The Appropriate Provision of Primary versus Specialist Palliative Care to Cancer Patients: Oncologists' Perspectives

Abstract

Background:

Many cancer patients do not receive recommended palliative care (PC). Oncologists' perspectives about PC have not been adequately described qualitatively and may explain some of the gaps in the delivery of PC.

Objective:

To characterize U.S. oncologists' perceptions of: primary and specialist PC; experiences interacting with PC specialists; and the optimal interface of PC and oncology in providing PC.

Design:

In-depth interviews with practicing oncologists.

Setting/subjects:

Oncologists working in: the general community, academic medical centers (AMC), and Veterans Health Administration.

Measurements:

Semistructured telephone interviews with 31 oncologists analyzed using matrix and thematic approaches.

Results:

Seven major themes emerged: PC was perceived as appropriate throughout the disease trajectory but due to resource constraints was largely provided at end of life; oncologists had three schools of thought on primary versus specialist PC; there was an under-availability of outpatient PC; poor communication about prognosis and care plans created tension between providers; PC was perceived as a “team of outsiders”; PC had too narrow a focus of care; and AMC-based PC evidence did not generalize to community practices. Oncologists noted three ways to improve the interface between oncologists and PC providers: a clear division of responsibility, in-person collaboration, and sharing of nonphysician palliative team members.

Conclusions:

Oncologists in our sample were supportive of PC, but they reported obstacles related to care coordination and inpatient PC. Inpatient PC posed some unique challenges with respect to conflicting prognoses and care practices that would be mitigated through the increased availability and use of outpatient PC.

Introduction

The National Academy of Medicine, the American Society of Clinical Oncology, and the National Comprehensive Cancer Network recommend palliative care (PC) to be provided for every patient with cancer, soon after diagnosis and concurrent with anti-neoplastic treatment.^1–3 These recommendations stem from robust evidence establishing the ability of PC to improve patient satisfaction,⁴ quality of life,^5,6 and overall survival.^5,7

Despite the positive evidence base for PC and resulting practice guidelines, recent research indicates that gaps persist between these recommendations and actual practice of PC. In the Veterans Health Administration (VA), an environment that has long been strongly supportive of PC, only 52% of decedent cancer patients received any PC consultation before death.⁸ There are a number of factors that may contribute to this low receipt of PC, including under-availability of PC specialists,⁹ lack of patient understanding of PC, patient/family reticence to PC, or oncologist-related factors. The latter may include oncologists providing PC themselves (called primary PC) rather than referring this care to another provider, and/or oncologists' perceptions of the merits of specialist PC.

Here, we characterize medical oncologists' perceptions of PC, including primary versus specialist PC, and whether/how anti-cancer treatment and specialist PC should be integrated. We focus on medical oncologists for two reasons. First, oncologists largely direct the care of cancer patients and, thus, serve as gatekeepers to other services, including PC. Second, evidence indicates that physician attributes and views are the strongest predictors of end-of-life care patterns.^10,11 Although PC is not simply end-of-life care, it is certainly recommended and appropriate for patients in their terminal stages. Thus, to bridge the gap between recommendations and practice of PC for cancer patients, it is important to understand oncologists' views toward these services. Our qualitative exploration of oncologists' attitudes provides depth and resonance that previous closed-ended surveys cannot achieve.

Methods

Participants/recruitment

We used snowball and maximum variation sampling to purposefully recruit a national sample of oncologists from July 2015 to December 2015. Snowball sampling, a recognized qualitative sampling technique, was used because oncologists are a difficult-to-reach population with demanding work schedules.¹² We used investigators' professional networks and oncology professional e-mail lists to initially contact oncologists. Maximum variation sampling was used as a secondary strategy to diversify our sample and to improve our understanding of how PC is delivered in different medical settings across the United States, specifically the community, academic medical centers (AMC), and the VA.¹² Analytic memos were written after each interview to document data saturation and preliminary results. Oncologists were recruited for the study on a rolling basis until theme exhaustion was reached.

