2. Describe population most likely to receive anti-cancer therapy at end of life in a community can-cer center.

3. Discuss implications and potential application ofthese findings.

Original Research Background. Studies haveshown that aggressive cancer care at end of life is asso-ciated with decreased quality of life, decreased mediansurvival, and increased cost of care.Research Objectives. This study describes the pa-tients most likely to receive aggressive anti-cancer ther-apy at the end of life in a community cancer institute.Methods. We performed a retrospective review of213 patients who received anti-cancer therapy inour institution and died between July 2016 andApril 2017. Data collected included primary malig-nancy, death date, date of last anti-cancer treatment,hospice enrollment, healthcare utilization,Oncology Care Model (OCM) enrollment, and clin-ical assessments at last office visit prior to a treat-ment decision before death. Data were analyzedusing univariate logistic regression to determinefeature importance.Results. Of the 201 patients who died of cancer,36(17%) received anti-cancer therapy within the last14 days of life. Several factors were significantly posi-tively correlated with receiving anti-cancer therapy atend of life, including enrollment in OCM (p <0.001), frequency of hospital utilization (p < 0.001),death in hospital (p<0.001), referral to hospice(p<0.001), and hematologic malignancy (p ¼ 0.014).Conclusion. In our community cancer institute,enrollment in OCM, frequency of hospitalizations,death in a hospital, referral to hospice, and hemato-logic malignancy diagnosis were predictive ofreceiving aggressive anticancer therapy at the end oflife, suggesting that these factors should have greaterimportance in our clinic.Implications for Research, Policy, orPractice. Taken as a whole, these data will helpinform clinicians and patients in choices regardingcare near the end of life.

Palliative Care Education in U.S. AdultNeuro-Oncology Fellowship Programs(S854)Ambereen Mehta, MD MPH, University of Virginia,Charlottesville, VA. Sarah Verga, DO, University of Vir-ginia, Charlottesville, VA. Camilo Fadul, MD, Univer-sity of Virginia, Charlottesville, VA.

Objectives1. Describe the current need for palliative educa-

tion in neuro-oncology fellowship programs asidentified by program directors.

2. List the preferred education methods and toolsof neuro-oncology fellowship program directorsfor teaching palliative care to fellows.

3. Know which barriers program directors facewhen providing palliative care education for neu-ro-oncology fellows.

Original Research Background. Palliative care (PC)for patients with neuro-oncological diseases positivelyimpacts morbidity and mortality. No studies have eval-uated whether neuro-oncology fellows receive formalPC education during fellowship.Research Objectives. The purpose of this study wasto describe the PC education and identify educationneeds of US neuro-oncology fellowship programs.Methods. Program directors (PDs) of US neuro-oncology fellowships were surveyed. The electronicsurvey included qualitative and quantitative questions.Results. Of 26 programs fellows, 17 completed sur-veys (65% response rate) of which 3 (18%) offeredno formal PC education. The methods most utilizedwere formal didactics (seminars/conferences) andself-directed reading materials. One-third of programshave developed their own teaching materials. Commu-nication was the domain identified as most important,the domain fellows were most well-trained in, and thedomain PDs felt most comfortable providing for theirown patients. Addressing spiritual distress and initi-ating life-prolonging therapies were the domains PDsidentified as being least important, fellows were leastwell trained in, and PDs were least comfortableproviding for their own patients. Most programs(83%) were satisfied with the PC education availableat their program. Time for teaching and faculty avail-ability were the most common barriers.Conclusion. Neuro-oncology PDs recognize the needfor PC education, which is currently offered in someform by most programs, but the content and methodsof delivery are heterogeneous. Interdisciplinary educa-tional teams and nationally-available PC educationalmaterial may improve implementation of PC educa-tion in neuro-oncology.Implications for Research, Policy, orPractice. This study implies the need for a dedicatedneuro-oncology palliative education curriculum.

Patient and Family Caregiver Perspectiveson Palliative Care Needs in End-StageLiver Disease: A Qualitative Study (S855)Sandhya Mudumbi, MD, University of Alabama at Bir-mingham, Birmingham, AL. Macy Stockdill, RN, Uni-versity of Alabama at Birmingham, Birmingham, AL.Nicholas Hoppmann, MD, University of Alabama atBirmingham, Birmingham, AL. James Dionne-Odom,PhD RN ACHPN, University of Alabama at Birming-ham, Birmingham, AL. Brendan McGuire, MD,




University of Alabama at Birmingham, Birmingham,AL. Cynthia Brown, MD MSPH, University of Alabamaat Birmingham, Birmingham, AL. Marie Bakitas,DNSc NP-C FAAN, University of Alabama at Birming-ham School of Nursing, Birmingham, AL.

Objectives1. Identify the specific impact ESLD has on patients

and their family caregivers from theirperspectives.

2. Identify the stated needs of patients and familycaregivers in ESLD care from their perspectives.

3. Identify the barriers to integrating PC into careof patients with ESLD from patient and caregiverperspectives.

Original Research Background. Palliative Care(PC) is underutilized in persons with end-stage liverdisease (ESLD) and little data exists on patient andfamily caregivers’ perspectives on PC needs and howPC can be integrated into ESLD care.Research Objectives. Identify ESLD patient andfamily caregiver perspectives on challenges of livingwith ESLD, potential PC needs, and barriers to inte-grating PC.Methods. Semi-structured one-on-one interviewswere conducted with purposively-sampled patientswith ESLD and their caregivers at a tertiary care aca-demic medical center. Patients and caregivers wereasked about: 1) challenges of living with ESLD,2) their unmet needs, and 3) their understandingand perceptions of PC and hospice, including access-ing these services. Interviews were digitally recordedand transcribed. Transcripts were entered into NVivosoftware and analyzed using thematic analysis.Results. Patients (n¼7) had a mean age of 67 andwere mostly female and white (70%) with ESLD dueto alcohol (43%), hepatitis C (57%), non-alcoholicsteatohepatitis (29%), and with concurrent hepatocel-lular carcinoma (43%). Most caregivers were femaleand white (83%), and were the patient’s spouse/part-ner (83%). Patients and caregivers perceived thatESLD challenges occurred in all four quality of life(QOL) domains (physical, emotional, social, and spir-itual). Participants’ needs included better communi-cation with providers, emotional support, caregiversupport, and practical needs. A majority of patientsand caregivers had a lack of understanding of PC.Conclusion. Thematic analysis identified a variety ofunmet patient and family caregiver needs in ESLDthat could be addressed by PC services. However, a ma-jor barrier is a lack of understanding of PC services.Implications for Research, Policy, orPractice. These results provide a first step in interven-tion development for a PC intervention to addressidentified patient/caregiver needs in ESLD with afocus on enhancing PC literacy.

Perceptions of Inappropriate Critical CareAre Decreasing (S856)Thanh Neville, MD, University of California, Los An-geles, Los Angeles, CA. Joshua Wiley, PhD, MonashUniversity, Clayton Campus Victoria, Australia. Mira-mar Kardouh, BSc, University of California, Los An-geles, Los Angeles, CA. Myrtle Yamamoto, RN,University of California, Los Angeles, Los Angeles,CA. J. Randall Curtis, MD MPH, University of Wash-ington, Seattle, WA. Neil Wenger, MD, University ofCalifornia, Los Angeles, Los Angeles, CA.

