Primary palliative care for heart failure provided within ambulatory cardiology: A randomized pilot trial

Abstract

Background:

Heart failure is characterized by physical and emotional symptoms and decreased quality of life (QoL). Palliative care can reduce burdens of serious illness but often is limited to inpatient or academic settings.

Objectives:

To develop and test the Primary Education for Nurses in Palliative care-HF (PENPal-HF) intervention, training outpatient cardiology nurses to address symptom burden, patient priorities for care and QoL, and advance care planning as part of quarterly HF visits.

Methods:

We conducted a pilot randomized clinical trial for adults with NYHA Stage III or IV HF and ≥ 2 hospitalizations in the past 12 months, recruited from a community-based cardiology clinic. Participants were randomized 2:1, PENPal-HF plus usual care versus usual care alone. Primary outcomes were feasibility and acceptability.

Results:

We randomized 30 adults with Stage III HF – 20 to PENPal-HF and 10 to usual care. Most in the intervention group (71%) and in the control group (62%) completed the study through the final outcome assessment in week 56; 5 participants died. Of 20 participants in the intervention, 14 (70%) remained in the study through the end of intervention visits; 11 (55%) completed all visits. Most intervention participants (93.75%) agreed or strongly agreed that they were satisfied with their care, and 87.5% agreed or strongly agreed that all people with HF should receive the intervention. Most intervention group participants (93.75%) reported a perceived improvement in physical symptoms, mood, and/or QoL.

Conclusions:

This pilot study suggests that nurse-led primary palliative care in outpatient cardiology settings is promising. Research is warranted to determine efficacy and effectiveness.

Introduction

More than 6 million adults in the United States and their caregivers suffer from the multifactorial burdens of heart failure (HF), characterized by progressive physical and emotional symptoms, loss of independence, and reduced quality of life (QoL). ^1–3 Evidence from randomized clinical trials suggests that specialist palliative care interventions are effective in reducing the burdens common in serious illnesses, although most of this evidence emanates from oncology.^4–6 Emerging evidence suggests that specialist palliative care interventions may hold similar promise in HF;^{4, 5, 7} yet gaps persist regarding delivery models of palliative care in HF that are generalizable to meet population-level demands.

The limited specialist palliative care workforce cannot match the volume of patients who could benefit from palliative intervention, particularly outside of tertiary or quaternary care centers.^{8, 9} Primary, or generalist, palliative care is an alternative approach in which clinicians focused in other specialties are trained in palliative principles such as basic symptom management and goals of care discussions.¹⁰ A 2016 systematic review documented 16 randomized clinical trials of primary palliative care interventions, yet heterogeneity among the interventions and their methodological quality leave room for improvement.^{4, 11} First, the interventions ranged widely in their delivery setting (e.g., home, inpatient), clinician type (e.g., physician, nurse, social worker), and focus (e.g., symptom management, advance care planning).¹¹ Second, of the 16 randomized clinical trials included, 88% (n=14) were deemed to be of high or unclear risk of bias. Third, none of the included trials of primary palliative care were designed for or conducted with patients with HF. Formative data by Kavalieratos and colleagues found strong support among cardiologists and cardiology nurses for primary palliative care delivered by cardiology nurses with additional training in palliative principles.¹²

To address the gap between the need for palliative care services among patients with HF and the challenge of providing palliative care in community settings, we developed and pilot tested PENPal-HF (Primary Education for Nurses in Palliative care-HF), training registered nurses already practicing in outpatient cardiology clinics to elicit and address symptom burden, patient priorities for care and QoL, and advance care planning, using enhanced communication skills.¹³ We then used the RE-AIM (reach, effectiveness, adoption, implementation, maintenance) framework to conduct a summative evaluation of this pilot trial to identify potential implementation barriers.^14–16 We used a pilot randomized clinical trial to assess the feasibility and acceptability of participant recruiting, intervention delivery and fidelity, and data collection before undertaking an adequately-powered efficacy trial.