Interviews

We conducted semistructured telephone interviews to characterize oncologists' perspectives about PC for patients with advanced cancer. Respondents were asked open-ended questions about their perceptions of primary and specialist PC for patients with advanced solid tumors (defined as late stage III or stage IV), including who should provide it, when it should be provided, and the integration and interface of PC and oncology (Appendix 1 includes interview questions). Questions were vetted with a panel of oncologists (D.W.B., M.P., K.R.), clinicians (S.M.A.), and qualitative experts (C.T., P.A.K.) and pilot tested before inclusion in the study. This study was approved by the Stanford University's Institutional Review Board; participants provided verbal informed consent.

Data analysis

A health economist with content expertise (R.G.) and an anthropologist with qualitative methods expertise (A.N.) conducted interviews lasting from 22 to 76 minutes. Interviews were digitally recorded and transcribed verbatim. We conducted multiphase qualitative analysis, which involved matrix¹³ and thematic analyses,¹⁴ to answer the following research questions: (1) What are oncologists' perceptions of PC? (2) What role do oncologists play in providing PC to their patients? (3) What is the current and optimal interface between oncologists and PC specialists? Investigators independently reviewed and summarized each transcript into a matrix (based on domains from the interview guide), resolving any discrepancies through consensus. Investigators each inductively identified candidate themes and collectively refined this list to develop final themes. Lastly, a coding by committee approach¹⁵ was used to apply the final themes to the matrix. Further details about the analysis can be found in Appendix 2.

Results

We started with a seed of 231 oncologists and tracked 35 snowball recommendations to yield a total of 31 oncologists for in-depth interviews. Nine oncologists practiced in the community, 11 in AMCs, 9 in VA, and 2 served in administrative roles (Table 1). Oncologists varied in length of professional experience, ranging from 4 to 40 years of practice. Most oncologists were full-time practicing clinicians.

Table 1.

Participant Characteristics

Characteristic	N (%)
Gender
Female	15 (48)
Male	16 (52)
Age
30–39	4 (13)
40–49	7 (23)
50–59	4 (13)
60+	10 (32)
Declined to state	6 (19)
Race/ethnicity
White	9 (29)
Asian/Asian American	2 (6)
Other	1 (3)
Declined to state	19 (61)
Geographic region
West	17 (55)
Northeast	3 (10)
South	11 (35)
Year completed residency/fellowship
2010–present	6 (19)
2000–2009	9 (29)
1990–1999	5 (16)
1980–1989	6 (19)
1970–1979	5 (16)
Foreign medical school graduate
Yes	8 (26)
No	22 (71)
Declined to state	1 (3)
Academic affiliation
Yes	20 (65)
No	11 (35)
% Time clinical
0–19	2 (6)
20–39	3 (10)
40–59	2 (6)
60–79	4 (13)
80–100	10 (32)
Declined to state	10 (32)
% of patients estimated to have metastatic cancer
0–19	5 (16)
20–39	7 (23)
40–59	10 (32)
60–79	2 (6)
80–100	4 (13)
Declined to state	3 (10)

We identified seven major themes regarding oncologists' views of providing primary PC and their interactions with PC specialists, and three oncologist-identified ways to improve the interface of PC and the oncology team. The themes are described in detail below with exemplary quotes for each theme provided in Table 2.

Table 2.