Objectives1. Explain what "inappropriate critical care’’ mean.2. Recognize that there has been a decrease in inap-

propriate critical care at one institution.Original Research Background. When a patient’schance of survival is very low or the quality of life ismarkedly diminished, intensive care interventionsthat prolong life without achieving the goals of medi-cine are often considered ‘‘inappropriate’’ by healthcare providers. In 2012, we showed that the prevalenceof inappropriate treatment was 11% at one academichealth system.Research Objectives. To assess whether the propor-tion of patients receiving inappropriate critical carehas changed from 2012 to 2017, we repeated the eval-uation in the same health system.Methods. On a daily basis from August 28 throughDecember 28, 2017, we surveyed critical careattending physicians in five intensive care units(ICUs) in one health system to ask whether each pa-tient was receiving inappropriate critical care and ifso, why the care was inappropriate. In-hospital and6-month mortality was collected. Receipt of inappro-priate critical care, patient characteristics andoutcomes were compared between 2017 and 2012.Results. Over 4 months, 55 physicians made 10,105assessments on 1424 critically ill patients. Of these,94 (6.6%) patients received perceived inappropriatecritical care, which was less than 11% (p<0.01) in2012. Comparing 2017 and 2012, patient age(mean 61.8 vs 60.6), MS-DRG (4.5 vs 4.5), lengthof stay (15 vs 14.9 days), and overall mortality(18% vs 20%) were not significantly different(p>0.05). The most common reason why treatmentwas inappropriate in 2012 was burdens grosslyoutweigh benefits (67% of patients) whereas in2017 it was that treatment cannot achieve patient’sgoals (70%). In 2017, inpatient mortality was 9%,44%, and 73% for patients receiving critical carethat was perceived to be appropriate, probably inap-propriate and inappropriate.Conclusion. Over five years at one health system theproportion of patients receiving perceived inappro-priate critical care dropped by 40%.



Implications for Research, Policy, orPractice. Understanding the reasons for this decreasein inappropriate critical care might elucidate how tofoster further improvement.

Lung Transplant Pulmonologists’ Views ofSpecialty Palliative Care for LungTransplant Recipients (S857)Eric Nolley, MD, University of Pittsburgh, Pittsburgh,PA. Jessica Fleck, MA, University of Pittsburgh, Pitts-burgh, PA. Dio Kavalieratos, PhD, University of Pitts-burgh, Pittsburgh, PA. Mary Amanda Dew, PhD,University of Pittsburgh, Pittsburgh, PA. Matthew Mor-rell, MD, University of Pittsburgh Medical Center,Pittsburgh, PA. Yael Schenker, MD, University of Pitts-burgh, Pittsburgh, PA.

Objectives1. Differentiate how lung transplant pulmonolo-

gists’ views of lung transplantation affect theiruse of specialty of palliative care (SPC).

2. Contrast lung transplant pulmonologists defini-tions of specialty palliative care with their pat-terns of specialty palliative care utilization forlung transplant recipients.

Original Research Background. Lung transplantrecipients face foreshortened life expectancies andfrequently experience significant symptoms. Theymay benefit from but rarely receive SPC services.Transplant pulmonologists’ views of SPC may be keyto understanding SPC utilization for this populationbut these have not been well characterized.Research Objectives. (1) Examine how lung trans-plant pulmonologists view SPC and make decisionsto refer transplant recipients to SPC and (2) identifyany unique aspects of lung transplantation affectingtransplant pulmonologists use of SPC.Methods. We conducted semi-structured interviewswith attending transplant pulmonologists at ninegeographically diverse high-volume transplant centerswith SPC services in the U.S. and Canada. All inter-views were audio-recorded and transcribed verbatim.The multidisciplinary team developed a qualitativecodebook using the constant comparative method.Two investigators coded all transcripts, with disagree-ments discussed and resolved by consensus.Results. We interviewed 37 transplant pulmonolo-gists. Only 2 participants had never referred a lungtransplant recipient to SPC. While most participantscorrectly defined SPC and differentiated SPC fromhospice, approximately half used SPC only when dis-ease-directed therapies failed. This approach was asso-ciated with a perception that transplant and SPC are‘‘not convergent paths’’ because transplant focuseson ‘‘survival and aggressive treatment,’’ particularlyin the first post-transplant year or when re-

transplantation is possible. Participants who reportedusing SPC alongside disease-directed therapies weremore likely to view transplant as a ‘‘palliative treat-ment’’ or a ‘‘terminal illness’’ with an uncertain ‘‘roll-ercoaster’’ course especially after the onset ofchronic rejection.Conclusion. Despite viewing SPC as more than solelyend-of-life care, many transplant pulmonologists viewSPC as incompatible with traditional post-transplantdisease-directed therapy.Implications for Research, Policy, orPractice. Efforts to integrate SPC into lung transplan-tation will require solutions that address transplantpulmonologists perception that transplant and SPCare divergent treatment paths.

The EFFECT (End-oF-liFE-CommunicaTion) Study: Acceptability,Feasibility, and Potential Impact of UsingMortality Prediction Scores for InitiatingEnd-of-Life Goals of Care Communication inthe Adult Intensive Care Unit (S859)Shelly Orr, PhD RN CNE, VCU School of Nursing,Richmond, VA.

Objectives1. Describe the role of the Sequential Organ Failure

Assessment (SOFA) in calculating mortality riskprediction scores.

2. Articulate the role of mortality risk predictionscores in promoting EOL goals-of-carecommunication.

3. Describe a patient example in which use of mor-tality prediction scores promoted earlier EOLgoals-of-care communication.

Original Research Background. Uncertainties inprognosis remain a barrier to end-of-life (EOL)communication in the intensive care unit (ICU).Mechanisms for increasing the accuracy and timeli-ness of EOL goals-of-care communication are needed.Research Objectives. This study evaluated: 1) theacceptability and feasibility of providers’ use of patientmortality prediction scores as part of routine practice,and 2) providers’ intentions to change practice,related to goals-of-care communication, as a result ofawareness of the scores.Methods. An explanatory mixed-methods approachwas used. Using Sequential Organ Failure Assess-ment (SOFA), patient mortality prediction scoreswere provided to ICU providers (12) at a large ur-ban medical university who then completed anacceptability and feasibility questionnaire. Follow-upinterviews were conducted to further understandand gain insight into providers’ perceptionsregarding EOL practice changes as a result of hav-ing the scores.




Results. Overall, use of mortality risk predictionscores was acceptable and feasible. There was somedisagreement related to the use of SOFA scores asan effective way for determining patient mortalityrisk. Providers with limited ICU experience were eagerand accepting of the scores while those with vast expe-rience found the scores to be an adjunct to their ownintuition. All providers acknowledged the benefit oflooking at daily scores or ‘trends’ and the most sub-stantial theme was the need to consider SOFA scoresin relation to patient context.Conclusion. Use of SOFA scores for potentiallyincreasing EOL goals-of-care conversations appearsto be most beneficial for providers with limited ICUexperience. A case example will be provided forattendees.Implications for Research, Policy, orPractice. Deficiencies in EOL care communicationcan compromise quality of EOL care and increaseresource utilization. Although large-scale studies areneeded to determine the effect on patient EOL out-comes, routine consideration of mortality predictionscores may provide an avenue for more accurate andtimelier EOL goals-of-care communication.

For Change You Need a Roadmap: AnImplementation Model to Improve SeriousIllness Communication Across HealthSystems (S860)Joanna Paladino, MD, Ariadne Labs, Boston, MA.Francine Maloney, MPH, Ariadne Labs, Boston, MA.Jane Kavanagh, BA, Ariadne Labs/Brigham andWomen’s Hospital, Boston, MA. Erik Fromme, MDMSCR FAAHPM, Ariadne Labs, Boston, MA.

Objectives1. Describe the three phases of a novel implementa-

tion model designed to achieve system-levelimprovement in serious illness communication.

2. Apply incremental milestones derived from im-plementation science and organizational changeto primary palliative care programs.