Methods

Intervention Development

We developed the intervention by triangulating data and input from a multidisciplinary stakeholder advisory board, qualitative research with cardiology clinicians and patients with heart failure,^{12, 17} and literature reviews,^{4, 6} using the Medical Research Council framework for complex intervention research, which emphasizes investigating not only whether an intervention achieves a set outcome but how it fits into the real-world context of resources, delivery environment, and decision making.¹⁸ The resulting intervention, PENPal-HF (NCT03170466), is informed by the Chronic Care Model, which aims to improve patient outcomes by restructuring ambulatory chronic illness care to promote patient activation and clinician support.^{19, 20} In PENPal-HF, we provide training and clinical decision support to existing cardiology clinic nurse care managers to enhance their skillset regarding palliative needs, such as symptom management, coping with serious illness, and advance care planning. Further, the operational workflows of outpatient cardiology clinics are modified to allow the additional time needed to deliver primary palliative care content.

PENPal-HF comprised three components: 1) Patients received four quarterly visits with the PENPal-HF nurse as part of regular outpatient cardiology visits; 2) the PENPal-HF nurse made monthly semi-scripted calls between visits to reinforce palliative principles covered in the visits and to identify any incident concerns; and 3) the PENPal-HF nurse communicated through the electronic medical record or interpersonal interactions about patient concerns with the patient’s cardiologist and primary care clinician and made outside referrals to additional resources as needed. The four intervention visits addressed physical and emotional symptom management while introducing and exploring palliative concepts such as prognostic awareness, goals of care, and advance care planning.²¹ Caregivers were invited to attend study visits and receive care suggestions from the nurse; however, caregiver enrollment was not mandatory. To ease uptake and fidelity, we distilled the intervention content to five “critical questions,” which were used to elicit patient priorities for palliative care and preferences for future care (Table 1).

Table 1.

Intervention Visit Topics and Critical Questions

Visit	Topics	Critical Questions
1	• Buildrapport • Demystify palliative care • Set visit agenda • Focused physical and emotional symptom management • Explore prognostic awareness • Introduce advance care planning	• What is the most stressful/difficult part about living with heart failure? • If we could change one thing about your health, what would it be? • Who do you trust to make decisions for you if you are no longer able to speak to your doctors? What have you told this person about what you would want if you were no longer able to speak for yourself? • As you think about your health and the future, what is most important to you?
2–4	• Set visit agenda • Focused physical and emotional symptom management • Explore views about illness and coping • Goals of care and advance care planning	• As you think about your own health and the future, what kinds of situations do you want to avoid? For example, some people say they do not want to have to depend on a machine to breathe; others say they wouldn’t want to have CPR.

PENPal-HF Nurse Training

All intervention visits were conducted by a cardiology registered nurse who held an Associate Degree in Nursing and had 15 years of nursing experience. Because we designed a pilot trial of feasibility and acceptability at a single site, we conducted the intervention with a single nurse due to limited financial and personnel resources available at the time. The nurse received training in primary palliative care skills relevant to advanced HF (e.g., diet management, discussing advance care planning in light of cardiac devices). Training included review of the 79-page study manual, which covered the structure of each visit as well as symptom management guides. The manual also included sample language for introducing palliative care principles into patient visits and tips on verbal and nonverbal cues to elicit patient reflection on sensitive topics. Prior to study initiation, the nurse spent 10.5 hours training with palliative care educators on principles of serious illness communication, role play with standardized patient actors, and skills booster sessions. The nurse completed checklists of topics addressed after each intervention visit as part of fidelity monitoring. In addition, we audio recorded all intervention visits and reviewed them using a fidelity rubric based on the intervention manual. We prepared supplemental trainings if intervention fidelity dipped below 80% over the course of the study.

Trial Design

We used a randomized pilot clinical trial design with both quantitative and qualitative methods to evaluate the feasibility, acceptability, and perceived efficacy of the PENPal-HF intervention. Because the goal of the pilot was to assess these outcomes to inform future, powered trials, we did not compare outcomes between intervention and control groups; the purpose of control and randomization was to assess feasibility and acceptability of trial conditions. Patients were randomly assigned in a 2:1 ratio (20 PENPal-HF, 10 usual care) to prioritize generating information regarding the intervention. Participants in the usual care arm attended clinic appointments with cardiology nurses and cardiologists approximately every quarter, according to clinic norms. Participants in both arms could access specialty palliative care if referred by their cardiologist. The study protocol was approved by the University of Pittsburgh Institutional Review Board, and all participants provided informed consent prior to enrollment.