Exemplary Quotes for Each Theme

Theme	Quote
Theme 1: Specialist PC is an extra layer of support that is perceived as appropriate at any point in the disease trajectory, but it is usually provided at end of life.	“We have consultations for people who have problems with pain or other symptoms, but a lot of it is geared towards end of life care only and not really palliative care earlier on in the course of the person's illness. So a lot of issues that people have don't really get addressed in that form, and I think it should occur far earlier in the course of their treatment. Maybe palliative care is brought in at the time the person is found to have incurable cancer or [for] people who don't have incurable cancer—if the usual things are not controlling their symptoms and nausea, that we would have a more broad-based approached to palliative care.”
	Group 1: Oncologist refers patient out for palliative care
	“The first time a patient comes with a metastatic tumor or some unresectable tumor, I'll refer them to palliative care if they are not already referred. It's really good timing. Actually, it does help us quite a bit. A lot of these patients come with discomfort and painful situations. [Palliative care] gets involved early on and can manage the pain. Plus, they give [patients] a lot of information about what's expected of the chemo and side effects So, they prepare the patients even before they come to see the oncologist. That helps a lot of the patient's expectations. For the families, it also prepares them to face the ongoing problem or the expected problems from any treatment. I think it's a very good service.”
	“In the beginning, in oncology, [symptom management was] all we could do. And when we didn't have so many treatment options, the oncologist would say, ‘Okay, you've got one or two chemotherapy choices,’ and then their job was symptom management. Now we have 15 treatment choices and the decisions, the discussions are so much more sophisticated that there really needs to be a division of labor. I feel, in the complicated world of molecular medicine, multiple treatment options—particularly in an academic center where you're talking about clinical trials—the ideal thing is that the oncologist really should be able to focus on explaining the patient's treatment options, and explaining the anticipated toxicity for a future treatment, and do the management when someone is on a particular regimen to make adjustments in the regimen. But ultimately people with advanced cancer have constipation, and have pain, and have neuropathy, and have insomnia, and have depression, and have anxiety, and that's where the hospice and palliative care symptom-management folks can really take over and help.”
Theme 2: Oncologists' views on primary vs. specialist PC fall into three distinct schools of thought.	Group 2: Oncologist provides some palliative care and refers difficult patients to specialists
	“Anyone that I think might benefit I refer to the palliative care physician. If someone looks like they're going to do very well for a long time, for example, a patient with metastatic cancer but who is completely asymptomatic, that's not someone I would necessarily try refer to palliative care. Part of it's resource management as well. Because when we only have one part-time palliative care physician, we can't just send every single patient. We have to prioritize. If it's pain, we can deal with that.”
	Group 3: Oncologist provides palliative care
	“I personally would be embarrassed to have a patient with cancer whose pain is poorly controlled, whose nausea is poorly controlled, whose bowel regimen is poorly controlled and that I need a palliative care specialist to help me do that. I think that much of the symptom management should be provided by the medical oncologist. It should not require a separate clinic or a separate individual to get involved in that. If the patient is being cared for in a medical oncology clinic, I believe there should be a social worker there who's going to help. There should be a nurse there who's going to help. They should have ready referrals to mental health professionals. So, I would be unenthusiastic about a practice in which my job was to pick the chemotherapy dose and delivery mechanism and everything else is done by the palliative care team. That would seem to me to be a poor process for patient care. And to make [chemotherapy] decisions while pretending that [patients'] pain is not an issue and their nausea is not an issue and their emotional health is not an issue is part of the reason why there are situations [with] tension between the palliative care providers and the medical non-palliative care providers. When the people making decisions about the cancer care are not paying attention to those other issues and deferring them to other providers, I think you make worse decisions about whether the patient should or should not receive chemotherapy in an advanced cancer situation.”