Original Research Background. Programs that aimto improve clinician-patient communication aboutvalues and goals (‘serious illness communication’)often focus on skills-training while neglecting the sys-tem in which communication takes place. Drivingmeasurable improvements in communication requiresorganization-level change.Research Objectives. Develop a novel implementa-tion model for improving serious illness communica-tion across a health system with practical steps forpalliative care leaders.Methods. Researchers and implementers with theSerious Illness Care Program (SICP) at Ariadne Labs

(AL) conducted a four-stage process: 1) Assembledan expert panel (n¼10) to draft a theory of change.2) Reviewed the implementation science and organi-zational change literature. 3) Synthesized learningsinto key implementation features and engaged adesigner to create a ‘‘roadmap.’’ 4) Refined the road-map with feedback from its application in threesystems.Results. The ‘Implementation Roadmap’ has threephases: Prepare; Train/Coach; Sustain. Phase onecreates a supportive environment for implementa-tion. Actions include engaging leaders and col-leagues to gain buy-in, assembling and training ateam of champions (trainers + implementationteam) with dedicated resources, selecting levers tosupport practice change, choosing pilot sites, andcustomizing the program (clinician training/coach-ing; workflow; EHR template; metrics for moni-toring/evaluation). Phase two launches theprogram in pilot sites with rapid-cycle-improvement.The team trains early-adopter clinicians in seriousillness communication who initiate the workflow, re-sulting in documented conversations with patients.The team uses metrics to track conversations, getfeedback, and provide support to frontline clini-cians. Phase three expands the program to new sitesand plans for evaluation and sustainability, e.g. dash-boards, automated triggers.Conclusion. An Implementation Roadmap providesincremental milestones and practical steps to sup-port palliative care leaders interested in organiza-tion-level, measurable improvements in seriousillness communication across populations andsettings.Implications for Research, Policy, orPractice. Improving communication for a populationrequires education plus systems-change; following aroadmap may increase the likelihood of reaching thedestination: every patient, every time.

Measuring the Quality of Palliative Carefor Patients with End Stage Liver Disease(S861)Arpan Patel, MD, Mount Sinai, New York, NY. StevenAsch, MD MPH, Palo Alto Health Care System, PaloAlto, CA. Anna Liza Antonio, DRPH, University of Cal-ifornia, Los Angeles Health, Los Angeles, CA. FasihaKanwal, MD MSHS, Baylor College of Medicine, Hous-ton, TX. Karl Lorenz, MD MS MSHS, Stanford - VAPalo Alto Health Care System, Palo Alto, CA. DeborahRiopelle, MSPH, VA Health Services Research andDevelopment Service/CSHIIP, Los Angeles, CA. Mar-tin Lee, PhD, University of California, Los Angeles,Fielding School of Public Health, Los Angeles, CA.




Anne Walling, MD, University of California, Los An-geles, Los Angeles, CA.

Objectives1. Discuss sampling strategies for measuring pallia-

tive care quality among Veterans with end stageliver disease (ESLD).

2. Describe approaches for measuring the quality ofcare for Veterans with (ESLD).

Original Research Background. Nineteen palliativecare quality indicators (QIs) for patients with EndStage Liver Disease (PC-ESLD) were recently devel-oped using the RAND/UCLA appropriatenessmethod (RAM).Research Objectives. These indicators were devel-oped to measure the quality of care for patients withdecompensated cirrhosis and a MELD > 19 or aCTP Score of 12 or greater.Methods. We identified patients newly diagnosed withdecompensated cirrhosis in 2012 who lived at least 30days after diagnosis at a single Veteran health center us-ing any of three ICD-9 based sampling strategies and fol-lowed them for one year or until death. A chartabstraction tool established a gold standard for clinicaldocumentation of ESLD to compare the sensitivity andspecificity of sampling strategies.Wealsopilot tested thePC-ESLD QIs in all patients confirmed to have ESLD.Results. Out of 167 patients identified using at leastone sampling strategy, 62 were confirmed to meet eligi-bility criteria after chart abstraction. The best perform-ing sampling strategy had a sensitivity of 62% andspecificity of 60%. The addition of laboratory valuescan improve specificity while limiting sensitivity. 98%of Veterans in the cohort were male, mean age at diag-nosis was 61, 74.2% were white, 14.5% African Amer-ican, 1.6% Asian, 6.5% other, and 3.2% did not havedataon race. 38%weremarried or livingwith significantother, 36% homeless, and 66% had a history of drug oralcohol use. The overall QI pass rate for our cohort was65% (59% for information care planning QIs and 76%for supportive care Qis). Patients receiving palliativecare consultation were more likely to meet criteria forinformation care planning QIs (68% vs. 36%, p¼0.02).Conclusion. PC-ESLD Quality Indicators are validand feasible.Implications for Research, Policy, orPractice. Measuring the quality of palliative care for pa-tients withESLD is feasible and can identify opportunitiesfor quality improvement in a vulnerable population.

Palliative Care Needs of Individuals withCystic Fibrosis: A National Survey ofPatients, Caregivers, and Cystic FibrosisCare Team Members (S862)Elisabeth Dellon, MD MPH, University of North Car-olina School of Medicine, Chapel Hill, NC. Mara

Hobler, PhD, University of Washington Medical Cen-ter, Seattle, WA. Anna Georgiopoulos, MD, Massachu-setts General Hospital, Boston, MA. Jessica Goggin,MAS RN, University of California, San Diego, San Die-go, CA. Elaine Chen, MD, Rush University MedicalCenter, Chicago, IL. Sarah Hempstead, MS, CysticFibrosis Foundation, Bethesda, MD. Albert Faro,MD, Cystic Fibrosis Foundation, Bethesda, MD. DioKavalieratos, PhD, University of Pittsburgh, Pitts-burgh, PA.

Objectives1. Describe patient, caregiver, and providers percep-

tions of how palliative care in cystic fibrosis dif-fers from palliative care in other conditions.

2. Identify the top palliative care needs of individ-uals with cystic fibrosis.

Original Research Background. Although palliativecare (PC) has been demonstrated to alleviate sufferingfor patients with serious illness, no evidence exists forits role in cystic fibrosis (CF).Research Objectives. To ensure the relevance offuture CF-specific PC education and interventions,we 1) identified top PC needs of individuals with CF;and 2) examined perceptions of if/how PC mayneed to be different in CF.Methods. Adults with CF (‘‘patients’’), family care-givers (‘‘caregivers’’), and CF care providers (‘‘pro-viders’’) recruited from online listservs maintainedby the Cystic Fibrosis Foundation completed an onlinesurvey. We used descriptive statistics to analyzeresponses.Results. All participant groups, including 85% of 70patients, 87% of 100 caregivers, and 84% of 350 pro-viders felt PC in CF differs from PC in other condi-tions. All groups ranked the unpredictable diseasecourse as the top difference (44% of patients, 42%of caregivers, 39% of providers). Other key differencesincluded shortened life expectancy and CF being alifelong disease. Emotional support was identified asthe top PC need of individuals with CF (ranked firstby 38% of patients, 45% of caregivers, 49% of pro-viders), with emotional symptom management also be-ing a top need per patients and caregivers. Patientsand providers felt advance care planning is a top PCneed. Adult providers ranked physical symptom man-agement as a top need more often than pediatric pro-viders (47% vs 29%, P¼0.001). Pediatric providersidentified caregiver support as a top need more oftenthan adult providers (28% vs 16%, P¼0.011).Conclusion. Most patients, caregivers, and providersfeel that PC is different in CF than in other serious ill-nesses and identify a variety of PC needs.Implications for Research, Policy, or Practice. PCeducation and interventions targeting key differencesand needs could enhance the impact of PC in CF.