Population

We recruited adults with HF from a single academic-affiliated community cardiology office. Potential participants were eligible if they had NYHA Class III or IV HF and two or more HF-related hospitalizations in the past year; or excluded if they were <50 years old, expected to receive a cardiac transplant or implantable device in the next 6 months, or had received outpatient specialist palliative care in the preceding year. Other exclusion criteria were lack of regular phone access, lack of ability to understand English, not regularly attending clinic in the prior year, or screening results indicating possible cognitive impairment.²² Potential participants were identified via chart review and approached by the HF nurse about the study during a regular clinic visit. A research coordinator screened for possible cognitive impairment,²² obtained informed consent, and enrolled interested, eligible patients. Par Participants randomized to receive the intervention had their first study visit on the day of randomization.

Outcomes and Data Collection

Given the pilot nature of this trial, our primary outcome was feasibility of implementing and delivering the intervention to plan a larger and adequately powered efficacy trial. Thus, we chose N=30 as an enrollment target appropriate to the study timeframe and resources. For pilot outcomes we focused on: 1) our ability to enroll and randomize 30 participants; 2) participant retention rates; 3) intervention visit completion rates, including length and modality (in-person vs telephone); 4) intervention content fidelity (defined as the percentage of intervention visits that adhered to ≥80% of predefined intervention topics based on review of audio with rubric); and 5) completion rates of patient-reported outcomes surveys. Other outcomes were intervention acceptability and perceived efficacy of the intervention, both evaluated through closed- and open-ended surveys and qualitative interviews of patient participants and clinic staff. Participants completed the Functional Assessment of Chronic Illness Therapy-Palliative Care, Kansas City Cardiomyopathy Questionnaire, FAMCARE-Patient, and Patient-Reported Outcomes Measurement Information System depression, anxiety, symptom management, medication management scales to report various patient-centered outcomes relevant to advanced HF; these measures were collected to evaluate the feasibility of data collection for a future Phase III trial. Participants completed baseline assessments via digital tablets while in clinic, and follow-up assessments every eight weeks for 56 weeks total via telephone. Participants received $10 for each of the seven surveys they completed, plus a $30 bonus for completing all possible surveys; payments amounts were determined based on the estimated time and inconvenience of completing multiple surveys over time.²³

Data Analysis

Because PENPal-HF involved both new settings and new delivery models for palliative care, we adapted the RE-AIM framework^{14, 16} constructs to assess the pilot intervention for future powered trials and possible eventual implementation. RE-AIM is a five-part implementation science framework that is widely used to evaluate barriers and facilitators to public health and clinical intervention impact and implementation. We slightly modified several RE-AIM constructs to reflect the pilot nature of this intervention; for example, because we did not test for differences in outcomes between trial arms, the “efficacy” domain was reconceptualized to evaluate patient participants’ perceptions of potential intervention benefit (Table 2).

Table 2.

Evaluation Metrics and Outcomes as Informed by RE-AIM Framework

RE-AIM Constructs as applied in this evaluation	Data source	Evaluation metrics	Associated outcome
Reach Number and proportion of individuals willing to participate	Screening and enrollment data	Target enrollment reached (n=30); enrollment rate; consent:approach ratio; approach: randomize ratio; % withdrawn or lost to follow-up	Feasibility
Efficacy (Perceived) Impact of intervention on important outcomes, including QOL	Interviews, surveys	Qualitative analysis of participant perception of intervention on living with HF; patient satisfaction surveys	Acceptability (patient)
Adoption Willingness ofsettings and staffto participate	Interviews	Clinician perceptions of barriers and facilitators to implementation in study setting	Acceptability (clinician)
Implementation Fidelity to protocol by PENPAL-HF nurse	Fidelity rubric	% of intervention visits addressing ≥80% of intervention topics	Fidelity
Maintenance Factors participants and staff perceive as related to likely success ofimplementation	Interviews	Clinician perception of whether and how intervention could be integrated into standard care for patients with HF	Not addressed

We conducted semi-structured interviews with a cardiologist, the PENPal-HF nurse, and four (20%) intervention arm participants about their experiences with PENPal-HF, barriers and facilitators to study participation, and barriers and facilitators to adding primary PEN-Pal to standard care for HF. We used template analysis,²⁴ a modified form of thematic analysis, to identify common themes deductively and inductively for efficacy and implementation. An experienced qualitative researcher (J.L.) coded all transcripts iteratively using the constant comparative method, consulting with the first author (D.K.) regarding codes and themes.