Theme 3: Specialist PC is heavily concentrated in the inpatient arena; oncologists would prefer greater availability of outpatient PC.	“The worst [system] is what in fact is the most common, which is the only palliative care is administered to people who get admitted to the hospital. They do not follow patients after they leave the hospital. The care is fragmented and episodic, and it provides absolutely no continuity. Also, in many, many respects, it gets patients at absolutely the wrong time. You have the intervention before the patient winds up in the hospital, not after the patient winds up the hospital. So, it's wrong place, wrong time. It would be a lot more beneficial if it could be delivered more routinely on an outpatient basis.”
Theme 4: Poor communication about prognostication and care creates tension between oncologists and PC specialists.	“[Palliative Care physicians] go too much into the prognosis part. Having cancer at stage 4 is actually curable. So, we're doing symptom management but we shouldn't put them as terminal.”
	“If the palliative care team has a particular pain regime, it isn't clear who's going to be providing the patient with their pain medications. And either we haven't talked about it with palliative care people or they say they're going to take over but they're not available 24/7 and then the patient comes in or is short of pain medication and not able to reach someone. Therefore, we need to order the pain medication and yet it might not be clear, exactly, what the palliative care team's plan is. So, having optimal communication is really key there.”
Theme 5: PC specialists are a “team of outsiders” that some oncologists viewed positively and others denoted negatively and that impacts patient care.	Positive perception
	“Sometimes patients knowing that the palliative care person is not actually making the treatment decision or that they're not letting that doctor down by saying ‘I don't want to have more treatment’—that maybe makes it easier at times for them to talk to someone who's not an interested party in the same way the treating physician is.”
	Negative perception
	“Oncologists, in many ways because they know more about the underlying diseases, are better at anticipating what might crop up and bother someone, and getting that taken care of in an anticipatory fashion rather than after the fact of pain already occurring or some other symptom already occurring. I think the patient care relationship is better served by having these discussions between the treating oncologist and the patient steadily as people are going through the process, rather than invoking another team of strangers later on in the disease when things are not going well. I do think that in effect, oncologists are delivering primary care because most of the time in advanced cancer patients, they're the main person in charge of the patient's care, and that they should regard themselves as the primary source of this kind of management. The other thing that is a problem for us on the inpatient side is the palliative care gets called when we already know the patient and have been caring for the patient for a long time, and I regard that as, in effect, unnecessary and potentially actually intrusive.”
Theme 6: PC provides too narrow a focus of care.	“I haven't had patients find [the palliative care service] very helpful. They're very focused on getting their advanced directive done and they don't even really help with the [patient's] pain. I think what most people need is the psychological support.”
Theme 7: The evidence base for specialist PC comes from tertiary care centers, and it does not reflect the reality or needs of many other practice settings.	“The problem is that some of this gets written about by people who are tertiary-referral hospitals. Tertiary referrals just don't have the same relationship with their patients that primary oncologists in the real world do, and tertiary-referral centers say, ‘Gee, it works really great when we send the patient to palliative care. They get better pain management and so forth.’ But I think that is a result of the fact that the aim of the oncologist in that system is really to study the disease, to get patients on trials. They have a different set of purposes than the primary care of the patient. I have actually been out in the real world as well and the relationship of the oncologist to the patient is not the same as it is in a tertiary-referral center. Here in a public hospital we do some clinical trials, but that's not our raison d'être, as it were. And so we are not as invested in that particular aspect of people's care and therefore less invested in their end-of-life care.”

