Caregiver Engagement in Serious Illness Communication in a Long-Term Acute Care Hospital Setting.

CONTEXT: Prolonged management of critical illnesses in long-term acute care hospitals (LTACH) makes serious illness communication (SIC), a clinical imperative. SIC in LTACH is challenging as clinicians often lack training and patients are typically unable to participate-making caregivers central. OBJECTIVES: This qualitative descriptive study characterized caregiver engagement in SIC encounters, while considering influencing factors, following the implementation of Ariadne Labs' SIC training at a LTACH in the Northeastern United States. METHODS: Clinicians' documented SIC notes (2019-2020) were analyzed using directed content analysis. Codes were grouped into four categories generated from two factors that influence SIC-evidence of prognostic understanding (yes/no) and documented preferences (yes/no)-and caregiver engagement themes identified within each category. RESULTS: Across 125 patient cases, 251 SIC notes were analyzed. In the presence of prognostic understanding and documented preferences, caregivers acted as upholders of patients' wishes (29%). With prognostic understanding but undocumented preferences, caregivers were postponers of healthcare decision-making (34%). When lacking prognostic understanding but having documented preferences, caregivers tended to be searchers, intent on identifying continued treatment options (13%). With poor prognostic understanding and undocumented preferences, caregivers were strugglers, having difficulty with the clinicians or family unit over healthcare decision-making (21%). CONCLUSION: The findings suggest that two factors-prognostic understanding and documented preferences-are critical factors clinicians can leverage in tailoring SIC to meet caregivers' SIC needs in the LTACH setting. Such strategies shift attention away from SIC content alone toward factors that influence caregivers' ability to meaningfully engage in SIC to advance healthcare decision-making.

Long-term Effect of Machine Learning-Triggered Behavioral Nudges on Serious Illness Conversations and End-of-Life Outcomes Among Patients With Cancer: A Randomized Clinical Trial.

Importance: Serious illness conversations (SICs) between oncology clinicians and patients are associated with improved quality of life and may reduce aggressive end-of-life care. However, most patients with cancer die without a documented SIC. Objective: To test the impact of behavioral nudges to clinicians to prompt SICs on the SIC rate and end-of-life outcomes among patients at high risk of death within 180 days (high-risk patients) as identified by a machine learning algorithm. Design, Setting, and Participants: This prespecified 40-week analysis of a stepped-wedge randomized clinical trial conducted between June 17, 2019, and April 20, 2020 (including 16 weeks of intervention rollout and 24 weeks of follow-up), included 20 506 patients with cancer representing 41 021 encounters at 9 tertiary or community-based medical oncology clinics in a large academic health system. The current analyses were conducted from June 1, 2021, to May 31, 2022. Intervention: High-risk patients were identified using a validated electronic health record machine learning algorithm to predict 6-month mortality. The intervention consisted of (1) weekly emails to clinicians comparing their SIC rates for all patients against peers' rates, (2) weekly lists of high-risk patients, and (3) opt-out text messages to prompt SICs before encounters with high-risk patients. Main Outcomes and Measures: The primary outcome was SIC rates for all and high-risk patient encounters; secondary end-of-life outcomes among decedents included inpatient death, hospice enrollment and length of stay, and intensive care unit admission and systemic therapy close to death. Intention-to-treat analyses were adjusted for clinic and wedge fixed effects and clustered at the oncologist level. Results: The study included 20 506 patients (mean [SD] age, 60.0 [14.0] years) and 41 021 patient encounters: 22 259 (54%) encounters with female patients, 28 907 (70.5%) with non-Hispanic White patients, and 5520 (13.5%) with high-risk patients; 1417 patients (6.9%) died by the end of follow-up. There were no meaningful differences in demographic characteristics in the control and intervention periods. Among high-risk patient encounters, the unadjusted SIC rates were 3.4% (59 of 1754 encounters) in the control period and 13.5% (510 of 3765 encounters) in the intervention period. In adjusted analyses, the intervention was associated with increased SICs for all patients (adjusted odds ratio, 2.09 [95% CI, 1.53-2.87]; P < .001) and decreased end-of-life systemic therapy (7.5% [72 of 957 patients] vs 10.4% [24 of 231 patients]; adjusted odds ratio, 0.25 [95% CI, 0.11-0.57]; P = .001) relative to controls, but there was no effect on hospice enrollment or length of stay, inpatient death, or end-of-life ICU use. Conclusions and Relevance: In this randomized clinical trial, a machine learning-based behavioral intervention and behavioral nudges to clinicans led to an increase in SICs and reduction in end-of-life systemic therapy but no changes in other end-of-life outcomes among outpatients with cancer. These results suggest that machine learning and behavioral nudges can lead to long-lasting improvements in cancer care delivery. Trial Registration: ClinicalTrials.gov Identifier: NCT03984773.

