Equity in Using Artificial Intelligence Mortality Predictions to Target Goals of Care Documentation.

BACKGROUND: Artificial intelligence (AI) algorithms are increasingly used to target patients with elevated mortality risk scores for goals-of-care (GOC) conversations. OBJECTIVE: To evaluate the association between the presence or absence of AI-generated mortality risk scores with GOC documentation. DESIGN: Retrospective cross-sectional study at one large academic medical center between July 2021 and December 2022. PARTICIPANTS: Hospitalized adult patients with AI-defined Serious Illness Risk Indicator (SIRI) scores indicating > 30% 90-day mortality risk (defined as "elevated" SIRI) or no SIRI scores due to insufficient data. INTERVENTION: A targeted intervention to increase GOC documentation for patients with AI-generated scores predicting elevated risk of mortality. MAIN MEASURES: Odds ratios comparing GOC documentation for patients with elevated or no SIRI scores with similar severity of illness using propensity score matching and risk-adjusted mixed-effects logistic regression. KEY RESULTS: Among 13,710 patients with elevated (n = 3643, 27%) or no (n = 10,067, 73%) SIRI scores, the median age was 64 years (SD 18). Twenty-five percent were non-White, 18% had Medicaid, 43% were admitted to an intensive care unit, and 11% died during admission. Patients lacking SIRI scores were more likely to be younger (median 60 vs. 72 years, p < 0.0001), be non-White (29% vs. 13%, p < 0.0001), and have Medicaid (22% vs. 9%, p < 0.0001). Patients with elevated versus no SIRI scores were more likely to have GOC documentation in the unmatched (aOR 2.5, p < 0.0001) and propensity-matched cohorts (aOR 2.1, p < 0.0001). CONCLUSIONS: Using AI predictions of mortality to target GOC documentation may create differences in documentation prevalence between patients with and without AI mortality prediction scores with similar severity of illness. These finding suggest using AI to target GOC documentation may have the unintended consequence of disadvantaging severely ill patients lacking AI-generated scores from receiving targeted GOC documentation, including patients who are more likely to be non-White and have Medicaid insurance.

AI-Generated Content in Cancer Symptom Management: A Comparative Analysis Between ChatGPT and NCCN.

CONTEXT: Artificial intelligence-driven tools, like ChatGPT, are prevalent sources for online health information. Limited research has explored the congruity between AI-generated content and professional treatment guidelines. OBJECTIVES: This study seeks to compare recommendations for cancer-related symptoms generated from ChatGPT with guidelines from the National Comprehensive Cancer Network (NCCN). METHODS: We extracted treatment recommendations for nine symptoms from NCCN, separated into four full Supportive Care sections and five subsections of the Palliative Care webpage. We entered "How can I reduce my cancer-related [symptom]" into ChatGPT- 3.5 for these same symptoms and extracted its recommendations. A comparative content analysis focused on recommendations for medications, consultations, and non-pharmacological strategies. We compared word count and Flesch-Kincaid Grade Level (FKGL) readability for each NCCN and ChatGPT section. RESULTS: The mean percent agreement between NCCN and ChatGPT recommendations was 37.3% (range 16.7%- 81.8%). NCCN offered more specific medication recommendations. ChatGPT did recommend medications in the constipation and diarrhea sections that were not recommended by NCCN. Significant differences in word count (p=0.03) and FKGL (p<0.01) were found for NCCN Supportive Care webpages, with ChatGPT having lower word count and reading level. In the NCCN Palliative Care webpage subsections, there was no significant difference in word count (p=0.076), but FKGL was significantly lower with ChatGPT (p<0.01). CONCLUSIONS: While ChatGPT provides concise, accessible supportive care advice, discrepancies with guidelines raise concerns for patient-facing symptom management recommendations. Future research should consider how AI can be used in conjunction with evidence-based guidelines to support cancer patients' supportive care needs. KEY MESSAGE: This study compares cancer-related symptom recommendations from ChatGPT with NCCN guidelines, revealing significant differences and highlighting the need for cautious integration of AI-generated content in patient care. It emphasizes the importance of professional guidance and collaboration to ensure data accuracy and quality.

Artificial intelligence for better goals of care documentation.

