Integrating Patient and Expert Perspectives to Conceptualize High-Quality Palliative Cancer Care for Symptoms in the US Veterans Health Administration: A Qualitative Study

Abstract

Quality measurement is typically the domain of clinical experts and health system leaders; patient/caregiver perspectives are rarely solicited. We aimed to describe and integrate clinician and patient/caregiver conceptualizations of high-quality palliative symptom care for patients receiving care for advanced cancer within the US Veterans Health Administration in the context of existing quality measures. We conducted a secondary qualitative analysis of transcripts from prioritization discussions of process quality measures relevant to cancer palliative care. These discussions occurred during 2 modified RAND-UCLA appropriateness panels: a panel of 10 palliative care clinical expert stakeholders (7 physicians, 2 nurses, 1 social worker) and a panel of 9 patients/caregivers with cancer experience. Discussions were recorded, transcribed, and independently double-coded using an a priori logical framework. Content analysis was used to identify subthemes within codes and axial coding was used to identify crosscutting themes. Patients/caregivers and clinical experts contributed important perspectives to 3 crosscutting themes. First, proactive elicitation of symptoms is critical. Patients/caregivers especially emphasized importance of comprehensive and proactive screening and assessment, especially for pain and mental health. Second, screening and assessment alone is not enough; information elicited from patients must inform care. Measuring screening/assessment and management care processes separately has important limitations. Lastly, high-quality symptom management can be broadly defined if it is patient-centered; high-quality care takes an individualized approach and might include non-medical or non-pharmacological symptom management. Integrating the perspectives of clinical experts and patients/caregivers is critical for health systems to consider as they design and implement quality measures for palliative cancer care.

• What do we already know about this topic?

• Patients/caregivers have different perspectives from clinical experts on the most important aspects of high-quality care and numerous quality measures addressing palliative care for cancer symptoms have been developed.

• How does your research contribute to the field?

• Groups of clinical experts and patients/caregivers engaged in discussions about high-quality palliative cancer care in the context of existing quality measures, contributing complementary perspectives on the importance of symptom screening and assessment, the need for stronger linkages between symptom identification and management, and recognition that symptom management can be broadly defined.

• What are your research’s implications toward theory, practice, or policy?

• Patients/caregiver perspectives provide critical and complementary perspectives when health systems are identifying priorities for quality measurement and improvement.

Introduction

The perspectives of patients and caregivers on the experience of care differ in important ways from the clinical experts who set standards for high quality care. Patients’ and caregivers’ unique perspectives are still underrepresented in health care systems’ quality measurement and improvement efforts compared to those of clinicians and administrators. While patient and caregiver voices are increasingly contributing to health services research and health systems governance,^1-3 patient and caregiver involvement in quality improvement is still relatively uncommon.⁴ Prior efforts to involve patients and caregivers in the development or prioritization of quality measures have found that patients bring a distinct voice from clinical experts.^5-7 Patients emphasize different aspects of high-quality care, like communication and access to care.^8-10

These aspects are especially important for patients with advanced cancer, who are often weighing the benefits and harms of continued aggressive treatment and need extra supportive care. Early, routine palliative care can greatly improve quality of life for patients with advanced cancer^11-14; and seriously ill patients, caregivers, and clinicians all consider pain and symptom management critical aspects of care for patients at the end of life.¹⁵ However, numerous care gaps in palliative care quality have been documented.^16-19 Quality measurement is a critical component for quality improvment,²⁰ but the quality measures health systems track may not capture aspects of pain and symptom care most important to patients and caregivers.

Palliative care for patients with advanced cancer is an important target for quality improvement.^16,17 The US Veterans Health Administration (VA), which is the largest provider of health care in the US, has been at the forefront of making palliative care widely accessible²¹ and implementing processes to enable goals-of-care communication for seriously ill Veterans.²² VA provide high-quality palliative care compared to other US providers,²³ but VA does not routinely measure or monitor the quality of palliative care they provide.

Numerous and diverse quality measures relevant to palliative care, such as cancer pain and symptom management have been developed.^24,25 Some common symptoms of advanced cancer, such as pain and dyspnea, have validated quality measures for each stage of the care process, from screening, diagnosis/assessment, to management and iterative follow-up. Other common symptoms, such as fatigue and anxiety, have fewer measures, and others have no measures at all.²⁶ In this study, we sought to engage both clinical experts and patients and caregivers to reflect on existing process quality measures’ importance to providing high-quality patient- and family-centered palliative cancer care.