An accelerated hospital observation pathway to reduce length of stay for patients with COVID-19.

OBJECTIVES: Strategies to maintain hospital capacity during the COVID-19 pandemic included reducing hospital length of stay (LOS) for infected patients. We sought to evaluate the association between LOS and enrollment in the COVID Accelerated Care Pathway, which consisted of a hospital observation protocol and postdischarge automated text message-based monitoring. STUDY DESIGN: Retrospective matched cohort study of patients hospitalized from December 14, 2020, to January 31, 2021. METHODS: Participants were patients who presented to the emergency department with acute infection due to COVID-19, required hospitalization, and met pathway inclusion criteria. Participants were compared with a propensity score-matched cohort of patients with COVID-19 admitted to the same hospital during the 7 weeks preceding and following pathway implementation. RESULTS: There were 44 patients in the intervention group and 83 patients in the propensity score-matched cohort. The mean (SD) hospital LOS for patients in the intervention group was 1.7 (2.6) days compared with 3.9 (2.3) days for patients in the matched cohort (difference, -2.2 days; 95% CI, -3.3 to -1.1). In the intervention group, 2 patients (5%; 95% CI, 0%-15%) were rehospitalized within 14 days compared with 8 (10%; 95% CI, 4%-17%) in the matched cohort. CONCLUSIONS: Patients with COVID-19 who were managed through an accelerated hospital observation protocol and postdischarge monitoring service had reduced hospital LOS compared with patients receiving standard care. Hospital preparedness for future public health emergencies may involve the design of pathways that reduce the time that patients spend in the hospital, lower cost, and ensure continued recovery upon discharge.

Serious Illness Care Programme-contextual factors and implementation strategies: a qualitative study.

OBJECTIVES: The Serious Illness Care Programme (SICP) is a multicomponent evidence-based intervention that improves communication about patients' values and goals in serious illness. We aim to characterise implementation strategies for programme delivery and the contextual factors that influence implementation in three 'real-world' health system SICP initiatives. METHODS: We employed a qualitative thematic framework analysis of field notes collected during the first 1.5 years of implementation and a fidelity survey. RESULTS: Analysis revealed empiric evidence about implementation and institutional context. All teams successfully implemented clinician training and an electronic health record (EHR) template for documentation of serious illness conversations. When training was used as the primary strategy to engage clinicians, however, clinician receptivity to the programme and adoption of conversations remained limited due to clinical culture-related barriers (eg, clinicians' attitudes, motivations and practice environment). Visible leadership involvement, champion facilitation and automated EHR-based data feedback on documented conversations appeared to improve adoption. Implementing these strategies depended on contextual factors, including leadership support at the specialty level, champion resources and capacity, and EHR capabilities. CONCLUSIONS: Health systems need multifaceted implementation strategies to move beyond the limited impact of clinician training in driving improvement in serious illness conversations. These include EHR-based data feedback, involvement of specialty leaders to message the programme and align incentives, and local champions to problem-solve frontline challenges longitudinally. Implementation of these strategies depended on a favourable institutional context. Greater attention to the influence of contextual factors and implementation strategies may enable sustained improvements in serious illness conversations at scale.

Clinician perspectives on machine learning prognostic algorithms in the routine care of patients with cancer: a qualitative study.