OBJECTIVES: Lower rates of goals of care (GOC) conversations have been observed in non-white hospitalised patients, which may contribute to racial disparities in end-of-life care. We aimed to assess how a targeted initiative to increase GOC documentation rates is associated with GOC documentation by race. METHODS: We retrospectively assessed GOC documentation during a targeted GOC initiative for adult patients with an artificial intelligence predicted elevated risk of mortality. Patients were admitted to an urban academic medical centre in Pittsburgh, Pennsylvania between July 2021 and 31 December 2022. RESULTS: The 3643 studied patients had a median age of 72 (SD 13.0) and were predominantly white (87%) with 42% admitted to an intensive care unit and 15% dying during admission. GOC documentation was completed for 28% (n=1019/3643). By race, GOC was documented for 30% black (n=105/351), 28% white (n=883/3161) and 24% other (n=31/131) patients (p=0.3933). There was no statistical difference in the rate of documented GOC among races over time (p=0.5142). CONCLUSIONS: A targeted initiative to increase documented GOC conversations for hospitalised patients with an elevated risk of mortality is associated with similar documentation rates across racial groups. Further research is needed to assess whether this initiative may promote racial equity in GOC documentation in other settings.

Associations of Health Care Utilization and Therapeutic Alliance in Patients with Advanced Cancer.

Introduction: Therapeutic alliance (TA), or the extent to which patients feel a sense of caring and trust with their physician, may have an impact on health care utilization. We sought to determine if TA is associated with: (1) emergency department (ED) visits within 30 days of death and (2) hospice enrollment. Methods and Materials: This is a secondary analysis of data from a randomized clinical trial. We used restricted cubic splines to assess the relationship between TA scores and health care utilization. Results: Six hundred seventy-two patients were enrolled in the study, with 331 (49.3%) dying within 12 months. Patients with higher TA were less likely to have an ED visit in the last 30 days of life, but there was no evidence of a relationship between TA and enrollment in hospice. Conclusions: Higher TA was associated with decreased ED visits within 30 days of death. There was no association between TA and rates of hospice enrollment. Clinical Registration Number: NCT02712229.

Implementing a Palliative Care Junior Faculty Visiting Professor Program: Pearls and Pitfalls.

MOTIVATION: Palliative Care (PC) is a small, relatively young interprofessional sub-specialty; hence mentorship for early-career research faculty is widely dispersed across schools and universities. We developed the Junior Visiting Professor Program (JVPP) to provide junior faculty in palliative care (PC) with opportunities to meet multidisciplinary PC researchers from other institutions and to advance their research through networking and presenting their work. We describe how we designed and implemented the program, and we report on the first cohort of participants. METHODS: We invited PC research groups from US schools of medicine and nursing to participate in this 5-year interprofessional exchange program by nominating junior faculty and serving as hosts. We matched nominees to host institutions based on nominee training experiences, nominee research interests, and host institution faculty expertise. In addition, we provided logistical guidance on visit planning. Post-visit, we surveyed both hosts and junior visiting professors (JVPs) regarding their satisfaction, perceived value, and suggestions regarding the program. RESULTS: We recruited 13 schools to participate and matched 10 nominees to host institutions in our first year. Nine JVPs completed their visit; 6 JVPs and 8 host faculty/staff responded to the post-visit survey. Overall, JVPs were highly satisfied with their matches and the visiting professor experience. Hosts were generally satisfied with their matches and believed the program to be mutually beneficial. The most frequent suggestion was for greater administrative support to plan visits. CONCLUSIONS: Structured, well-supported opportunities for networking across institutions is beneficial for emerging PC researchers and for building PC research capacity.

Conversational Agents in Palliative Care: Potential Benefits, Risks, and Next Steps.

Conversational agents (sometimes called chatbots) are technology-based systems that use artificial intelligence to simulate human-to-human conversations. Research on conversational agents in health care is nascent but growing, with recent reviews highlighting the need for more robust evaluations in diverse settings and populations. In this article, we consider how conversational agents might function in palliative care-not by replacing clinicians, but by interacting with patients around select uncomplicated needs while facilitating more targeted and appropriate referrals to specialty palliative care services. We describe potential roles for conversational agents aligned with the core domains of quality palliative care and identify risks that must be considered and addressed in the development and use of these systems for people with serious illness. With careful consideration of risks and benefits, conversational agents represent promising tools that should be explored as one component of a multipronged approach for improving patient and family outcomes in serious illness.