Quantitative data were analyzed using descriptive statistics by a study team member (Z.P.H.). Continuous variables were summarized using means and standard deviations, while categorical outcomes were described using proportions. In line with best practices for feasibility trials, we did not test for differences in outcomes across trial arms.²⁵

Results

Feasibility

We screened 5154 patient records from a single community-based cardiology clinic between October 2017-May 2019 (Figure 1). We consented 30 participants (52%) over 78 weeks of recruiting, with a mean age of 73 years (SD 11), of whom 83% were white (Table 3). All participants had NYHA Stage III HF at baseline and reported a mean 2.9 (SD 1.3) HF-related hospitalizations in the prior year. They also endorsed a median of 3 symptoms at moderate or higher burden, most commonly pain, shortness of breath, lack of energy, and poor mobility.

Figure 1.

Participant Flow. Fractions in assessment completion indicate number of assessments completed (numerator) vs number of participants remaining in study at each assessment timepoint (denominator)

Table 3.

Baseline Characteristics of Enrolled Patient Participants

	All Participants (N=30)	Intervention (N=20)	Usual care (N=10)
Age, mean (SD), years	73 (11)	70 (11)	81 (5)
Female sex	18 (60%)	11 (55%)	7 (70%)
Race
White	25 (83%)	16 (80%)	9 (90%)
Black	3 (10%)	3 (15%)	0 (0%)
Other	1 (3%)	1 (5%)	0 (0%)
Not reported	1 (3%)	0 (0%)	1 (10%)
Hispanic ethnicity	0 (0%)	0 (0%)	0 (0%)
Relationship status
Single/Never married	2 (7%)	1 (5%)	1 (10%)
Married or partnered	11 (37%)	10 (50%)	1 (10%)
Widowed	13 (43%)	5 (25%)	8 (80%)
Divorced or separated	4 (13%)	4 (20%)	0 (0%)
Education
Some high school or less	3 (10%)	3 (15%)	0 (0%)
High school or GED	18 (60%)	13 (65%)	5 (50%)
Some college or vocational school	5 (17%)	2 (10%)	3 (30%)
College degree	4 (13%)	2 (10%)	2 (20)
Postgraduate degree	0 (0.0%)	0 (0.0%)	0 (0%)
Work Status
Unemployed	0 (0%)	0 (0%)	0 (0%)
Employed	0 (0%)	0 (0%)	0 (0%)
Full time homemaker	6 (20%)	3 (15%)	3 (30%)
Not attending school/work due to health	4 (13%)	2 (10%)	2 (20%)
Not working for other reasons (e.g., retired)	18 (60%)	14 (70%)	4 (40%)
Not reported	2 (7%)	1 (5%)	1 (1%)
Finances at end of month
Money left over	11 (37%)	10 (50%)	1 (10%)
Just enough money	13 (43%)	6 (30%)	7 (70%)
Not enough money	6 (20%)	4 (20%)	2 (20%)
Health Status
NYHA Class III	30 (100%)	20 (100%)	10 (100%)
HF-related hospitalizations in prior 12 months, no., mean (SD)	3 (1)	3 (1)	3 (1)
Patient-reported outcomes, mean (SD)
FACIT-Pal total scorea	127 (29)	129 (29)	125 (29)
KCCQclinical summaryb	59 (22)	59 (22)	59 (22)
FAMCARE P13c	56 (6)	56 (6)	55 (7)
PROMIS Instruments, t-score (SD)
Anxietyd	52 (11)	52 (12)	55 (12)
Depressione	49 (11)	48 (10)	51 (11)
Self-efficacy to manage symptomsf	49 (10)	48 (10)	46 (6)
Self-efficacy to manage medications/treatmentsg	49 (8)	48 (7)	42 (8)
Advance Care Planning
Had living will or advance directive	21 (72%)*	12 (60%)	9 (100%)*
Had discussions regarding wishes with doctor	16 (55%)	9 (45%)	7 (78%)

^*One participant did not report.

^aThe FACIT-Pal instrument’s total score ranges from 0 to 184, where higher scores indicate better perceived quality of life.

^bThe KCCQ instrument ranges from 0 to 100, where higher scores indicate better perceived HF-related health status.

^cThe FAMCARE P13 instrument ranges from 1 to 65, where higher scores indicate better satisfaction with care.

^dThe PROMIS Anxiety instrument has a mean (SD) of 50 (10), where higher scores indicate greater levels of anxiety.

^eThe PROMIS Self Efficacy for Managing Chronic Conditions - Symptoms instrument has a mean (SD) t-score of 50 (10), where higher scores indicate greater confidence in managing or controlling symptoms.