PURPOSE: Oncologists may overestimate prognosis for patients with cancer, leading to delayed or missed conversations about patients' goals and subsequent low-quality end-of-life care. Machine learning algorithms may accurately predict mortality risk in cancer, but it is unclear how oncology clinicians would use such algorithms in practice. METHODS: The purpose of this qualitative study was to assess oncology clinicians' perceptions on the utility and barriers of machine learning prognostic algorithms to prompt advance care planning. Participants included medical oncology physicians and advanced practice providers (APPs) practicing in tertiary and community practices within a large academic healthcare system. Transcripts were coded and analyzed inductively using NVivo software. RESULTS: The study included 29 oncology clinicians (19 physicians, 10 APPs) across 6 practice sites (1 tertiary, 5 community) in the USA. Fourteen participants had previously had exposure to an automated machine learning-based prognostic algorithm as part of a pragmatic randomized trial. Clinicians believed that there was utility for algorithms in validating their own intuition about prognosis and prompting conversations about patient goals and preferences. However, this enthusiasm was tempered by concerns about algorithm accuracy, over-reliance on algorithm predictions, and the ethical implications around disclosure of an algorithm prediction. There was significant variation in tolerance for false positive vs. false negative predictions. CONCLUSION: While oncologists believe there are applications for advanced prognostic algorithms in routine care of patients with cancer, they are concerned about algorithm accuracy, confirmation and automation biases, and ethical issues of prognostic disclosure.

Hospice Enrollment, Future Hospitalization, and Future Costs Among Racially and Ethnically Diverse Patients Who Received Palliative Care Consultation.

BACKGROUND: Palliative care consultation to discuss goals-of-care ("PCC") may mitigate end-of-life care disparities. OBJECTIVE: To compare hospitalization and cost outcomes by race and ethnicity among PCC patients; identify predictors of hospice discharge and post-discharge hospitalization utilization and costs. METHODS: This secondary analysis of a retrospective cohort study assessed hospice discharge, do-not-resuscitate status, 30-day readmissions, days hospitalized, ICU care, any hospitalization cost, and total costs for hospitalization with PCC and hospitalization(s) post-discharge among 1,306 Black/African American, Latinx, White, and Other race PCC patients at a United States academic hospital. RESULTS: In adjusted analyses, hospice enrollment was less likely with Medicaid (AOR = 0.59, P = 0.02). Thirty-day readmission was less likely among age 75+ (AOR = 0.43, P = 0.02); more likely with Medicaid (AOR = 2.02, P = 0.004), 30-day prior admission (AOR = 2.42, P < 0.0001), and Black/African American race (AOR = 1.57, P = 0.02). Future days hospitalized was greater with Medicaid (Coefficient = 4.49, P = 0.001), 30-day prior admission (Coefficient = 2.08, P = 0.02), and Black/African American race (Coefficient = 2.16, P = 0.01). Any future hospitalization cost was less likely among patients ages 65-74 and 75+ (AOR = 0.54, P = 0.02; AOR = 0.53, P = 0.02); more likely with Medicaid (AOR = 1.67, P = 0.01), 30-day prior admission (AOR = 1.81, P = 0.0001), and Black/African American race (AOR = 1.40, P = 0.02). Total future hospitalization costs were lower for females (Coefficient = -3616.64, P = 0.03); greater with Medicaid (Coefficient = 7388.43, P = 0.01), 30-day prior admission (Coefficient = 3868.07, P = 0.04), and Black/African American race (Coefficient = 3856.90, P = 0.04). Do-not-resuscitate documentation (48%) differed by race. CONCLUSIONS: Among PCC patients, Black/African American race and social determinants of health were risk factors for future hospitalization utilization and costs. Medicaid use predicted hospice discharge. Social support interventions are needed to reduce future hospitalization disparities.

Factors Associated with Health Care Utilization at the End of Life for Patients with Acute Myeloid Leukemia.

Background: Patients (≥60 years) with acute myeloid leukemia (AML) often receive intense health care utilization at the end of life (EOL). However, factors associated with their health care use at the EOL are unknown. Methods: We conducted a secondary analysis of 168 deceased patients with AML within the United States. We assessed quality of life (QOL) (Functional-Assessment-Cancer-Therapy-Leukemia), and psychological distress (Hospital-Anxiety-and-Depression Scale [HADS]; Patient-Health-Questionnaire-9 [PHQ-9]) at diagnosis. We used multivariable logistic regression models to examine the association between patient-reported factors and the following outcomes: (1) hospitalizations in the last 7 days of life, (2) receipt of chemotherapy in the last 30 days of life, and (3) hospice utilization. Results: About 66.7% (110/165) were hospitalized in the last 7 days of life, 51.8% (71/137) received chemotherapy in the last 30 days of life, and 40.7% (70/168) utilized hospice. In multivariable models, higher education (odds ratio [OR] = 1.54, p = 0.006) and elevated baseline depression symptoms (PHQ-9: OR = 1.09, p = 0.028) were associated with higher odds of hospitalization in the last seven days of life, while higher baseline QOL (OR = 0.98, p = 0.009) was associated with lower odds of hospitalization at the EOL. Higher baseline depression symptoms were associated with receipt of chemotherapy at the EOL (HADS-Depression: OR = 1.10, p = 0.042). Higher education was associated with lower hospice utilization (OR = 0.356, p = 0.024). Conclusions: Patients with AML who are more educated, with higher baseline depression symptoms and lower QOL, were more likely to experience high health care utilization at the EOL. These populations may benefit from interventions to optimize the quality of their EOL care.