Hope and illness expectations: A cross-sectional study in patients with advanced cancer.

BACKGROUND: The fear of taking away hope hinders clinicians' willingness to share serious news with patients with advanced disease. Unrealistic illness expectations, on the other hand, can complicate decision making and end-of-life care outcomes. Exploration of the association between hope and illness expectations can support clinicians in better communication with their patients. AIM: The aim of this study was to explore whether realistic illness expectations are associated with reduced hope in people with advanced cancer. DESIGN: This is a cross-sectional secondary analysis of baseline data from a primary palliative care cluster-randomized trial CONNECT (data collected from July 2016 to October 2020). Hope was measured by Herth Hope Index. Illness expectations were measured by assessing patients' understanding of their treatment goals, life expectancy, and terminal illness acknowledgement. Multivariable regression was performed, adjusting for demographical and clinical confounders. SETTING/PARTICIPANTS: Adult patients with advanced solid cancers recruited across 17 oncology clinics. RESULTS: Data from 672 patients were included in the study, with mean age of 69.3 years (±10.2), 53.6% were female. Proportion of patients indicating realistic expectations varied based on which question was asked from 10% to 46%. Median level of hope was 39 (IQR = 36-43). Multivariate non-inferiority regression did not find any significant differences in hope between patients with more and less realistic illness expectations. CONCLUSIONS: Our results suggest that hope can be sustained while holding both realistic and unrealistic illness expectations. Communication about serious news should focus on clarifying the expectations as well as supporting people's hopes.

Defining the Time-limited Trial for Patients with Critical Illness: An Official American Thoracic Society Workshop Report.

In critical care, the specific, structured approach to patient care known as a "time-limited trial" has been promoted in the literature to help patients, surrogate decision makers, and clinicians navigate consequential decisions about life-sustaining therapy in the face of uncertainty. Despite promotion of the time-limited trial approach, a lack of consensus about its definition and essential elements prevents optimal clinical use and rigorous evaluation of its impact. The objectives of this American Thoracic Society Workshop Committee were to establish a consensus definition of a time-limited trial in critical care, identify the essential elements for conducting a time-limited trial, and prioritize directions for future work. We achieved these objectives through a structured search of the literature, a modified Delphi process with 100 interdisciplinary and interprofessional stakeholders, and iterative committee discussions. We conclude that a time-limited trial for patients with critical illness is a collaborative plan among clinicians and a patient and/or their surrogate decision makers to use life-sustaining therapy for a defined duration, after which the patient's response to therapy informs the decision to continue care directed toward recovery, transition to care focused exclusively on comfort, or extend the trial's duration. The plan's 16 essential elements follow four sequential phases: consider, plan, support, and reassess. We acknowledge considerable gaps in evidence about the impact of time-limited trials and highlight a concern that if inadequately implemented, time-limited trials may perpetuate unintended harm. Future work is needed to better implement this defined, specific approach to care in practice through a person-centered equity lens and to evaluate its impact on patients, surrogates, and clinicians.

Talking About Dying.

This essay describes a palliative medicine physician's experience with a dying patient and reflections on life.

Strategies to Improve Perioperative Palliative Care Integration for Seriously Ill Veterans.

CONTEXT: Seriously ill patients are at higher risk for adverse surgical outcomes. Palliative care (PC) interventions for seriously ill surgical patients are associated with improved quality of patient care and patient-centered outcomes, yet, they are underutilized perioperatively. OBJECTIVES: To identify strategies for improving perioperative PC integration for seriously ill Veterans from the perspectives of PC providers and surgeons. METHODS: We conducted semistructured, in-depth individual and group interviews with Veteran Health Administration PC team members and surgeons between July 2020 and April 2021. Participants were purposively sampled from high- and low-collaboration sites based on the proportion of received perioperative palliative consults. We performed a team-based thematic analysis with dual coding (inter-rater reliability above 0.8). RESULTS: Interviews with 20 interdisciplinary PC providers and 13 surgeons at geographically distributed Veteran Affairs sites converged on four strategies for improving palliative care integration and goals of care conversations in the perioperative period: 1) develop and maintain collaborative, trusting relationships between palliative care providers and surgeons; 2) establish risk assessment processes to identify patients who may benefit from a PC consult; 3) involve both PC providers and surgeons at the appropriate time in the perioperative workflow; 4) provide sufficient resources to allow for an interdisciplinary sharing of care. CONCLUSION: The study demonstrates that individual, programmatic, and organizational efforts could facilitate interservice collaboration between PC clinicians and surgeons.