Methods

Approach

We conducted this work as part of the Improving Palliative Measurement Application with Computer-Assisted Abstraction (ImPACS) Project (Veterans Affairs [VA] HSR&D Investigator-Initiated Research [IIR] 17-277), which aimed to identify patient and caregiver priorities for cancer quality measurement and to operationalize prioritized measures using computer-assisted methods. For this analysis, we leveraged qualitative data obtained during 2 prioritization panels on palliative care quality measurement. One panel comprised health system clinical expert stakeholders and the second comprised patients and caregivers with cancer experience. We conducted a qualitative content analysis of verbatim panel transcripts using the logical process of the symptom care trajectory, using each stage as the code list (Figure 1; code list in Supplemental Appendix). A methodological summary is shown in Table 1.

Figure 1.

Logical process of the symptom care trajectory.

Table 1.

Methodological Summary.

Study activity	Quality measure synthesis	Clinician panel	Patient/caregiver panel	Panel transcript coding	Subtheme generation	Crosscutting theme generation
Approach	Literature scan and content analysis	Adapted RAND-UCLA Appropriateness panel	Adapted RAND-UCLA Appropriateness panel	Qualitative coding (deductive)	Content analysis (inductive)	Axial coding (inductive)
Description	Systematic literature scan for existing, validated quality measures relevant to palliative care of patients with advanced cancer	VA and non-VA clinicians rated and discussed measures on 1-9 scales on “importance for providing patient- and family-centered palliative cancer care” and “potential for quality improvement” before and after a 1-day in-person meeting	VA and non-VA cancer patients and caregivers rated and discussed measures on 1-9 scale on “importance for providing patient- and family-centered palliative cancer care” before and after a 1-day in-person meeting	Coders marked segments of transcript text using codes from an a priori logical framework representing patients’ trajectory of symptom care	Within each code, similar segments were iteratively grouped together and subthematic statements generated	Subthematic statements grouped together into crosscutting themes
Data and/or Results Generated	List of 226 existing measures. Similar measures were grouped resulting in 64 measure concepts.	Quantitative ratings of 64 measure concepts on 1-9 scale, nominations of “top 5” measure concepts, verbatim transcripts of discussions of reasons behind ratings and nominations	Quantitative ratings of 21 measure concepts rates as high priority by clinician experts on 1-9 scale, nominations of “top 5” measure concepts, verbatim transcripts of discussions of reasons behind ratings and nominations	Segments of panel transcripts coded with 5 codes representing patients’ trajectory of symptom care	14 subthemes	3 cross-cutting themes
Analytic Team	CO, AW, KL, CL, KG	CO, AW	CO, AW	CO, RG, KG	CO, KG, RG	CO, KG

Qualitative Data From Stakeholder Panel Meetings

Data are verbatim transcripts from 2 stakeholder panels on palliative care quality measurement priorities for health care facilities administered by the U.S. Department of Veterans Affairs. Both panels used modified RAND-UCLA Appropriateness Panel methods to rate and discuss relative merits of concepts of interest.²⁷ Each panel was held in-person over a single day, recorded with participant consent, professionally transcribed and de-identified. We aimed to recruit at least 9 participants per panel, which promotes diversity of viewpoints while being small enough to allow participants to engage in the conversation.²⁷

Prior to the panels, we conducted a measure scan and synthesis for the concepts discussed by the panels; we grouped 226 existing process quality measures into 64 “measure concepts”;²⁵ similar measures with different operationalizations were grouped together (eg, “pain screening” or “depression management”). Materials summarizing the evidence for the clinician panel were adapted into lay-friendly materials for the patient/caregiver panel (available from the authors upon request).

Qualitative data for the clinician panel comprise the verbatim panel transcript (length: 6 h 43 min); panel recruitment and discussion elicitation methods have been described previously.²⁸ Briefly, we aimed to recruit clinical and subject matter experts with experience in palliative care and oncology who represented critical decision makers in palliative care quality measurement (Table 2). Participants represented VA’s Office of Nursing Services, National Center for Ethics in Health Care, Pain Program Office, Oncology Program Office, Office of Patient-Centered Care and Cultural Transformation, and the Performance Accountability Workgroup, as well as expertise in health services research, quality measurement, and ethics. The panel was chaired by a palliative care physician. Prior to the meeting, panelists were asked to review evidence on existing measures synthesized into 64 measure concepts and provide numerical ratings for each respect to importance for providing patient- and family-centered palliative cancer care and potential for quality improvement in advance of the meeting (rating sheet available in the Supplemental Appendix). At the meeting, participants received summary information on all ratings and their own ratings, and the panel chair went through each measure and elicited comments from participants, probing areas where there was more disagreement and allowing for discussion. After these discussions, participants re-rated the measure concepts. At the end of the meeting, participants were asked to nominate their “top 5” measure concepts with highest importance and potential for quality improvement, and participants discussed their choices. Participants were offered the opportunity to comment exhaustively through structured conversations around ratings and unstructured conversations of additional input. There were balanced opportunities for discussion of all topics; no topics were rushed through. Quantitative rating results have been previously reported²⁸; the meeting transcript was analyzed to identify themes.