Barriers to Palliative Care in CysticFibrosis (S863)Elisabeth Dellon, MD MPH, University of North Car-olina School of Medicine, Chapel Hill, NC. Mara Ho-bler, PhD, University of Washington Medical Center,Seattle, WA. Anna Georgiopoulos, MD, MassachusettsGeneral Hospital, Boston, MA. Elaine Chen, MD,Rush University Medical Center, Chicago, IL. JessicaGoggin, MAS RN, University of California, San Diego,San Diego, CA. Sarah Hempstead, MS, Cystic FibrosisFoundation, Bethesda, MD. Albert Faro, MD, CysticFibrosis Foundation, Bethesda, MD. Dio Kavalieratos,PhD, University of Pittsburgh, Pittsburgh, PA.

Objectives1. Identify barriers to palliative care for individuals

with cystic fibrosis from the perspectives of pa-tients, caregivers, and CF providers.

2. Describe ways to overcome barriers to palliativecare in cystic fibrosis.

Original Research Background. In the absence ofclear guidelines for palliative care (PC) in cysticfibrosis (CF), little is known about PC provision to in-dividuals with CF and barriers to PC.Research Objectives. To identify barriers to PC inCF.Methods. CF care team members (‘‘providers’’),adults with CF (‘‘patients’’), and family caregivers(‘‘caregivers’’) recruited from listservs maintained bythe Cystic Fibrosis Foundation completed an onlinesurvey about barriers to PC in CF. Responses wereanalyzed using descriptive statistics.Results. Surveys were completed by 70 patients, 100caregivers, and 350 providers. All providers and 96%of patients and caregivers expressed beliefs that PCis valuable for individuals with CF. Nearly half of pa-tients and caregivers recalled knowing someone whoreceived PC, but rarely had personal experience.While 73% of providers reported introducing PC topatients, only 26% of patients and 12% of caregiversrecalled learning about PC from providers. From alist of common barriers to PC, all groups identifiedthe same top three: perception that palliative care isonly for dying people, lack of access to outpatientPC, and lack of PC training for CF providers. Nearly1/3 of providers felt patient reluctance to engage inPC conversations is an important barrier, versus 5%of patients and caregivers. Many patients (64%) andcaregivers (74%) were not aware of availability of PCspecialists, and 19% of providers did not know if PCspecialists were available to their patients.Conclusion. Patients with CF and caregivers reportlittle experience with PC despite a majority of pro-viders reporting they introduce PC to their patients.The most common barriers to PC reflect gaps in

knowledge and provider concerns about patient will-ingness to engage in PC conversations.Implications for Research, Policy, or Practice. PCeducation and tools to promote patient-providercommunication may help alleviate barriers to PC inCF.

Non-Surgical Patients with AdvancedGynecologic Cancer Discharged to SubacuteRehabilitation Centers Have Low Rates ofSubsequent Chemotherapy (S864)Nathan Riley, MD, UC San Diego/Scripps Health, SanDiego, CA. Camille Clefton, MD, Kaiser Permanente,Los Angeles, CA. Carolyn Lefkowits, MD, Universityof Colorado, Denver, CO.

Objectives1. Interpret survival data presented as median days

after discharge.2. Interpret significant of results from a univariate

analysis.Original Research Background. Women withadvanced gynecologic cancer are often enticed bythe promise of additional chemotherapy even whenthe potential for prolonged life or a cure is slim.Furthermore, many of these patients don’t qualifyfor additional chemotherapy due to preclusive weak-ness or medical comorbidities. Despite these observa-tions, they are often discharged from the hospital toskilled rehabilitation centers (SRC) to recuperate thestrength to qualify for additional chemotherapy.Research Objectives. We sought to characterizeprognosis and rates of subsequent chemotherapyamong non-surgical patients with advanced gyneco-logic cancer who were discharged to SRC.Methods. Patients with stage III/IV or recurrent gy-necologic malignancy who were admitted to the KaiserPermanente Southern California health care systemand discharged to SRC over a ten-year period wereincluded in this retrospective cohort. Patients whohad surgery during their admission or who wereenrolled in hospice prior to discharge were excluded.A univariate analysis was performed to identify patientcharacteristics that might be predictive of receivingfuture chemotherapy. Statistical significance was ascer-tained in all analyses through a Fisher’s exact test.Results. 35 patients met inclusion criteria. The major-ity were Caucasian (75%), and the median age was 70.Survival after discharge ranged from 5 to 1463 days,with a median survival of 57.5 days. Seven patients(20%) received future chemotherapy after discharge.None of the examined patient or admission factors(age, primary site of malignancy, upfront vs recurrentdisease, indication for admission) was predictive offuture chemotherapy. Trends toward longer survival




were seen in patients with age #75 (p¼0.27), cervicalcancer (p¼0.11), non-recurrent disease (p¼0.69) anda non-infectious admission indication (p¼0.49).Conclusion. In this cohort of non-surgical patientswith advanced gynecologic malignancy discharged toSRC, only 20% received additional chemotherapy,and the median survival was 58 days.Implications for Research, Policy, orPractice. This may be an appropriate population totarget for advanced care planning prior to discharge.

Why Do Some Patients Regret TheirDecision to Initiate Dialysis? (S865)Fahad Saeed, MD, University of Rochester MedicalCenter, Rochester, NY. Susan Ladwig, MPH, Universityof Rochester, Rochester, NY. Muhammad Sardar, MD,Monmouth Medical Center, Long Branch, NJ. PaulDuberstein, PhD, University of Rochester, Rochester,NY.

Objectives1. Discuss the importance of patient autonomy in

dialysis decision-making.2. Discuss the impact of a physician’s paternalistic

style of decision making on patients’ inner peacewith the decision.

Original Research Background. Nearly 25% of pa-tients receiving maintenance dialysis withdraw fromdialysis each year. Many patients regret their decisionto start dialysis.Research Objectives. To identify factors associatedwith dialysis regret.Methods. A forty-one item questionnaire was admin-istered to adult patients receiving maintenance dialysisin seven dialysis units located in Cleveland, Ohio andits suburbs. Of the 450 patients who were asked toparticipate in the study, 423 agreed. The question-naire items assessed patients’ knowledge of their kid-ney disease, attitudes toward chronic kidney disease(CKD) treatment, and preference for end-of-life(EoL) care. A single question was used to assess dial-ysis regret, ‘‘Do you regret your decision to start dial-ysis?’’ We used logistic regression to identifypredictors of decisional regret. Candidate predictorswere patient demographics, attitudes toward CKDtreatment, beliefs about the dialysis decision-makingprocess, and EoL care preferences.Results. Eighty of 395 respondents (20.2%) reporteddialysis regret. Three variables were associated withdialysis regret: (1) patients chose dialysis over conser-vative management to please doctors or family mem-bers (adjusted odds ratio (AOR) 3.33, confidenceinterval 1.73, 6.37), p¼<0.0001; (2) patients thoughtit was important for their families to be activelyinvolved in dialysis decision-making (AOR 1.97, CI1.73, 6.37), p¼0.0001; (3) patients reported not

having prognostic discussions with the kidney doctors(AOR 2.60, CI 5.85,1.15, CI), p¼0.0414.Conclusion. Dialysis regret was not uncommon inthis sample. Regret is associated with beliefs aboutthe dialysis decision-making process. There was no ev-idence of demographic (age, gender, race, income)differences in regret.Implications for research. Future researchinvolving multiple stakeholders (e.g., patients, care-givers, physicians) is warranted to identify modifiablerisk factors for dialysis regret and to improve dialysisdecision-making.

Trends in Hospital-Based SpecialtyPalliative Care: Insights from a NationalPalliative Care Quality ImprovementCollaborative (S866)Laura Schoenherr, MD, University of California, SanFrancisco, San Francisco, CA. Kara Bischoff, MD, Uni-versity of California, San Francisco, San Francisco, CA.Angela Marks, MSED, University of California, SanFrancisco, San Francisco, CA. David O’Riordan, PhDMPH, University of California, San Francisco, SanFrancisco, CA. Steven Pantilat, MD FACP FAAHPM,University of California, San Francisco, San Francisco,CA.