Serious Illness Communication Practices in Glioblastoma: An Institutional Perspective.

Background: Early, high-quality advance care planning discussions are essential for supporting goal-concordant care among glioblastoma (GBM) patients. Objective: Using mixed methods, we sought to characterize current serious illness (SI) communication practices at our institution. Methods: The electronic medical records of 240 deceased GBM patients cared for at the Abramson Cancer Center in Philadelphia, PA between 2017 and 2019 were systematically reviewed for documented SI conversations about four domains: prognosis, goals, end-of-life planning, and code status. Patient outcomes and SI conversation characteristics were analyzed using descriptive statistics. Standardized interviews about GBM care were held with five clinicians. Interview transcripts were analyzed using grounded-theory coding to identify emergent themes. Results: Nearly all patients (96%) had at least one documented SI conversation (median: 4, interquartile range [IQR] 2-7), mostly outpatient with medical oncology physicians. Median timing of first SI conversation was 360 days before death. SI conversations were not significantly associated with patient outcomes, including inpatient death and hospice enrollment. Seven themes emerged from clinician interviews: balancing hope and reality, anticipatory guidance, neglect of the "big picture," need for earlier conversations, care coordination, the role of clinical expertise, and communication training. Conclusion: SI conversations were documented early and often in our sample, but their quality was difficult to assess. Contrary to our quantitative findings, interviewees reported that SI conversations were late, infrequent, inadequate, and fragmented across specialties, failing to explore critical issues such as prognosis and functional decline.

Impact of Behavioral Nudges on the Quality of Serious Illness Conversations Among Patients With Cancer: Secondary Analysis of a Randomized Controlled Trial.

PURPOSE: Serious Illness Conversations (SICs) are structured conversations between clinicians and patients about prognosis, treatment goals, and end-of-life preferences. Although behavioral interventions may prompt earlier or more frequent SICs, their impact on the quality of SICs is unclear. METHODS: This was a secondary analysis of a randomized clinical trial (NCT03984773) among 78 clinicians and 14,607 patients with cancer testing the impact of an automated mortality prediction with behavioral nudges to clinicians to prompt more SICs. We analyzed 318 randomly selected SICs matched 1:1 by clinicians (159 control and 159 intervention) to compare the quality of intervention vs. control conversations using a validated codebook. Comprehensiveness of SIC documentation was used as a measure of quality, with higher integer numbers of documented conversation domains corresponding to higher quality conversations. A conversation was classified as high-quality if its score was ≥ 8 of a maximum of 10. Using a noninferiority design, mixed effects regression models with clinician-level random effects were used to assess SIC quality in intervention vs. control groups, concluding noninferiority if the adjusted odds ratio (aOR) was not significantly < 0.9. RESULTS: Baseline characteristics of the control and intervention groups were similar. Intervention SICs were noninferior to control conversations (aOR 0.99; 95% CI, 0.91 to 1.09). The intervention increased the likelihood of addressing patient-clinician relationship (aOR = 1.99; 95% CI, 1.23 to 3.27; P < .01) and decreased the likelihood of addressing family involvement (aOR = 0.56; 95% CI, 0.34 to 0.90; P < .05). CONCLUSION: A behavioral intervention that increased SIC frequency did not decrease their quality. Behavioral prompts may increase SIC frequency without sacrificing quality.

Behavioral economic implementation strategies to improve serious illness communication between clinicians and high-risk patients with cancer: protocol for a cluster randomized pragmatic trial.