Implementing a Serious Illness Risk Prediction Model: Impact on Goals of Care Documentation.

CONTEXT: Goals of care conversations can promote high value care for patients with serious illness, yet documented discussions infrequently occur in hospital settings. OBJECTIVES: We sought to develop a quality improvement initiative to improve goals of care documentation for hospitalized patients. METHODS: Implementation occurred at an academic medical center in Pittsburgh, Pennsylvania. Intervention included integration of a 90-day mortality prediction model grouping patients into low, intermediate, and high risk; a centralized goals of care note; and automated notifications and targeted palliative consults. We compared documented goals of care discussions by risk score before and after implementation. RESULTS: Of the 12,571 patients hospitalized preimplementation and 10,761 postimplementation, 1% were designated high risk and 11% intermediate risk of mortality. Postimplementation, goals of care documentation increased for high (17.6%-70.8%, P< 0.0001) and intermediate risk patients (9.6%-28.0%, P < 0.0001). For intermediate risk patients, the percentage of goals of care documentation performed by palliative medicine specialists increased from pre- to postimplementation (52.3%-71.2%, P = 0.0002). For high-risk patients, the percentage of goals of care documentation completed by the primary service increased from pre-to postimplementation (36.8%-47.1%, P = 0.5898, with documentation performed by palliative medicine specialists slightly decreasing from pre- to postimplementation (63.2%-52.9%, P = 0.5898). CONCLUSIONS: Implementation of a goals of care initiative using a mortality prediction model significantly increased goals of care documentation especially among high-risk patients. Further study to assess strategies to increase goals of care documentation for intermediate risk patients is needed especially by nonspecialty palliative care.

Rural Hospital Disparities in Goals of Care Documentation.

CONTEXT: Goals of care conversations for seriously ill hospitalized patients are associated with high-quality patient-centered care. OBJECTIVES: We aimed to assess the prevalence of documented goals of care conversations for rural hospitalized patients compared to nonrural hospitalized patients. METHODS: We retrospectively assessed goals of care documentation using a template note for adult patients with predicted 90-day mortality greater than 30% admitted to eight rural and nine nonrural community hospitals between July 2021 and April 2023. We compared predictors and prevalence of goals of care documentation among rural and nonrural hospitals. RESULTS: Of the 31,098 patients admitted during the study period, 21% were admitted to a rural hospital. Rural patients were more likely than nonrural patients to be >65 years old (89% vs. 86%, P = <.0001), more likely to live in a neighborhood classified in the highest quintile of socioeconomic disadvantage (40% vs. 16%, P = <.0001), and less likely to receive a palliative care consult (8% vs. 18%, P = <.0001). Goals of care documentation occurred less often for patients admitted to rural vs. nonrural community hospitals (2% vs. 7%, P < .0001). In the base multivariable logistic regression model adjusting for patient characteristics, the odds of goals care documentation were lower in rural vs. nonrural community hospitals (aOR 0.4, P = .0232). In a second multivariable logistic regression model including both patient characteristics and severity of illness, the odds of goals of care documentation in rural community hospitals were no longer statistically different than nonrural community hospitals (aOR 0.5, P = .1080). Patients who received a palliative care consult had a lower prevalence of goals of care documentation in rural vs. nonrural hospitals (16% vs. 37%, P = <.0001). CONCLUSION: In this study of 17 rural and nonrural community hospitals, we found low overall prevalence of goals of care documentation with particularly infrequent documentation occurring within rural hospitals. Future study is needed to assess barriers to goals of care documentation contributing to low prevalence of goals of care conversations in rural hospital settings.

The feasibility, acceptability, and preliminary efficacy of a self-advocacy serious game for women with advanced breast or gynecologic cancer.