Objectives1. Describe at least two trends in processes of care

provided by inpatient palliative care (PC) teamsover time.

2. Describe one major change in a clinical outcomeachieved by inpatient PC teams over time.

3. Discuss how these trends relate to evolving expec-tations of and norms within the field.

Background. The field of palliative care (PC) isgrowing and evolving rapidly in response to increaseddemand and recognition of its benefits.Objectives. Describe how processes of care and out-comes achieved by inpatient PC teams have changedover time.Methods. Data for this study were extracted from thePalliative Care Quality Network database on 03/06/2018 and pertain to 135,197 patients referred to 88inpatient PC consult teams between 01/01/2013 and12/31/2017.Results. The most common diagnoses leading toinpatient PC consult were cancer (32.0%, range be-tween teams: 11.3%e93.9%), cardiovascular disease(13.2%, 0%e29.0%), and pulmonary disease (11.3%,0%e26.0%). The percentage of referred patientswith cancer decreased between 2013 and 2017(39.0% to 30.0%, p<0.0001), while there was an in-crease in the percentage of patients with cardiovascu-lar disease (12.0% to 14.0%, p<0.0001) andpulmonary disease (10.0% to 12.0%, p<0.0001).




Most patients were discharged from the hospital alive(78.7%, range between teams: 44.7%e99.4%), andthe percentage of patients discharged alive increasedover time (75.0% to 80.0%, p<0.0001). Between2013 and 2017, there was a substantial decrease in hos-pice referrals (46.0% to 31.0%, p<0.0001) and an in-crease in referrals to clinic-based (2.0% to 4.0%,p<0.0001) and home-based PC services (2.0% to4.0%, p<0.0001).Conclusions. There is wide variation in practiceacross inpatient PC teams. Overall, teams are seeingmore patients with diagnoses other than cancer andare doing so earlier in the course of illness, whichmay account for lower rates of hospice referral.Teams are connecting slightly more patients withoutpatient PC services at the time of hospitaldischarge.Implications. Changes in the practice of PC overtime are consistent with guidelines calling for PC forall patients with serious illness earlier in the courseof illness. Variations in practice between teams estab-lish benchmarks and reveal opportunities for improve-ment.

Finding the Path: Incorporating PatientPreferences into an Interactive ClinicalPathway Platform (S867)Dena Schulman-Green, PhD, Yale University School ofNursing, West Haven, CT. Emily Cherlin, PhD MSW,Yale School of Public Health, New Haven, CT. ReneeCapasso, BS, Yale University, New Haven, CT. SarahMougalian, MD, Yale Cancer Center, New Haven,CT. Shi-Yi Wang, MD, PhD, Yale School of PublicHealth, New Haven, CT. Cary Gross, MD, Yale Schoolof Medicine, New Haven, CT. Preeti Bajaj, PhD, Gen-entech, Inc., South San Francisco, CA. Katherine Ea-kle, PharmD, University of California, San Francisco,San Francisco, CA. Kerin Adelson, MD, Yale CancerCenter, New Haven, CT. Sharmi Patel, PharmD, Gen-entech, San Francisco, CA.

Objectives1. Identify preferences that affect treatment deci-

sion-making among women with Stage I-III breastcancer.

2. Describe a new clinical pathway model that willincorporate women’s preferences for individual-ized assessments of prognosis and treatment ben-efits and challenges, and that will facilitateshared-decision making and use of palliativecare services.

Original Research Background. Clinical pathwaytools offer physicians a selection of cost-effective, evi-dence-based treatment options to discuss with pa-tients. Although shared decision-making is essentialto patient-centered care, clinical pathway tools have

not integrated patient preferences around treatmentburden and outcomes.Research Objectives. We sought to identify patientpreferences for incorporation into MyPATHway, a pa-tient-centered, interactive clinical pathway platformfor patients with breast cancer.Methods. Using interpretive description as anapproach, we conducted individual qualitative inter-views with women aged 18+ treated for Stage I-IIIbreast cancer at Smilow Cancer Hospital in New Ha-ven, CT. We asked participants if there was additionalinformation they would have liked prior to treatmentdecision-making, their preferences for and factorsaffecting decision-making, and their attitude towardsan electronic platform, including preferences forlearning about and weighing treatment burdens andoutcomes. We coded transcribed interviews andanalyzed them for themes.Results. The sample’s (n¼21) mean age was 56.4(range 29-74). Breakdown of cancer stage was I(33.3%), II (42.9%), and III (23.8%). Participants re-ported receiving adequate information prior to treat-ment decision-making, although some wanted moreregarding what to expect during and after treatment.Several factors affected treatment decision-making,including physical (e.g., symptom burden), lifestyle(e.g., ability to function in normal roles), provider(e.g., provider opinion), and health care system(e.g., consistency of care) factors. Participants stronglyfelt they had ‘‘final say’’ in treatment decision-makingwhile preferring involvement of providers and familycaregivers. Most were open to using an electronic plat-form at home or in clinic, but to augment versus toreplace face time with providers.Conclusion. Participants identified preferencesregarding content and format that build on currentlyavailable clinical pathway tools.Implications for Research, Policy, or Practice. Wewill utilize data to design and test MyPATHway so thatit facilitates shared decision-making and, potentially,use of palliative care services to better set and meet pa-tient expectations.

Are Pediatric Patients Just Short Adults?Most Commonly Prescribed Drugs forPediatric Hospice Patients (S868)Leah Sera, PharmD MA BCPS, University of MarylandSchool of Pharmacy, Baltimore, MD. Jill Morgan,PharmD, University of Maryland School of Pharmacy,Baltimore, MD. Mary Lynn McPherson, PharmD MAMDE BCPS, University of Maryland School of Phar-macy, Baltimore, MD.

Objectives1. Describe prescribing practices for pediatric pa-

tients receiving hospice care.




2. Compare and contrast prescribing practices forpatients with different admitting diagnoses.

Original Research Background. Palliative care forchildren focuses on holistic care, including the allevi-ation of physical, psychosocial, and spiritual suffering.Medications are used in hospice patients to palliatephysical symptoms of terminal illness, includingpain, dyspnea, nausea, and fatigue.Research Objectives. The purpose of this study wasto characterize the most commonly prescribed medi-cations and medication classes in a population of pedi-atric hospice patients.Methods. We conducted a retrospective review of apatient information database compiled by a nationalhospice organization. The database contained demo-graphic information, as well as information on drugname, dosage, formulation, and strength. Wecompared proportions of the most commonly pre-scribed pharmacological classes among the threemost common admitting diagnoses: cancer, centralnervous system disorders, and genetic disorders.Results. A total of 3,017 medication orders wereevaluated. Six of the 10 most commonly prescribeddrugs (morphine, lorazepam, acetaminophen, hyo-scyamine, prochlorperazine, and haloperidol) wereincluded in symptom management medication kitsprovided to most patients at admission. Other drugsprescribed for over 20% of patients included meto-clopramide, diphenhydramine, albuterol, altropine,ondansetron, diazepam, polyethylene glycol, and lev-etiracetam. Opioid analgesics, anxiolytics, anticholin-ergics, and antiemetics were prescribed to over 50%of patients at some point during admission. Otherfrequently prescribed medication classes includednon-opioid analgesics, anticonvulsants, antiinfectives,laxatives, corticosteroids, acid reducers, antipsy-chotics, and vitamins/supplements.Of the 20 mostcommonly prescribed drug classes, patients with can-cer were significantly more likely than those withCNS disorders or genetic disorders to be prescribedanticholinergics (p¼0.03), antiemetics (p<0.0001),non-opioid analgesics (p¼0.003), laxatives(p¼0.003), corticosteroids (p¼0.0004), antihista-mines (p¼0.01), acid reducers (p¼0.03), and anti-psychotics (p<0.0001).Conclusion. Medications commonly prescribed forchildren receiving hospice care include those in-tended to treat symptoms including pain, dyspnea,nausea, seizures, and constipation.Implications for Research, Policy, or Practice. Ageneral understanding of medications used in hospicecare may be helpful in the development of educa-tional materials, medications guidelines and proto-cols, and questions for future research.