BACKGROUND: Serious illness conversations (SICs) are an evidence-based approach to eliciting patients' values, goals, and care preferences that improve patient outcomes. However, most patients with cancer die without a documented SIC. Clinician-directed implementation strategies informed by behavioral economics ("nudges") that identify high-risk patients have shown promise in increasing SIC documentation among clinicians. It is unknown whether patient-directed nudges that normalize and prime patients towards SIC completion-either alone or in combination with clinician nudges that additionally compare performance relative to peers-may improve on this approach. Our objective is to test the effect of clinician- and patient-directed nudges as implementation strategies for increasing SIC completion among patients with cancer. METHODS: We will conduct a 2 × 2 factorial, cluster randomized pragmatic trial to test the effect of nudges to clinicians, patients, or both, compared to usual care, on SIC completion. Participants will include 166 medical and gynecologic oncology clinicians practicing at ten sites within a large academic health system and their approximately 5500 patients at high risk of predicted 6-month mortality based on a validated machine-learning prognostic algorithm. Data will be obtained via the electronic medical record, clinician survey, and semi-structured interviews with clinicians and patients. The primary outcome will be time to SIC documentation among high-risk patients. Secondary outcomes will include time to SIC documentation among all patients (assessing spillover effects), palliative care referral among high-risk patients, and aggressive end-of-life care utilization (composite of chemotherapy within 14 days before death, hospitalization within 30 days before death, or admission to hospice within 3 days before death) among high-risk decedents. We will assess moderators of the effect of implementation strategies and conduct semi-structured interviews with a subset of clinicians and patients to assess contextual factors that shape the effectiveness of nudges with an eye towards health equity. DISCUSSION: This will be the first pragmatic trial to evaluate clinician- and patient-directed nudges to promote SIC completion for patients with cancer. We expect the study to yield insights into the effectiveness of clinician and patient nudges as implementation strategies to improve SIC rates, and to uncover multilevel contextual factors that drive response to these strategies. TRIAL REGISTRATION: ClinicalTrials.gov , NCT04867850 . Registered on April 30, 2021. FUNDING: National Cancer Institute P50CA244690.

Posttraumatic stress disorder symptoms in patients with acute myeloid leukemia.

BACKGROUND: Patients with acute myeloid leukemia (AML) receiving intensive chemotherapy face a life-threatening illness, isolating hospitalization, and substantial physical and psychological symptoms. However, data are limited regarding risk factors of posttraumatic stress disorder (PTSD) symptoms in this population. METHODS: The authors conducted a secondary analysis of data from 160 patients with high-risk AML who were enrolled in a supportive care trial. The PTSD Checklist-Civilian Version was used to assess PTSD symptoms at 1 month after AML diagnosis. The Brief COPE and the Functional Assessment of Cancer Therapy-Leukemia were to assess coping and quality of life (QOL), respectively. In addition, multivariate regression models were constructed to assess the relation between PTSD symptoms and baseline sociodemographic factors, coping, and QOL. RESULTS: Twenty-eight percent of patients reported PTSD symptoms, describing high rates of intrusion, avoidance, and hypervigiliance. Baseline sociodemographic factors significantly associated with PTSD symptoms were age (B = -0.26; P = .002), race (B = -8.78; P = .004), and postgraduate education (B = -6.30; P = .029). Higher baseline QOL (B = -0.37; P ≤ .001) and less decline in QOL during hospitalization (B = -0.05; P = .224) were associated with fewer PTSD symptoms. Approach-oriented coping (B = -0.92; P = .001) was associated with fewer PTSD symptoms, whereas avoidant coping (B = 2.42; P ≤ .001) was associated with higher PTSD symptoms. CONCLUSIONS: A substantial proportion of patients with AML report clinically significant PTSD symptoms 1 month after initiating intensive chemotherapy. Patients' baseline QOL, coping strategies, and extent of QOL decline during hospitalization emerge as important risk factors for PTSD, underscoring the need for supportive oncology interventions to reduce the risk of PTSD in this population.

Integrating Conservative kidney management Options and advance care Planning Education (COPE) into routine CKD care: a protocol for a pilot randomised controlled trial.