^fThe PROMIS Self Efficacy for Managing Chronic Conditions - Medications/Treatments instrument has a mean (SD) t-score of 50 (10), where higher scores indicate greater confidence in managing medications.

Abbreviations: NYHA, New York Heart Association; HF, heart failure; FACIT-Pal, Functional Assessment of Chronic Illness Therapy – Palliative; KCCQ, Kansas City Cardiomyopathy Questionnaire; FAMCARE P13, Family Satisfaction with End-of-Life Care, Patient scale; PROMIS, Patient-Reported Outcomes Measurement System.

Study completion

In the intervention group, 12 of 17 (71%) patients completed the study through the final outcome assessment in week 56; 3 of 20 (15%) patients in the intervention group died before study completion. In the control group, 62% (5 of 8) patients completed the final outcome assessment and 2 of 10 (20%) died. Two patients in the intervention group (10%) were removed administratively (i.e., conflicting study enrollment, participant had a stroke), and one patient in the usual care arm (10%) became ineligible after transitioning to hospice. Two patients in each group (10%, 20% respectively) withdrew consent during the study, and 1 (5%) in the intervention group was lost to follow up. Completion of outcome assessments ranged from 70.6% to 100% in the intervention group to 66.7% to 100% in the control group, based on number of participants able and eligible to complete each of the bimonthly assessments.

Among the 20 patients randomized to the intervention group, 14 (70%) remained in the study through the end of intervention visits and 11 (79%) of those completed all visits. Due to the COVID-19 pandemic, four intervention visits were conducted via telephone without issue.

Intervention fidelity

The first intervention visit lasted a mean 46 minutes (SD 12.75); visits 2–4 took a mean 31.82 (SD 11.96) minutes to complete. Across all visits, mean fidelity, measured by comparing audio recordings of visits against a protocol checklist, was 80.6%.

Acceptability of intervention

Across the four intervention visits, 16 participants in the intervention group completed at least one survey evaluating satisfaction with and perceived efficacy of the intervention (Table 4). Among the 16 participants, 15 (93.75%) agreed or strongly agreed that they were satisfied with their care, and 14 (87.5%) agreed or strongly agreed that all patients with HF should receive the intervention. Most participants (15, or 93.75%) agreed or strongly agreed that (1) the intervention was not burdensome, (2) improved their understanding of the future with HF, and (3) improved their ability to cope with HF.

Table 4.

Satisfaction with and Perceived Efficacy of Intervention among Participants Randomized to PENPal-HF (Agree or strongly agree)

	Visit 1 (n=15)	Visit 2 (n=13)	Visit 3 (n=13)	Visit 4 (n=13)	Average (n=16)*
“I have been satisfied with the care that I received from the [study] nurse.”	15 (100%)	12 (92%)	12 (92%)	12 (92%)	15 (94%)
“I think that all patients with heart failure should receive the same care I did through the [study] nurse.”	14 (93%)	12 (92%)	12 (92%)	12 (92%)	14 (88%)
“Interacting with the [study] nurse in addition to my usual healthcare was burdensome.”	14 (93%)	11 (85%)	10 (77%)	12 (92%)	15 (94%)
“I believe that seeing the [study] nurse improved my physical symptoms.”	13 (86%)	12 (92%)	12 (92%)	10 (77%)	14 (88%)
“I believe that seeing the [study] nurse improved my mood.”	13 (86%)	12 (92%)	12 (92%)	12 (92%)	15 (94%)
“I believe that seeing the [study] nurse improved my quality of life.”	12 (80%)	12 (92%)	12 (92%)	12 (92%)	15 (94%)
“I believe that seeing the [study] nurse improved my understanding of my heart failure.”	14 (93%)	12 (92%)	12 (92%)	12 (92%)	15 (94%)
“I believe that seeing the [study] nurse improved my understanding ofwhat the future holds regarding my heart failure.”	12 (80%)	12 (92%)	11 (85%)	12 (92%)	15 (94%)
“I believe that seeing the [study] nurse improved my ability to cope with my heart failure.”	12 (80%)	12 (92%)	11 (85%)	12 (92%)	15 (94%)

^*16 participants completed at least 1 survey. Response categories were strongly disagree, disagree, neither agree nor disagree, agree, and strongly agree.

Most intervention group participants (n=15, 93.75%) reported a perceived improvement in physical symptoms, mood, and/or QoL over the course of the study.