The Comparison of State POLST Forms:Scope of Life-Sustaining Treatment (S869)Hannah Spero, BA BSN AGPCNP Program, EmoryUniversity, Atlanta, GA. Hyejin Kim, PhD, Emory Uni-versity School of Nursing, Atlanta, GA. Mi-KyungSong, PhD RN FAAN, Emory University, Atlanta, GA.

Objectives1. Analyze the structure and categories included in

state POLST form to compare to a form in aneighboring state, and be able to state two simi-larities and two differences between forms.

2. Explain the potential implications of excluding a‘‘Goals of Care’’ section from POLST forms, bothfrom the perspective of the provider and thepatient.

Original Research Background. Physician Ordersfor Life-Sustaining Treatment (POLST) is a portablemedical order that delineates patients’ and/or surro-gates’ care preferences. Currently, all states havePOLST programs, either endorsed or in development.However, we do not know how consistent the treat-ments in POLST forms are across the states.Research Objectives.

1. Identify the variations in life-sustaining treat-ments offered in state POLST forms.

2. Discuss the potential implications of variability instate POLST forms.

Methods. State POLST forms were retrieved in May2018 from official websites or email correspondencewith state program coordinators. Data on the presenceof Goals of Care, Cardiopulmonary Resuscitation(CPR), Medical Interventions, Artificial Nutrition, An-tibiotics, and other treatment categories were ex-tracted and analyzed using descriptive statistics.Results. Of the 45 state POLST forms reviewed, onlysix included a goals-of-care section. POLST formsincluded two to eight treatment categories: one withtwo categories (CPR and medical interventions), 17with three categories (CPR, medical interventions,and artificial nutrition), 11 with four categories(CPR, medical interventions, artificial nutrition, andantibiotics), and 16 with four or more categories ofother treatments. The CPR category had CPR anddo-not-resuscitate choices in all forms. In 41 forms,the Medical Interventions category had choices ofComfort Measures Only, Limited Interventions, andFull Treatment. Forty-four forms had the ArtificialNutrition category with three choices (None, DefinedPeriod, and Long-Term) with or without additionalchoices (e.g., hydration, parenteral nutrition). Seven-teen forms included the Antibiotics category withtwo to four choices. Other treatment categoriesincluded dialysis, transfer to hospital, and bloodtransfusion.



Conclusion. Although there is some consistency inPOLST forms, significant variation exists in treatmentcategories and choices.Implications for Research, Policy, orPractice. Varied POLST forms create differences inlife-sustaining treatment options given to patientsand/or surrogates based on their state of residence.Greater discussion is needed among healthcare pro-viders, policymakers, and researchers to reconcilethis variation.

Impact of Palliative Care Interventions onHealth-Related Quality of Life (HRQOL):A Secondary Analysis of the PromotingResilience in Stress Management (PRISM)Randomized Controlled Trial (RCT) (S870)Angela Steineck, MD, Seattle Children’s Hospital, Se-attle, WA. Miranda Bradford, MS, Seattle Children’sResearch Institute, Seattle, WA. Nancy Lau, PhD, Uni-versity of Washington/Seattle Children’s, Seattle, WA.Samantha Scott, BA, Seattle Children’s Research Insti-tute, Seattle, WA. Joyce Yi-Frazier, PhD, Seattle Chil-dren’s Research Institute, Seattle, WA. AbbyRosenberg, MD MA MS, Seattle Children’s Hospital,Seattle, WA.

Objectives1. Review the concept of health-related quality of

life as a study outcome.2. Evaluate the effectiveness of an intervention us-

ing patient-reported health-related quality oflife as an outcome.

Original Research Background. PRISM is a novelresilience-building intervention for adolescents andyoung adults (AYAs). Primary analysis of the RCT inAYAs with cancer showed PRISM improved HRQOL.Research Objectives. Secondary analysis exploredchanges in HRQOL domains and differences betweenpatient groups.Methods. English-speaking AYAs (12-25 years) wererandomized to PRISM or usual care (UC) from 1/2015 e 10/2016. Surveys were completed at enroll-ment and six months later, using the Pediatric Qualityof Life Inventory (PedsQL) Generic Short Form (SF-15) and Cancer Module to assess HRQOL. Wecompared change scores (PRISM vs UC) by domain(PedsQL SF-15: physical, emotional, social, school;Cancer: pain, nausea, procedure anxiety, treatmentanxiety, worry, cognition, perceived appearance,communication). Participants were stratified by age(12-17 years vs 18-25 years) and advanced cancer status(yes/no).

Results. 74 patients (36 PRISM, 38 UC) completed 6-month assessments. 72% were 12-17 years old. 23%had advanced cancer at enrollment. PRISM improvedpatient-reported communication (UC: median [inter-quartile range, IQR] 0 [-17, 8]; PRISM: 8 [0, 25]).Younger patients benefited more, especially in PedsQLSF-15 school (12-17: UC 0 [-8, 0], PRISM 13 [0, 17]; 18-25: UC 0 [-33, 17], PRISM 0 [-25, 17]) and social do-mains (12-17: UC 0 [-33, 0], PRISM 0 [0, 8]; 18-25: UC0 [-25, 4], PRISM -17 [-25, 8]), and cancer-specificperceived appearance (12-17: UC -4 [-25, 0], PRISM 8[-8, 25]; 18-25: UC 0 [-21, 0], PRISM -8 [-25, 17]). Pa-tients with advanced cancer benefited more in cancer-specific domains nausea (no: UC 0 [-10, 15], PRISM10 [-10, 40]; yes: UC 6 [-15, 25], PRISM 35 [25, 50])and pain (no: UC 13 [-13, 25], PRISM 6 [-13, 25]; yes:UC -13 [-25, 0], PRISM 6 [-13, 25]).Conclusion. With PRISM, younger AYAs coped betterwith age-appropriate challenges and AYAs withadvanced cancer improved physical symptomHRQOL.Implications for Research, Policy, orPractice. Efficacious psychosocial intervention forAYAs.

Recruitment Outcomes Among African-American and Rural Populations withHeart Failure to an Early PalliativeCare Clinical Trial (S871)Macy Stockdill, RN, University of Alabama at Birming-ham, Birmingham, AL. Rachel Wells, MSN RN, CNL,University of Alabama at Birmingham, Birmingham,AL. James Dionne-Odom, PhD RN ACHPN, Universityof Alabama at Birmingham, Birmingham, AL. AndresAzuero, PhD, University of Alabama at Birmingham,Birmingham, AL. Deborah Ejem, PhD, University ofAlabama at Birmingham, Birmingham, AL. Sally Eng-ler, MPH, University of Alabama at Birmingham, Bir-mingham, AL. Karen Steinhauser, PhD, Duke andVA Medical Center, Durham, NC. Salpy Pamboukian,MD MSPH, University of Alabama at Birmingham, Bir-mingham, AL. Kathryn Burgio, PhD, Birmingham VAMedical Center, Birmingham, AL. Raegan Durant,MD, University of Alabama at Birmingham, Birming-ham, AL. Elizabeth Kvale, MD FAAHPM, Dell MedicalSchool, University of Texas, Austin, TX. RodneyTucker, MD MMM FAAHPM, University of Alabamaat Birmingham Center for Palliative and SupportiveCare, Birmingham, AL. Keith Swetz, MD MA HMDCFACP FAAHPM, University of Alabama at BirminghamCenter for Palliative and Supportive Care, Birming-ham, AL. Marie Bakitas, DNSc NP-C FAAN, University




of Alabama at Birmingham School of Nursing, Bir-mingham, AL.