Table 2.

Description of Panels and Participants.

	Clinical experta panel	Patient/caregivera panel
Participants	10 (8 in-person, 2 telephone)	9 (all in-person)
Date	September 18, 2019	March 5, 2020
Panel location	Washington, DC	Santa Monica, CA
Women/Men	6 women, 4 men	2 women, 7 men
VA/non-VA	7 VA, 3 non-VA	8 VA, 1 non-VA
Health system/cancer experience	7 physicians, 2 nurses, 1 social worker	7 patients (including 3 with advanced cancer)a, 2 caregiversa
Geography	California (1), Connecticut (1), District of Columbia (2), Maryland (1), New York (2), North Carolina (2), Washington (1)	California (3), Florida (1), Massachusetts (1), Michigan (1), New Jersey (1), Virginia (1), Washington (2)

^aClinicians primarily reflected on their experiences in health systems as clinicians, leaders, and researchers, and occasionally on their personal and family experiences of illness. Although we recruited participants with respect to understanding of their status as “patients” or “caregivers,” nearly all participants reflected on their experiences in health systems as both a patient and caregiver, including personal experiences and experiences of family and friends with cancer and non-cancer illnesses and injuries. Patients/caregivers similarly had a variety of professional experiences, including in medicine and law, that also informed their perspectives.

Qualitative data for the patient/caregiver panel (6 h 22 min) comprise the verbatim panel transcript; panel recruitment and discussion elicitation methods have been described previously.²⁹ Briefly, we aimed to recruit patients and caregivers with lived experience of cancer, representing diverse geographies, demographics, and cancer diagnoses and stages (Table 2). We recruited patients with advanced and less advanced cancers; participants had to be well enough to travel and participate in an all-day in-person meeting. Participants were primed to represent the perspectives of patients who were too ill to represent themselves and think about their own needs in the face of cancer progression. The panel was co-chaired by a patient/caregiver with experience in patient engagement in research and a physician clinical expert (SMA). Prior to the meeting, panelists were asked to review lay-friendly materials on existing measures synthesized into 20 measure concepts rated most highly by the expert panel and provide numerical ratings for each respect to importance for providing patient- and family-centered palliative cancer care in advance of the meeting (rating sheet available in the Supplemental Appendix). At the meeting, participants received summary information on all ratings and their own ratings, and the panel chair went through each measure, explaining clinical concepts if necessary, and elicited comments from participants, probing areas where there was more disagreement and allowing for discussion. After these discussions, participants re-rated the measure concepts. At the end of the meeting, participants were asked to nominate their “top 5” measure concepts with highest importance and participants discussed their choices. Participants were also asked to review the list of 44 measures and discuss any they thought were too important to leave out, as well as a discussion of ideas for which there was no measure concept presented. These discussions resulted in an additional measure concept (pain screening) being added back into consideration during the discussion. Participants were offered the opportunity to comment exhaustively through structured conversations around ratings and unstructured conversations of additional input. There were balanced opportunities for discussion of all topics; no topics were rushed through. Quantitative rating results have been previously reported²⁹; the meeting transcript was analyzed to identify themes.

Qualitative Data Analysis

The analytic team included 3 social scientists with qualitative methods expertise who were present at both panel meetings (CEO, RCG, and KFG). We used a combination of deductive and inductive approaches to identify themes in the panel discussion transcripts. An analysis of the panel discussions was pre-planned, however the approach was developed following a review of the transcripts to focus on palliative symptom care (an equal amount of time was spent on measures related to communication with the patient and family, but those discussions will be synthesized in another analysis). CEO and RCG independently coded transcripts using a deductive approach using the trajectory of the symptom care process as codes, which mapped well to the topical structure of the panel conversations (Figure 1, additional details available in Supplemental Appendix). These codes were: (1) the experience of illness (36 text sections); (2) screening, diagnosis, and/or assessment (132 text sections); (3) management, treatment, plan (105 text sections); (4) follow-up or iterative management (44 text sections); and (5) documentation and quality monitoring (38 text sections). Text segments corresponding to each code were reconciled through discussion until consensus around code definitions and application was reached. Then, content analysis methods were applied within each code to identify subthemes. Axial coding was then performed on the subthemes to synthesize similar subthemes across codes into cross-cutting themes. Analysis was done in qualitative analytic software Atlas.ti version 8.