INTRODUCTION: Predialysis education for patients with advanced chronic kidney disease (CKD) typically focuses narrowly on haemodialysis and peritoneal dialysis as future treatment options. However, patients who are older or seriously ill may not want to pursue dialysis and/or may not benefit from this treatment. Conservative kidney management, a reasonable alternative treatment, and advance care planning (ACP) are often left out of patient education and shared decision-making. In this study, we will pilot an educational intervention (Conservative Kidney Management Options and Advance Care Planning Education-COPE) to improve knowledge of conservative kidney management and ACP among patients with advanced CKD who are older and/or have poor functional status. METHODS AND ANALYSIS: This is a single-centre pilot randomised controlled trial at an academic centre in Philadelphia, PA. Eligible patients will have: age ≥70 years and/or poor functional status (as defined by Karnofsky Performance Index Score <70), advanced CKD (estimated glomerular filtration rate<20 mL/min/1.73 m2), prefer to speak English during clinical encounters and self-report as black or white race. Enrolled patients will be randomised 1:1, with stratification by race, to receive enhanced usual care or usual care and in-person education about conservative kidney management and ACP (COPE). The primary outcome is change in knowledge of CKM and ACP. We will also explore intervention feasibility and acceptability, change in communication of preferences and differences in the intervention's effects on knowledge and communication of preferences by race. We will assess outcomes at baseline, immediately post-education and at 2 and 12 weeks. ETHICS AND DISSEMINATION: This protocol has been approved by the Institutional Review Board at the University of Pennsylvania. We will obtain written informed consent from all participants. The results from this work will be presented at academic conferences and disseminated through peer-reviewed journals. TRIAL REGISTRATION NUMBER: This trial is registered at ClinicalTrials.gov under NCT03229811.

Effectiveness of Integrated Palliative and Oncology Care for Patients With Acute Myeloid Leukemia: A Randomized Clinical Trial.

IMPORTANCE: Patients with acute myeloid leukemia (AML) receiving intensive chemotherapy experience substantial decline in their quality of life (QOL) and mood during their hospitalization for induction chemotherapy and often receive aggressive care at the end of life (EOL). However, the role of specialty palliative care for improving the QOL and care for this population is currently unknown. OBJECTIVE: To assess the effect of integrated palliative and oncology care (IPC) on patient-reported and EOL outcomes in patients with AML. DESIGN, SETTING, AND PARTICIPANTS: We conducted a multisite randomized clinical trial of IPC (n = 86) vs usual care (UC) (n = 74) for patients with AML undergoing intensive chemotherapy. Data were collected from January 2017 through July 2019 at 4 tertiary care academic hospitals in the United States. INTERVENTIONS: Patients assigned to IPC were seen by palliative care clinicians at least twice per week during their initial and subsequent hospitalizations. MAIN OUTCOMES AND MEASURES: Patients completed the 44-item Functional Assessment of Cancer Therapy-Leukemia scale (score range, 0-176) to assess QOL; the 14-item Hospital Anxiety and Depression Scale (HADS), with subscales assessing symptoms of anxiety and depression (score range, 0-21); and the PTSD Checklist-Civilian version to assess posttraumatic stress disorder (PTSD) symptoms (score range, 17-85) at baseline and weeks 2, 4, 12, and 24. The primary end point was QOL at week 2. We used analysis of covariance adjusting and mixed linear effect models to evaluate patient-reported outcomes. We used Fisher exact test to compare patient-reported discussion of EOL care preferences and receipt of chemotherapy in the last 30 days of life. RESULTS: Of 235 eligible patients, 160 (68.1%) were enrolled; of the 160 participants, the median (range) age was 64.4 (19.7-80.1) years, and 64 (40.0%) were women. Compared with those receiving UC, IPC participants reported better QOL (adjusted mean score, 107.59 vs 116.45; P = .04), and lower depression (adjusted mean score, 7.20 vs 5.68; P = .02), anxiety (adjusted mean score, 5.94 vs 4.53; P = .02), and PTSD symptoms (adjusted mean score, 31.69 vs 27.79; P = .01) at week 2. Intervention effects were sustained to week 24 for QOL (ß, 2.35; 95% CI, 0.02-4.68; P = .048), depression (ß, -0.42; 95% CI, -0.82 to -0.02; P = .04), anxiety (ß, -0.38; 95% CI, -0.75 to -0.01; P = .04), and PTSD symptoms (ß, -1.43; 95% CI, -2.34 to -0.54; P = .002). Among patients who died, those receiving IPC were more likely than those receiving UC to report discussing EOL care preferences (21 of 28 [75.0%] vs 12 of 30 [40.0%]; P = .01) and less likely to receive chemotherapy near EOL (15 of 43 [34.9%] vs 27 of 41 [65.9%]; P = .01). CONCLUSIONS AND RELEVANCE: In this randomized clinical trial of patients with AML, IPC led to substantial improvements in QOL, psychological distress, and EOL care. Palliative care should be considered a new standard of care for patients with AML. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT02975869.