Objectives1. Examine relationships between sociodemo-

graphic factors and enrollment status in an earlypalliative care interventional clinical trial in theDeep South.

2. Describe baseline patient-reported outcome mea-sures of quality of life, symptom burden, mood[anxiety/depression], activation, coping style, so-cial support, self-reported health, health literacy,and spiritual/religious coping.

Original Research Background. Improving minor-ity and underserved rural patient recruitment to palli-ative care (PC) clinical trials is a high clinical researchpriority area. We conducted a randomized trial ofENABLE CHF-PC (Educate, Nurture, Advise, BeforeLife Ends Comprehensive Heartcare for Patients andCaregivers) early PC telehealth intervention for per-sons with heart failure.Research Objectives. To explore racial (Black vsWhite) and rural (vs. urban/suburban dwellers)ENABLE CHF-PC participant enrollment rates anddifferences.Methods. Using bivariate tests and measures of asso-ciation, we compared consented and non-consentedpatients by race and residency status. Among random-ized participants we examined baseline differences insociodemographic/clinical characteristics and base-line patient-reported outcomes (PROs) (quality oflife [QOL], symptom burden, mood, resource use,activation, coping style, social support, health literacy,and spiritual/religious coping).Results. Of 564 eligible patients, 82% (n¼461) con-sented and 90% (n¼415) of those were randomized.Of 103 eligible participants who declined, 73.5%(n¼75) were White compared to only 25% (n¼26)who were Black (d-equivalent ¼0.57; p<0.01). Ofthose randomized, 26 % (n¼108) resided in a ruralarea (by RUCA criteria). Among rural residentsthere was a higher portion of White 58.3%(nw¼63) compared to Black participants 40.7%(nB¼44) (d-equivalent¼0.34; p¼0.124). BaselinePROs were not statistically different based on ru-ral/urban location.Conclusion. Higher proportions of Black and ruralpatients were recruited, relative to the Alabama popu-lation (26% African American; 23% Rural). At base-line, patients reported a fair QOL, low anxiety/depression and average pain. Minority and rural pa-tient recruitment is possible despite documented PCdisparities.Implications for research, policy orpractice. High recruitment rates of Black patientsmay be attributable to racial congruence between

recruiters and patients. We found no PRO differencesbased on racial or rural status; lack of disparities maybe due to similar access to care and socioeconomic sta-tus.

An Opportunity for Palliative Care:Symptom Burden in Patients withAmyloidosis (S872)Jessica Tsukanov, DO, Virginia Commonwealth Uni-versity, Richmond, VA. Egidio Del Fabbro, MD, Vir-ginia Commonwealth University, Richmond, VA.Malisa Dang, BS, Virginia Commonwealth University,Richmond, VA. Madeline Underwood, MS, VirginiaCommonwealth University School of Medicine, Rich-mond, VA.

Objectives1. Evaluate symptom burden of patients undergo-

ing treatment for amyloidosis.2. Justify the need for palliative specialists to be a

part of standard of care for patients with amyloid.3. Propose a model for palliative care to be inte-

grated with standard of care for patients withamyloidosis.

Original Research Background. Although signifi-cant symptom burden has been described in patientswith amyloidosis, few studies have used a validatedtool to assess symptom burden or severity. Amyloidosishas many similarities to cancer; prognosis is poor, mul-tiple organs may be involved, and treatment toxicitiesare common.Research Objectives. To evaluate baseline symptomseverity using the Edmonton Symptom AssessmentSystem (ESAS), and the change in total symptomburden between outpatient hematology-oncologyclinic visits in patients with amyloidosis.Methods. Retrospective chart review of 48 consecu-tive patients with amyloidosis presenting to a cancercenter outpatient clinic. ESAS and Total SymptomDistress Scores (TSDS) representing the combinedphysical, emotional, and well-being intensities were re-ported at each visit. Individual ESAS scores $4 wereconsidered moderate to severe intensity. The minimalclinically important difference (MCID) was calculatedfor TSDS between first and second visits. Based on pastresearch cutoffs used were $+3/90 for improvementand #-4/90 for deterioration. Patients being co-managed with at least one appointment with palliativecare (PC) during the study period, were noted.Results. Baseline ESAS scores were collected on 48 pa-tients with 31 also receiving a second. At baseline 34(71%) had at least one or more moderate-severe symp-toms. Fatigue in23 (48%)and insomnia 42%weremostcommon. The MCID for TSDS showed 35% of patientsimproved while 35% showed deterioration. Only 6(13%) of patients received palliative care.



Conclusion. Many patients with amyloidosis experi-ence high symptom burden, and although the major-ity report moderate/ severe fatigue, and insomnia, feware co-managed by PC. About one third had improvedsymptom burden at their second visit.Implications for Research, Policy, orPractice. Patients with amyloidosis have high symp-tom burden and may benefit from specialist palliativecare. Further research is required to establish practiceprotocols and evaluate outcomes.

Testing Usability and Acceptance of theElectronic Patient Visit Assessment (ePVA)for Head and Neck Cancer: An IterativeProcess (S873)Janet Van Cleave, PhD RN, New York UniversityMeyers College of Nursing, New York, NY. Mei Fu,PhD, New York University, New York, NY. Antonia Ben-nett, PhD, University of North Carolina, Chapel Hill,NC. Mark Persky, MD, New York University MedicalCenter, New York, NY.

Objectives1. Describe usability testing of a web-based patient-

report application for assessment of symptoms.2. Evaluate the effectiveness of the usability testing

and refinement of the electronic Patient VisitAssessment by the study population.

Original Research Background. Patients with headand neck cancer experience substantial symptomburden. A clinically useful tool is needed to evaluatesymptoms for early detection of symptoms and func-tional limitations. Therefore, we developed a web-based electronic patient visit assessment (ePVA) forhead and neck cancer. Using an iterative process toidentify issues related to usability of the tool is imper-ative for the implementation of the ePVA in clinicalsettings.Research Objectives. 1) Determine usability of theePVA in head and neck cancer, and 2) refine theePVA system over time based on patients’ suggestionsfrom the iterative process.Methods. Study design consisted of usability testingusing the Think Aloud technique to guide the iterativeprocess to refine the ePVA based on participants’ eval-uations. After informed consent, 30 participants withhead and neck cancer (Mean age ¼ 61, 67% Male,70% White, 83% stage IV cancer) completed theePVA using iPads while thinking aloud about ease ofuse. Following ePVA completion, participantsanswered a valid and reliable survey about usability.All patient conversations were recorded, transcriptswere analyzed using thematic analysis.Results. Majority of participants reported symptoms(oral symptoms: 93%, fibrosis: 60%, fatigue: 60%);90% strongly agreed/agreed that the system was easy

to use and 80% were very satisfied. Only minor usabil-ity problems were reported, decreasing in frequencyover the study period. No usability problems were re-ported by the last 3 participants who completed theePVA. Based on patients’ suggestions from the itera-tive process, refinement of the ePVA includedincreased touch sensitivity and customized error mes-sages to improve ease of use.Conclusion. Study findings indicate that the ePVA iseasy to use and has good acceptance by the studypopulation.Implications for Research, Policy, orPractice. Future research using the ePVA includesmapping of longitudinal trajectory of symptoms andclinical usefulness studies.