BACKGROUND: Cancer clinicians and systems aim to provide patient-centered care, but not all patients have the self-advocacy skills necessary to ensure their care reflects their needs and priorities. This study examines the feasibility, acceptability, and preliminary efficacy of a self-advocacy serious game (an educational video game) intervention in women with advanced breast or gynecologic cancer. METHODS: Women with recently diagnosed (<3 months) metastatic breast or advanced gynecologic cancer were randomized 2:1 to receive a tablet-based serious game (Strong Together) (n = 52) or enhanced care as usual (n = 26). Feasibility was based on recruitment, retention, data completion, and intervention engagement. Acceptability was assessed via a postintervention questionnaire and exit interview. Preliminary efficacy was assessed on the basis of change scores from baseline to 3 and 6 months in self-advocacy (Female Self-Advocacy in Cancer Survivorship Scale) using intention-to-treat analysis. RESULTS: Seventy-eight women (55.1% with breast cancer; 44.9% with gynecologic cancer) were enrolled. Feasibility was demonstrated by satisfactory recruitment (69% approach-to-consent rate; 93% enroll-to-randomize rate), retention (90% and 86% at 3 and 6 months, respectively; 85% data completion), and intervention engagement (84% completed ≥75% of the game). Participants endorsed the intervention's (75%) and trial's (87%) acceptability. Participants in the intervention group experienced significant improvements in self-advocacy at 3 and 6 months compared to participants in the control group. CONCLUSIONS: Strong Together is feasible and acceptable among women with advanced breast or gynecologic cancer. This intervention demonstrates promising evidence of clinical efficacy. A future confirmatory trial is warranted to test the efficacy of the intervention for patient and health system outcomes.

Primary Palliative Care Improves Uptake of Advance Care Planning Among Patients With Advanced Cancers.

BACKGROUND: Palliative care specialists are experts in conducting advance care planning (ACP) but are a limited resource. Oncology nurses often have special relationships with their patients and thus may be poised to provide primary palliative care. We sought to determine the impact of a nurse-led primary palliative care intervention on ACP uptake among patients with advanced cancer. METHODS: We performed a secondary analysis of a cluster randomized controlled trial examining the impact of nurse-based primary palliative care. In the parent trial, patients with advanced cancer received either monthly primary palliative care visits with trained nurses within their cancer center or standard care. Nurses in the intervention arm received special training in ACP. ACP uptake was assessed at enrollment and 3 months later evaluating (1) whether an end-of-life conversation (EOLC) occurred with one's oncologist, and (2) completion of an advance directive (AD). Multivariable logistic regression tested differences in ACP uptake by treatment arm adjusted for age, religious importance, education, time with current oncologist, and performance status. RESULTS: Of 672 patients enrolled, 182/336 (54%) patients in the intervention arm and 196/336 (58%) in the standard care arm lacked an EOLC at baseline and completed the 3-month assessment. Of those, 82/182 (45.1%) patients in the intervention arm and 29/196 (14.8%) in the standard care arm reported having an EOLC at 3 months (adjusted odds ratio, 5.28; 95% CI, 3.10-8.97; P<.001). Similarly, 111/336 (33%) patients in the intervention arm and 105/336 (31%) in the standard care arm lacked an AD at baseline and completed the 3-month assessment. Of those, 48/111 (43.2%) patients in the intervention arm and 19/105 (18.1%) in the standard care arm completed an AD over the study period (adjusted odds ratio, 3.68; 95% CI, 1.89-7.16; P<.001). CONCLUSIONS: Nurse-led primary palliative care increased ACP uptake among patients with advanced cancer. Training oncology nurses embedded within community cancer centers to provide primary palliative care may help improve ACP access.

Disparities in Unilateral Do Not Resuscitate Order Use During the COVID-19 Pandemic.