Effect of Integrating Machine Learning Mortality Estimates With Behavioral Nudges to Clinicians on Serious Illness Conversations Among Patients With Cancer: A Stepped-Wedge Cluster Randomized Clinical Trial.

IMPORTANCE: Serious illness conversations (SICs) are structured conversations between clinicians and patients about prognosis, treatment goals, and end-of-life preferences. Interventions that increase the rate of SICs between oncology clinicians and patients may improve goal-concordant care and patient outcomes. OBJECTIVE: To determine the effect of a clinician-directed intervention integrating machine learning mortality predictions with behavioral nudges on motivating clinician-patient SICs. DESIGN, SETTING, AND PARTICIPANTS: This stepped-wedge cluster randomized clinical trial was conducted across 20 weeks (from June 17 to November 1, 2019) at 9 medical oncology clinics (8 subspecialty oncology and 1 general oncology clinics) within a large academic health system in Pennsylvania. Clinicians at the 2 smallest subspecialty clinics were grouped together, resulting in 8 clinic groups randomly assigned to the 4 intervention wedge periods. Included participants in the intention-to-treat analyses were 78 oncology clinicians who received SIC training and their patients (N = 14 607) who had an outpatient oncology encounter during the study period. INTERVENTIONS: (1) Weekly emails to oncology clinicians with SIC performance feedback and peer comparisons; (2) a list of up to 6 high-risk patients (≥10% predicted risk of 180-day mortality) scheduled for the next week, estimated using a validated machine learning algorithm; and (3) opt-out text message prompts to clinicians on the patient's appointment day to consider an SIC. Clinicians in the control group received usual care consisting of weekly emails with cumulative SIC performance. MAIN OUTCOMES AND MEASURES: Percentage of patient encounters with an SIC in the intervention group vs the usual care (control) group. RESULTS: The sample consisted of 78 clinicians and 14 607 patients. The mean (SD) age of patients was 61.9 (14.2) years, 53.7% were female, and 70.4% were White. For all encounters, SICs were conducted among 1.3% in the control group and 4.6% in the intervention group, a significant difference (adjusted difference in percentage points, 3.3; 95% CI, 2.3-4.5; P < .001). Among 4124 high-risk patient encounters, SICs were conducted among 3.6% in the control group and 15.2% in the intervention group, a significant difference (adjusted difference in percentage points, 11.6; 95% CI, 8.2-12.5; P < .001). CONCLUSIONS AND RELEVANCE: In this stepped-wedge cluster randomized clinical trial, an intervention that delivered machine learning mortality predictions with behavioral nudges to oncology clinicians significantly increased the rate of SICs among all patients and among patients with high mortality risk who were targeted by the intervention. Behavioral nudges combined with machine learning mortality predictions can positively influence clinician behavior and may be applied more broadly to improve care near the end of life. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT03984773.

Validation of a Machine Learning Algorithm to Predict 180-Day Mortality for Outpatients With Cancer.

IMPORTANCE: Machine learning (ML) algorithms can identify patients with cancer at risk of short-term mortality to inform treatment and advance care planning. However, no ML mortality risk prediction algorithm has been prospectively validated in oncology or compared with routinely used prognostic indices. OBJECTIVE: To validate an electronic health record-embedded ML algorithm that generated real-time predictions of 180-day mortality risk in a general oncology cohort. DESIGN, SETTING, AND PARTICIPANTS: This prognostic study comprised a prospective cohort of patients with outpatient oncology encounters between March 1, 2019, and April 30, 2019. An ML algorithm, trained on retrospective data from a subset of practices, predicted 180-day mortality risk between 4 and 8 days before a patient's encounter. Patient encounters took place in 18 medical or gynecologic oncology practices, including 1 tertiary practice and 17 general oncology practices, within a large US academic health care system. Patients aged 18 years or older with outpatient oncology or hematology and oncology encounters were included in the analysis. Patients were excluded if their appointment was scheduled after weekly predictions were generated and if they were only evaluated in benign hematology, palliative care, or rehabilitation practices. EXPOSURES: Gradient-boosting ML binary classifier. MAIN OUTCOMES AND MEASURES: The primary outcome was the patients' 180-day mortality from the index encounter. The primary performance metric was the area under the receiver operating characteristic curve (AUC). RESULTS: Among 24 582 patients, 1022 (4.2%) died within 180 days of their index encounter. Their median (interquartile range) age was 64.6 (53.6-73.2) years, 15 319 (62.3%) were women, 18 015 (76.0%) were White, and 10 658 (43.4%) were seen in the tertiary practice. The AUC was 0.89 (95% CI, 0.88-0.90) for the full cohort. The AUC varied across disease-specific groups within the tertiary practice (AUC ranging from 0.74 to 0.96) but was similar between the tertiary and general oncology practices. At a prespecified 40% mortality risk threshold used to differentiate high- vs low-risk patients, observed 180-day mortality was 45.2% (95% CI, 41.3%-49.1%) in the high-risk group vs 3.1% (95% CI, 2.9%-3.3%) in the low-risk group. Integrating the algorithm into the Eastern Cooperative Oncology Group and Elixhauser comorbidity index-based classifiers resulted in favorable reclassification (net reclassification index, 0.09 [95% CI, 0.04-0.14] and 0.23 [95% CI, 0.20-0.27], respectively). CONCLUSIONS AND RELEVANCE: In this prognostic study, an ML algorithm was feasibly integrated into the electronic health record to generate real-time, accurate predictions of short-term mortality for patients with cancer and outperformed routinely used prognostic indices. This algorithm may be used to inform behavioral interventions and prompt earlier conversations about goals of care and end-of-life preferences among patients with cancer.