Qualitative Evaluation

In analysis of qualitative interview data, we identified three themes. First, patients found having primary palliative care integrated into usual cardiology clinic visits beneficial. Patient participants reported that the PENPal-HF nurse had time to answer questions and was knowledgeable about their symptoms. Participants said the nurse respected their comfort level with serious topics like advance care planning: “It’s when she gets around to talking about if you die. . . your will, living will and stuff like that, and right now, I don’t want to hear about that in my condition [. . .] I wasn’t ready yet. [. . .] She helped me to get ready.” (Male, Age 74). Participants noted that they built rapport with the intervention nurse over the course of multiple study visits. “Even me, at first, ‘Ah, I don’t want to do this.’ Before it was over, I was missing it, when it was going to be over, if I make it that long. Who’s going to help without this program for me?” (Male, Age 68)

Second, co-location of palliative care within the HF clinic and on the same day as regular HF follow-up visits (as opposed to a standalone intervention) was an important facilitator for reducing patient burden, while also enhancing communication among clinical team members regarding patients’ palliative needs. A cardiologist also noted improvements in clinical communication as a result of providing palliative care within standard HF care. “Doing this the way it was done, you were really integrating it with the physician who really knows them well. If you start to separate it out, you may not have that same rapport between the two teams. ... All they have to do is walk to the back office and talk to the doctor or get on the phone quickly and call them.” (Cardiologist)

Third, strong buy-in and investment is necessary to successfully implement and deliver the intervention. Recognizing disruptions to clinic operations, the study financially offset a portion of the PENPal-HF nurse’s time to accommodate training, intervention delivery, and study-related tasks. “If it were to be rolled out in offices everywhere, it would require [planning] for staffing because that was a difficult issue in the beginning, too: I still had my own job to do, plus this, and I had to do it all in a day. (Cardiology nurse) Clinic staff identified other factors that could affect successful implementation of primary palliative care into a community HF clinic setting, including having a clinician champion, educating clinic staff about the role of palliative care, establishing clear communication channels between the nurse performing primary palliative care and other clinicians, and integrating palliative care visits into the clinic workflow. For the PENPal-HF nurse, access to additional resources, such as social work referrals, would help address patient needs outside the scope of clinical HF management.

Discussion

In this pilot randomized trial, we developed and tested an intervention to train registered nurses in community-based cardiology clinics in core palliative care skills relevant to adults with advanced heart failure: symptom assessment and management, goals of care elicitation, advance care planning, decision-making support, and care coordination. This intervention is innovative for three key reasons. First, we formally integrated primary palliative care within a community-based cardiology clinic, as opposed to larger and more well-resourced advanced HF clinics in tertiary or quaternary care centers, where most HF palliative care research has been conducted to date. Second, the intervention engages and enhances the skills of existing cardiology nurses to deliver primary palliative care, a model that has potential to greatly expand the avenues by which palliative approaches can reach broader populations. Third, the intervention minimizes burden for patients by integrating palliative care principles into routine outpatient cardiology clinic visits.

Participants in our intervention endorsed the benefits of palliative care delivered by their cardiology team – namely, the logistical convenience of working with clinicians with whom they already were familiar without additional office visits or expenses. The intervention nurse and clinic cardiologist likewise endorsed the efficiency of being able to communicate informally about cases within the clinic, underscoring a finding from formative – work that cardiology clinicians view themselves as providers of primary palliative care already yet seek additional skills to do so confidently and competently.^{12, 26} Participants also reported high satisfaction with the intervention and perceived improvement in symptoms and their understanding of HF, with 87.5% agreeing or strongly agreeing that others with HF receive the intervention. This pilot study thus provides evidence that leveraging primary clinicians and settings can extend core palliative care principles into communities and populations who might not otherwise have access to palliative care specialists.