Results

Three cross-cutting themes were identified from the 14 subthemes (Table 3). Additional exemplar quotes supporting thematic statements for each stage of the patient’s trajectory of illness are shown in the Supplemental Appendix.

Table 3.

Thematic Statements with Underlying Subthemes.

Theme	Cross-cutting thematic statements	Subthematic statements
I	High-quality palliative care proactively elicits patients’ physical and mental symptoms	High quality palliative care acknowledges patients’ reluctance to bring up their symptoms
		High quality palliative care takes a holistic and comprehensive view of patient needs
		High quality palliative care includes pain and mental health screening because patients may not bring up these issues
		High quality palliative care symptom management can only occur if patients are appropriately screened and assessed
		High quality palliative care should balance encouraging comprehensive assessments with the negative consequences of overly structuring patient interactions
II	High-quality palliative care uses information elicited from or provided by patients to inform care	High quality palliative care promotes confidence that health systems can and will address patient concerns
		High quality palliative care consistently follows up with issues identified through screening
		High quality palliative care leverages the interdisciplinary team
		High quality palliative care uses information elicited from patients to determine care priorities
III	High-quality palliative care for management of symptoms uses an individualized approach	High quality palliative care overcomes the inertia of health care systems
		High quality palliative care symptom management is conceptualized more broadly than medications
		High quality palliative care treats to patient’s definition of success, not an arbitrary standard
		High quality palliative care symptom management anticipates and acknowledges the side effects of curative and palliative treatments
		High quality palliative care follows up with patients on both efficacy and side effects of management plans

Theme I: “How Are You Going to Know?” High Quality Palliative Care Proactively Elicits Patients’ Physical and Mental Symptoms

Both stakeholder panels emphasized the importance of proactively eliciting physical and mental health symptoms from cancer patients. The patient/caregiver panel emphasized why patients might not bring up pain and mental health symptoms unprompted for a variety of reasons. First, they believed that clinicians would ask them about symptoms if they were relevant to their care and if their symptoms were addressable. Patients assumed there was no point to bringing up their symptoms if they thought there were “no options” other than “a bunch of drugs, and I don’t want a bunch of drugs.” Second, they may not bring up their symptoms or minimize their symptoms to avoid being seen as needy or “weak.” This issue was especially salient for Veterans with cancer as military culture prizes pain tolerance and can make people feel stigmatized asking for help. One patient/caregiver remarked: “Yeah, we tend to be our own worst enemy in that department. . .I’ll say—Yeah, you know, it hurts, but it’s not a 10, it’s a 3. . .I will diminish my pain, even though I know it hurts, just so I can continue on.” Last, patients may have adjusted to worsening chronic symptoms, such as shortness of breath. One patient/caregiver noted, “there’s the assumption that it gets taken seriously, but in the reality is, in that in chronic things, where it’s expected, it gets downplayed.”

Both panels were generally enthusiastic about broad, structured questionnaires such as symptom screening tools and assessments of physical, psychological, social, and spiritual needs (often referred to as “comprehensive assessments”). According to one patient: “I think [comprehensive assessment] goes to the very heart of palliative care. I can’t imagine how a palliative care physician could operate without having done this. Because if they don’t do this, what it is means is they’ve only looked at your medical conditions, which is only a small part of the picture.” One clinical expert related: “We had a survey in our clinic and [emotional needs assessment] is the number one thing that everybody says. . .it’s what we’ve seen from our patients.”