Serious Illness Conversations With Outpatient Oncology Clinicians: Understanding the Patient Experience.

PURPOSE: Guidelines recommend earlier advance care planning discussions focused on goals and values (serious illness communication) among oncology patients. We conducted a prospective, cross-sectional quality improvement evaluation of patients who had a serious illness conversation (SIC) with an oncology clinician using the Serious Illness Conversation Guide to understand patient perceptions of conversations using a structured guide. METHODS: We contacted 66 oncology patients with an SIC documented in the electronic health record. Thirty-two patients (48%) responded to survey and/or structured interview questions by telephone. We used summary statistics and thematic analysis to analyze results. RESULTS: Twenty-eight respondents (90%) reported that the SIC was worthwhile. Seventeen respondents (55%) reported that the conversation increased their understanding of their future health, and 18 (58%) reported that the conversation increased their sense of closeness with their clinician. Although the majority of respondents (28 [90%]) reported that the conversation increased (13 [42%]) or had no effect (15 [48%]) on their hopefulness, a small minority (3 [10%]) reported a decrease in hopefulness. Qualitative analysis revealed 6 themes: clinician-patient relationship, impact on well-being, memorable characteristics of the conversation, improved prognostic understanding, practical planning, and family communication. CONCLUSION: SICs are generally acceptable to oncology patients (nonharmful to the vast majority, positive for many). Our qualitative analysis suggests a positive impact on prognostic understanding and end-of-life planning, but opportunities for improvement in the delivery of prognosis and preparing patients for SICs. Our data also identify a small cohort who responded negatively, highlighting an important area for future study.

Patient Risk Factor Profiles Associated With the Timing of Goals-of-Care Consultation Before Death: A Classification and Regression Tree Analysis.

BACKGROUND: Early palliative care consultation ("PCC") to discuss goals-of-care benefits seriously ill patients. Risk factor profiles associated with the timing of conversations in hospitals, where late conversations most likely occur, are needed. OBJECTIVE: To identify risk factor patient profiles associated with PCC timing before death. METHODS: Secondary analysis of an observational study was conducted at an urban, academic medical center. Patients aged 18 years and older admitted to the medical center, who had PCC, and died July 1, 2014 to October 31, 2016, were included. Patients admitted for childbirth or rehabilitationand patients whose date of death was unknown were excluded. Classification and Regression Tree modeling was employed using demographic and clinical variables. RESULTS: Of 1141 patients, 54% had PCC "close to death" (0-14 days before death); 26% had PCC 15 to 60 days before death; 21% had PCC >60 days before death (median 13 days before death). Variables associated with receiving PCC close to death included being Hispanic or "Other" race/ethnicity intensive care patients with extreme illness severity (85%), with age <46 or >75 increasing this probability (98%). Intensive care patients with extreme illness severity were also likely to receive PCC close to death (64%) as were 50% of intensive care patients with less than extreme illness severity. CONCLUSIONS: A majority of patients received PCC close to death. A complex set of variable interactions were associated with PCC timing. A systematic process for engaging patients with PCC earlier in the care continuum, and in intensive care regardless of illness severity, is needed.