Bereavement Interventions for GrievingFamily Members: A Systematic Review(S874)Debra Wiegand, PhD RN CHPN FAAN FPCN, Univer-sity of Maryland, Baltimore, MD. In Seo La, MSN RN,University of Maryland School of Nursing, Baltimore,MD.

Objectives1. Describe psychoeducational bereavement inter-

ventions that can be offered to family membersduring the first year of bereavement.

2. Identify family member outcomes that can bemeasured in an effort to determine if bereave-ment interventions are effective.

Background and Objectives. Bereavement care ispart of palliative care that continues after patientdeaths. It is important to determine what interven-tions are helpful and contribute to positive outcomessince supporting the bereaved may prevent long-term negative problems. The purpose of this system-atic review was to review and evaluate the evidenceregarding bereavement interventions for bereavedadults during the first year of bereavement.Study Identification. Searches of MEDLINE, Em-base, CINAHl, PscyINFO, and the Cochrane CentralRegister of Controlled Trials were conducted. Searchterms used were grief, bereavement, mourning, inter-vention, program, support, therapy, outreach andcounsel. A comprehensive review was conducted ofbereavement intervention investigations published be-tween 1979 and 2018.Data Extraction and Synthesis. Twenty-five investi-gations were identified and analyzed. Of the 25 inves-tigations 16 were randomized control trials and 9 werequasi experimental investigations. The interventionsdesigned and tested included psychoeducational in-terventions (6), support group interventions (7), 1:1support interventions provided by professionals, non-professionals or a combination of both (4), and




additional interventions (8). Examples of additionalinterventions included post death letters and postdeath telephone calls.Results. The review demonstrated that psycho-educa-tional interventions (including two web-based inter-ventions) were most effective. Support groupinterventions and 1:1 support interventions hadmixed results. Effective additional interventionsincluded family focused grief therapy and an interven-tion that taught breathing and relaxation exercises.Positive family member outcomes included a decreasein symptoms of anxiety, depression and post-traumaticstress, decreased levels of grief and increased coping.Conclusions and Implications for Practice, Policy,and Research. The results of this systematic reviewdemonstrated that there are grief interventions thatare effective and can be used to support adult familymembers during the first year of bereavement. Basedon the results of this systematic review specific bereave-ment interventions can be safely offered to bereavedfamily members. Additional research is needed withdiverse populations.

Cannabis Use Among Patients PrescribedOpioids in a Palliative Care Clinic (S875)Matthew Wilson, MD, Dartmouth Hitchcock MedicalCenter, Lebanon, NH. Emily Masterson, MPH, GeiselSchool of Medicine at Dartmouth, Hanover, NH.Kathleen Broglio, DNP ACHPN ANP-BC CPE FPCN,Dartmouth Hitchcock Medical Center, Lebanon, NH.

Objectives1. State the approximate percentage of patients

receiving opioids in a palliative care clinic whoalso use cannabis.

2. Describe the most common symptoms for whichpatients use cannabis in our clinic population.

Original Research Background. Cannabis use isincreasing in the United States. Palliative care patientsutilize it to manage various symptoms, but there is lit-tle data about its use in this population.Research Objectives. To assess cannabis use amongpalliative care clinic patients being prescribed opioids.Methods. We conducted a retrospective chart review(October 2017, January 2018 and April 2018) usinga convenience sample of patients being prescribedopioid therapy in a palliative care clinic at a rural, ter-tiary care, academic system.Results. During this time period, 174 patientsreceived prescriptions for opioids (98% for cancer-related pain, 4% dyspnea). Seventy-eight patients(45%) were using cannabis, and of these, 57 (73%)had the indication documented. Among those usingcannabis, only 3 (4%) indicated using cannabis recrea-tionally. Documented reasons for cannabis useincluded: pain (55, 96%), insomnia (12, 21%), nausea

(10, 18%), appetite (10, 18%), anxiety (5, 9%),depression (3, 5%), and seizures (1, 2%). Twenty pa-tients (35%) used therapeutic cannabis for morethan one indication. Routes of administration forthe 38 patients for whom this was documentedincluded: smoking (12, 32%), vaping (15, 40%), tinc-ture (7, 18%), edible (6, 16%), and topical (6, 16%).Data was lacking on the formulations (CBD:THC ra-tios) of cannabis utilized and frequency of use.Conclusions. Cannabis use is common among pa-tients being prescribed opioids by a palliative careclinic. Indications and routes of administration arenumerous. Our data lacks information on frequencyand effectiveness of use for a given symptom. Use ofcannabis and other complementary and alternativetreatments should be assessed in palliative carepatients.Implications for Research, Policy, orPractice. Future research should investigate: betterunderstanding of cannabinoid component formula-tions utilized for symptom relief, patients’ perceivedefficacy and side effects of cannabis as compared toFDA-approved medications for symptom manage-ment, safety profile of cannabis in combination withopioids, and whether cannabis use affects usage of opi-oids for cancer-related pain.

How are Pediatric Tracheostomy DecisionsDiscussed? An Analysis of PediatricTracheostomy Decision Processes (S876)Haoyang Yan, MS, University of Michigan, Ann Arbor,MI. Cynthia Arslanian-Engoren, PhD RN, University ofMichigan, Ann Arbor, MI. Stephanie Kukora, MD,University of Michigan Medical Center, Ann Arbor,MI. Kenneth Pituch, MD, University of Michigan,Ann Arbor, MI. Tessie October, MD, Children’s Na-tional Medical Center, Washington, D.C. Patricia Del-din, PhD, University of Michigan, Ann Arbor, MI.Sandra Graham-Bermann, PhD BA MA, University ofMichigan, Ann Arbor, MI. J Frank Yates, PhD, the Uni-versity of Michigan, Ann Arbor, MI.

Objectives1. Describe pediatric tracheostomy decision discus-

sions through the lens of the 10 Cardinal Issuesfrom the Cardinal Issues Perspective (CIP) on de-cision making.

2. Interpret pediatric tracheostomy decision discus-sions through the lens of the 10 Cardinal Issuesfrom the Cardinal Issues Perspective (CIP) on de-cision making.

Original Research Background. Parents of criticallyill children who are ventilator dependent are oftenasked to consider the placement of a tracheostomy.This decision has substantial implications over the




life of the child and the family. Many healthcare pro-fessionals and parents find this decision process diffi-cult. Understanding tracheostomy decision processesmay inform researchers to develop decisions toolsfor promoting high-quality provider-parent communi-cation and decision making.Research Objectives. Our study aimed to describeand to interpret pediatric tracheostomy decision dis-cussions according to the 10 Cardinal Issues from theCardinal Issue Perspective (CIP) on decision making.The Cardinal Issue Perspective is a well-recognizedapproach for satisfying decisions in non-medicalhigh-stakes economic situations.Methods. We conducted a qualitative deductive the-matic analysis of 19 de-identified transcripts of familyconferences that discussed tracheostomy decisions.Results. Five Cardinal IssuesdNeed, Options, Possibil-ities, Judgment, and Implementationdwere discussed

more frequently than Mode, Investment, Value, Tradeoffs,and Acceptability.Conclusions. Our qualitative analysis of real-life fam-ily conferences revealed that some aspects of decisionmaking were discussed more frequently than others.In particular, decision process costs, values, and goalswere discussed infrequently and thus suggest potentialfor improving tracheostomy decision processes.Implications for Research, Policy, orPractice. Our work provides researchers, healthcareproviders, and policy makers a better understandingof pediatric tracheostomy decision processes. CardinalIssues may have the potential to structure and facilitatepediatric tracheostomy and other life-support-relatedmedical decision making and communication pro-cesses. Our research findings may also inform medicalschool curriculum, to educate medical students howto facilitate discussions with families.


scientiﬁc and quality improvement project poster...

Documents