Both the patient/caregiver panel and the clinician panel grappled with the limitations of structured instruments that screen and assess patient symptoms and needs. Clinical experts questioned whether patients should be screened for issues that the care team may not be able to help them with or if it was better to screen and acknowledge the patient’s experience. As one expert said: “a lot of people are really fatigued and they’re frustrated nobody acknowledges it, so asking them about it is important, but if you ask everybody about it and then you do nothing about it, then that is even more frustrating. I don’t know which is worse.” Clinicians also worried that rigidly defining deadlines for structured assessments (eg, during an initial visit) could negatively affect care: “You made the first visit so complicated that sometimes very brief visits that could help with guiding the care are not done. . .I can’t see you as a walk-in, because I’m not going to fit you in my lunch break. . .I think we just need to be a little bit careful there, right? However, one patient/caregiver emphasized the importance of the conversations structured assessments enable rather than focusing on which specific assessment should be used: ‘if they don’t actually ask the question, a lot of things don’t happen. . .the most fruitful sort of assessments and plans came out in the conversation about comprehensive needs, rather than a specific checkbox or ‘tell me what your pain scale is.’”

Theme II: “Assessment Isn’t Enough” High-Quality Palliative Care Uses Information Elicited From or Provided by Patients to Inform Care

Both the clinical expert and patient/caregiver panels emphasized the inextricable link between management of symptoms and the screening and diagnosis/assessment processes that precede it. Both panels struggled with the way quality measures tend to separate these components. Separation of the care trajectory into different quality measures was viewed as diminishing the importance of the symptom care process overall; both groups found it challenging to prioritize aspects of symptom care processes over others. According to one expert discussing several spinal cord compression measures: “I wish there would be more like, a summary, you know? I think having 4 measures for these maybe dilutes the importance of each individual one.” In the words of a patient discussing several pain measures: “I think they kind of go hand in hand.”

Both the patient/caregiver and expert panels agreed that elicitation and documentation of symptoms alone was not enough—that information must be used to inform patients’ care. Patients/caregivers valued clinicians having a comprehensive understanding of their needs, but were skeptical about their ability to use the information they provided. One patient/caregiver noted, “I think most of us have had the experience of having to fill in a million patient forms. . .every single time we go to the doctor. And they end up in the trash, nobody looks at them. Unless that information is not just assessed, but actually forces the creation of a plan or an intervention, it’s useless.” Clinicians noted that increasing use of structured screening and assessment tools could elicit so much information that care could be compromised by making it harder to triage and prioritize the most important issues. One expert noted: “It can also be a slippery slope around sort of just a pan-positive review of systems and which ones you chase or pay attention to, versus not.”

Patients/caregivers and clinical experts grappled with between identification of issues through screening and diagnosis/assessment and managing these issues. One common area of frustration was the lack of coordination within teams. As one patient/caregiver related about their own cancer experience: “There was never any comprehensive care, because you didn’t get all of the players involved.” Another said, “my experience is that at world class facilities, they do an overall assessment, but different people each have their compartmentalized assessment to do, and it’s not clear that they’re necessarily talking to each other in terms of coming up with a conclusion.” While clinicians expressed concern about being penalized for the actions or inactions of other team members outside of their control, they agreed that there should be quality measures that capture the patient’s overall experience with the entire interdisciplinary team. As one expert put it: “I think part of it is what does the medical subspecialty do and what does the big picture team do? . . .to say that [mental health] is not part of medical oncology is legitimate, but to say is it part of what should be measured overall in the experience of that patient is a different question.”

Both patients/caregivers and clinical experts underscored how health systems can fail patients “at their most desperate time” by inadequately assessing symptoms, leading to unnecessary suffering. According to one patient: “I had my surgery, came home, was in excruciating pain for 2 days, called 4 physicians for help. Finally, in the middle of the night, one of the physicians, exasperated, said—Okay, bring him back into the hospital. And they took care of the pain within 2 hours. But there was no initial—no one was willing to really to believe that the pain existed or assessed it.” Clinical experts worried patients might not know how or when to seek help: “If a patient sat at home. . .and didn’t know how to engage with the medical system. . .then yes, we should be accountable. . .did patients have access to services that could help them and help them recognize that particular symptoms were actually—you know, the longer you have them, the higher chance that there’s going to be prolonged disability.”

Theme III: “It Doesn’t Mean Don’t Have a Plan When There Aren’t That Many Options” High-Quality Palliative Care for Management of Symptoms Uses an Individualized Approach

Both panels conceptualized high-quality management of symptoms in a broad way, describing how acceptable high-quality symptom management plans could incorporate elements beyond pharmacological or medical approaches, such as observation, integrative health modalities, or other non-pharmacological approaches. According to one expert, “Therapy may or may not be part of the management plan, right? It may be you’re going to just watch. So that could be part of a management plan, is observation.” One patient noted: “I really enjoy the fact that the VA now has introduced the Whole Health opportunities. . .they can go get a massage or they can get acupuncture or acupressure.”