This study has several noteworthy limitations. First, it was conducted within one general cardiology clinic with one nurse in a suburban, socioeconomically disadvantaged community, yet affiliated with a large integrated delivery and finance system. The latter offered resources and streamlined processes to conduct clinical research in community cardiology clinics that may not be generalizable at clinics not affiliated with a large network; nevertheless, many of the challenges encountered in this trial likely will be applicable to other contexts, including scheduling, disruptions in clinical operations (e.g., room availability), and financial support to offset lost revenues from nurse training time. Second, relative to other outpatient palliative care interventions for HF, this study had high mortality: 17% of randomized participants, compared with 9% in the ENABLE CHF-PC study²⁷ and 3% in the CASA trial,²⁸ respectively. Relatedly, while we report greater all-cause attrition (40% intervention, 50% control) than other pilot trials of primary palliative care interventions, it is important to note that our trial’s primary endpoint of 56 weeks is much longer than the ENABLE or CASA trials (i.e., 12 weeks). ENABLE similarly recruited patients with Stage III or IV HF but did not have a requirement of past-year hospitalization, while CASA recruited adults with HF who had one or more hospitalizations in the past year or one or more of several therapeutic and biomarker criteria. Our attrition rates at 8 and 16 weeks are comparable to those trials, which may have had less severely ill populations.

Other questions also must be resolved to optimize delivery of primary palliative care in a community subspecialty setting such as cardiology. First, when is the most beneficial time to introduce primary palliative care in seriously ill populations, and in what dose? One recent Phase III trial of primary palliative care, delivered over three months by oncology nurses to adults with advanced metastatic cancer and <1 year estimated survival showed no benefit over usual care, raising the question of whether the intervention was too dilute to meet the needs of patients with advanced disease and high palliative needs.²⁹ What timing and intensity of palliative care will increase the likelihood that patients will benefit both from symptom management support and advance care planning?

Second, to enhance feasibility of delivering the intervention during busy clinical encounters, we distilled core palliative principles into five “critical questions” (Table 1). Is primary palliative care best addressed through a tightly scripted intervention or by weaving the essence of it into usual care, and what tradeoffs does each pose for research fidelity and real-world implementation? Third, PENPal-HF prepared cardiology nurses to address symptom management, goals of care, decision making and other aspects of palliative care, but patients’ unmet needs may extend to social, financial, or logistical challenges that inhibit their ability to manage their HF. What training, resources, and linkages to community services would enhance PENPal-HF to alleviate non-medical burdens and other social determinants of health that harm QoL? Third, while PENPal-HF established a closed loop between the intervention nurse and clinic cardiologists, it lacked a formalized referral pathway for specialist palliative care, if needed. Yet, this reflects the reality that exceedingly few outpatient palliative care programs work routinely with patients with chronic, non-cancer illnesses.⁷ What are the mechanisms by which specialist palliative care can support patients and clinicians in community settings (for example, telemedicine)? Further, primary palliative care delivery was feasible under research conditions that paid for the PENPal-HF nurse’s time. Currently fee-for-service payment models limit the time and resources outpatient clinicians have available to address the components of primary palliative care. What payment models, such as global capitation, might financially support primary palliative care? Finally, our guiding premise was that it was important for primary palliative care to be delivered during usual cardiology clinic visits. Yet, like the proliferation of telemedicine during the COVID-19 pandemic, we successfully delivered a portion of PENPal-HF visits via telephone. Can PENPal-HF be delivered remotely or even asynchronously with outpatient HF visits?

Conclusion

Broadly, this study illustrates the challenge and the potential in evolving palliative care from tightly controlled academic research environments to the real world. It also represents an additional step toward understanding the role of palliative care in cardiology and more broadly, in chronic, non-cancer illness.⁶ The results of this pilot trial suggest that PENPal-HF is generally feasible, acceptable, and perceived to be effective in improving patient-centered outcomes for people living with HF. Yet, as we show, developing and testing palliative care interventions in HF, and in particular, interventions that attempt to infuse primary palliative care into existing outpatient care settings (such as cardiology clinics) has challenges, which we sought to overcome.²⁹ Results of a pilot study such as ours should not be used to drive clinical care decisions until subsequent fully-powered trials demonstrate efficacy, effectiveness, and potential unintended consequences. Nevertheless, this intervention presents an exciting model of how subspecialty care can be reconceptualized to incorporate primary palliative care principles into usual care. Future research is needed that tailor interventions to the illness trajectory, symptom burden, and care delivery patterns of HF to make palliative care widely accessible. Although our findings are promising, our results reinforce the need for additional work, such as a multiphase optimization strategy trial,³⁰ to refine the components of PENPal-HF before proceeding to a multi-site efficacy trial.

Abbreviations:

CASA

Collaborative Care to Alleviate Symptoms and Adjust to Illness

ENABLE-CHF

Educate Nurture Advise Before Life Ends

PENPal-HF

Primary Education for Nurses in Palliative care-HF

RE-AIM

reach effectiveness adoption implementation maintenance