OBJECTIVES: A unilateral do-not-resuscitate (UDNR) order is a do-not-resuscitate order placed using clinician judgment which does not require consent from a patient or surrogate. This study assessed how UDNR orders were used during the COVID-19 pandemic. DESIGN: We analyzed a retrospective cross-sectional study of UDNR use at two academic medical centers between April 2020 and April 2021. SETTING: Two academic medical centers in the Chicago metropolitan area. PATIENTS: Patients admitted to an ICU between April 2020 and April 2021 who received vasopressor or inotropic medications to select for patients with high severity of illness. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The 1,473 patients meeting inclusion criteria were 53% male, median age 64 (interquartile range, 54-73), and 38% died during admission or were discharged to hospice. Clinicians placed do not resuscitate orders for 41% of patients ( n = 604/1,473) and UDNR orders for 3% of patients ( n = 51/1,473). The absolute rate of UDNR orders was higher for patients who were primary Spanish speaking (10% Spanish vs 3% English; p ≤ 0.0001), were Hispanic or Latinx (7% Hispanic/Latinx vs 3% Black vs 2% White; p = 0.003), positive for COVID-19 (9% vs 3%; p ≤ 0.0001), or were intubated (5% vs 1%; p = 0.001). In the base multivariable logistic regression model including age, race/ethnicity, primary language spoken, and hospital location, Black race (adjusted odds ratio [aOR], 2.5; 95% CI, 1.3-4.9) and primary Spanish language (aOR, 4.4; 95% CI, 2.1-9.4) had higher odds of UDNR. After adjusting the base model for severity of illness, primary Spanish language remained associated with higher odds of UDNR order (aOR, 2.8; 95% CI, 1.7-4.7). CONCLUSIONS: In this multihospital study, UDNR orders were used more often for primary Spanish-speaking patients during the COVID-19 pandemic, which may be related to communication barriers Spanish-speaking patients and families experience. Further study is needed to assess UDNR use across hospitals and enact interventions to improve potential disparities.

"I Refused to Get Addicted to Opioids": Exploring Attitudes About Opioid Use Disorder in Patients With Advanced Cancer Pain and Their Support People.

Patients with advanced cancer are commonly prescribed opioids, yet patient attitudes about opioid risks (eg, opioid use disorder, or OUD) are understudied. Our objective was to use in-depth qualitative interviews to understand perceptions of opioid prescribing and OUD in patients with advanced, solid-tumor cancers and their support people. We conducted a qualitative study using a rigorous inductive, qualitative descriptive approach to examine attitudes about OUD in patients with advanced cancer (n = 20) and support providers (n = 11). Patients with cancer hold 2 seemingly distinct views: prescription opioids are addictive, yet OUD cannot happen to me or my loved one. Participants described general concerns about the addictive nature of prescription opioids ("My biggest concern… would just be the risk of getting addicted to the medication or even like, overdosing it"), while separating cancer pain management from OUD when considering prescription opioid risks and benefits ("They need to make sure they get the right ones, when they're taking it away from you."). Finally, participants identified personal characteristics and behaviors that they felt were protective against developing OUD (commonly control, willpower, and responsibility). This rigorous qualitative study demonstrates that patients with advanced cancer and their support people simultaneously hold concerns about the addictive nature of prescription opioids, while distancing from perceptions of OUD risks when using opioids for cancer pain management. Given high rates of opioid exposure during advanced cancer treatment, it is important to explore opportunities to promote a balanced understanding of prescription opioid use and OUD risks in this population. PERSPECTIVE: Though prescription opioids carry risk of OUD, there is little data to help guide patients with advanced cancer. Findings suggest that there is a need to develop new, innovative strategies to promote effective pain management and minimize opioid risks in this complex population.

An Educational Intervention to Enhance Palliative Care Training at HBCUs.

CONTEXT: Primary palliative care training is important for clinicians at Historically Black Colleges and Universities (HBCUs) given the decreased access to specialty palliative care among Black patients and patients' preferences for race concordant care. OBJECTIVES: To describe the impact of a palliative care educational intervention at two HBCUs. METHODS: We administered a palliative care educational intervention in family and internal medicine residency programs at Morehouse School of Medicine and Howard University College of Medicine. Pre- and post-intervention surveys were sent to residents assessing attitudes toward their palliative care education and their perceived competency in specific palliative care domains. The results were analyzed using Chi-squared analysis. RESULTS: A total of 105 of 191 (response rate 55%) residents completed pre-intervention surveys and 101 of 240 (42%) completed post-intervention surveys. Prior to the intervention, 50% of residents rated their overall preparedness in palliative care as a 7 or above (0-10 Likert scale); 78% (P < .01) of respondents reported ≥7/10 after the educational intervention. While post-intervention residents did not feel better prepared to treat symptoms, a higher percentage reported feeling well prepared to deliver bad news (41% post-intervention vs. 23% pre-intervention) and conduct a family meeting (44% post-intervention vs. 27% pre-intervention) (P < .05). Pre-intervention, 14% of residents felt their overall palliative care education was very good or excellent, while post-intervention ratings increased to 30% (P < .01). CONCLUSION: Residents' confidence in their preparedness to provide palliative care, particularly in their communication skills increased after an intervention at two HBCUs.