Both patients/caregivers and clinical experts emphasized that communication and shared decision making about management plans was an important aspect of high-quality palliative care for symptoms. Since patients’ symptoms may not be attributable to the disease but instead to side effects of curative or palliative treatments, clinical experts emphasized that high-quality management should include discussions of the pros and cons of continuing or discontinuing these treatments. One expert asserted, “Like any medical intervention is, you are responsible for assessing the efficacy and toxicity of it.” Clinical experts cautioned against the subjectivity and negative consequences of “treating toward numbers” in terms of improvement in symptoms, emphasizing person-centered iterative management and monitoring processes. One expert expressed this idea in the context of dyspnea treatment: “I think what we’re trying to get at is that the patients are at a dyspnea level that is satisfactory to them or that is to a certain extent comfortable to them.”

Patients/caregivers and clinical experts also emphasized the importance of follow-up care beyond the initial prescription or recommendation. Monitoring and iterative management are especially important in cancer care because patients may not know the expected efficacy window of a new treatment or realize new symptoms may be side effects of new treatments. According to one patient/caregiver: “Pain management is just essential and having a feedback loop to measure the effectiveness of the plan versus actual, you know, it’s important.” One clinical expert discussed this idea in the context of appetite stimulants, saying: “There’s very few other things we start that have a 50/50 chance of working, where actually making the decision soon is actually important. . .if it doesn’t work now, it’s not going to work later. . .I think this is a particular wrong that happens a lot.”

Discussion

Both patients/caregivers and clinicians provided important perspectives for health systems to consider when pursuing high-quality palliative care quality measurement for patients with advanced cancer. First, proactive elicitation of patient symptoms or concerns is critical. Second, elicitation alone is not enough—there needs to be a strong link between the information patients provide and the care provided. Third, high-quality symptom management is not one-size-fits-all; it can incorporate a broad variety of patient-centered approaches.

These themes reveal important ways that existing measures fall short of measuring high-quality symptom care in patients with cancer. The use of structured screening tools falls short if it does not promote understanding of the patient as a whole person and facilitate a conversation about their most important symptoms and concerns. Measuring management and the screening, assessment, and diagnostic processes that preceded as separate processes makes sense when quality is measured at the clinician level, but does not capture patient experiences with the whole care team or health system; care coordination may be a fruitful area for development of new process quality measures in palliative care.^26,30 Lastly, high-quality symptom management could incorporate evidence-based non-medical and non-pharmacological approaches to management, as well shared decisions to stop management. As these care processes may not be detectable in structured administrative data,^31,32 this may be a use case for augmenting quality measures with clinical notes using natural language processing.^33-36 Existing process quality measures may be best used in concert with patient-reported outcome measures³⁷ by more tightly linking care processes to the outcomes that matter to patients.

Limitations of this analysis include the framing of these discussions around the 64 quality measures concepts from our measure scan²⁵ (in the case of the clinician panel²⁸) and the 20 rated high-priority by the clinician panel (in the case of the patient panel²⁹). Some important domains of palliative care received little discussion due to a relatively small number of relevant validated measures, though both panels were given the opportunity to bring up additional measures or concepts without associated measures that they believed were important to consider. The panels also sometimes focused on different aspects of the symptom care trajectory. Discussion about patients’ experience of illness mostly occurred during the patient/caregiver panel, while discussions about documentation and quality monitoring were more common during the clinician panel. Patients/caregivers and clinicians were similarly engaged in dialog that was coded for the middle stages of the care trajectory (screening and diagnosis/assessment, management, and follow-up). The present analysis was limited to care for patient symptoms and concerns; future work will focus on advance care planning, goals of care conversations, and communication with loved ones as critical components of high-quality palliative care in a separate analysis. Though our panel sizes were in line with recommendations for our chosen study design, they are small and mostly represent perspectives of VA patients.

In this secondary analysis of transcripts from 2 quality measure prioritization panels, the voices of patients and caregivers augmented those of clinical experts in important ways, and vice versa. Patient and caregiver voices added emotional depth to the discussion of relevant priorities, relaying how pain has been minimized or ignored, or how open-ended questions at the right moment averted mental health crises. At the same time, clinicians contributed important views on the difficulties and potentially negative unintended consequences of potential approaches. Health systems may find that integrating the perspectives of both clinical experts as well as patients and their loved ones will help them best achieve their aims of providing patient- and family-centered care.