Theme 1: Specialist PC is an extra layer of support that is perceived as appropriate at any point in the disease trajectory, but it is usually provided at end of life

Oncologists described specialist PC as an important layer of support that should be provided at cancer diagnosis, during treatment, when treatment is no longer advantageous, at end of life, and even after death to family members. Although PC was seen as quality-of-life enhancing care that could be provided throughout the course of illness, oncologists noted that in their practice environments, specialist PC was typically provided to patients at end of life. This was, in part, due to limited availability of PC specialists and/or outpatient PC, forcing providers to prioritize those patients with the most severe symptoms or the most advanced disease as candidates for PC. Oncologists acknowledged that more concerted efforts are needed to address this gap in the delivery of PC and to provide earlier access to specialist PC.

Theme 2: Oncologists' views on primary versus specialist PC fall into three distinct schools of thought

The first school of thought was represented by oncologists who did not provide PC but referred their patients to PC specialists. These oncologists did not see themselves as providers of PC due to limited training or time to focus on palliation and a preference for focusing on oncologic skills such as chemotherapy. For example, these oncologists noted that they were not able to properly elicit patients' goals or preferences or that there were symptoms or patient personality types that posed large challenges. Some noted that oncologists focusing on providing chemotherapy while PC specialists focusing on providing valuable services regarding goals-of-care discussions, symptom management, or psychosocial support created an optimal situation in which each provider was practicing at the top of her skillset. The second school was represented by oncologists who provided some PC themselves (primary PC) and referred patients to PC specialists when needed for various reasons such as illness progression, uncontrolled symptoms, or patients who were not accepting of their prognosis. They felt it important that oncologists provide some level of PC, but noted that they were not able (due to limited skills, limited time, or both) to provide the intensive PC that certain patients needed. The third school was composed of oncologists who saw themselves as the sole providers of PC and did not typically refer patients to PC specialists. These oncologists noted that their medical training provided them with the skills necessary to provide PC, that they had inadequate access to PC specialists within their medical center or rural environment, or that good PC was an integral component of high-quality oncologic care.

Theme 3: Specialist PC is heavily concentrated in the inpatient arena; oncologists would prefer greater availability of outpatient PC

Oncologists reported that the most common way that specialist PC was provided to their patients was through inpatient consults. Oncologists expressed several concerns about PC in the inpatient setting, including that patients were not provided follow-up PC after leaving the hospital and that patients may feel deserted when being “shunted off” to PC specialists at the end of life. However, the most serious concern expressed was that inpatient PC teams only see cancer patients at their most severe stage, leading them to believe that all cancer patients in the same stage of illness are similarly ill. For example, oncologists noted situations in which patients with Stage IV cancer were admitted to the hospital for pneumonia and received a hospitalist-activated PC consult, resulting in the PC physician telling the patient that he was going to die and switching him from anti-neoplastic treatment to comfort care. This posed concerns for oncologists, both because they felt they would be able to extend patient survival by treating the pneumonia and because it created problems in communicating with patients and families. One oncologist noted about PC physicians: “They don't understand that a cancer patient can look really crappy but not be on death's door; it's the side effects of chemo or they have a kind of cancer that's going to respond really well. [Palliative care physicians have] told people they were going to die that aren't even dying, and then it's this whole awful backpedaling and making us look bad. It's just ridiculous. We've had a lot of meetings with them, trying to explain this, and now they have a rule where they're supposed to call us first but it doesn't always happen. I don't expect them to understand oncology and to understand what diseases might really turn around with treatment, but it's done a lot of damage, actually, between our relationship with our patients.” Oncologists also noted problems with care coordination with the inpatient PC team, noting that it was often the hospitalist who consulted PC, and that the medical oncologist was frequently left out of both treatment decisions and communications regarding the patient visit and goals of care discussions. In another, minority, view, some oncologists felt that inpatient PC specialists were useful, because they were able to attend to hospitalized patients, obviating the need for the oncologist to leave her clinic to do so.

Oncologists emphasized a great need for PC in the outpatient setting, which could provide earlier access to PC, a better opportunity for care coordination, increased rapport with patients and PC teams, and improved continuity of PC services. Despite the potential benefits of outpatient PC, oncologists reported an under-availability of PC-trained providers in the outpatient setting, which they anticipated was due to healthcare administrators choosing to provide clinic space to higher-revenue generating providers. Although this was not a major area of emphasis among the oncologists we interviewed, one concern raised was that patients might perceive outpatient PC as burdensome or would be unwilling to utilize this service due to potential unwillingness to come in to clinic while battling serious illness or symptoms. One oncologist noted that outpatient PC visits should be scheduled on the same day as oncology clinic visits to avoid this barrier.

Theme 4: Poor communication about prognostication and care creates tension between oncologists and PC specialists

Oncologists described having poor communication with PC specialists about prognosis and care, and described the following challenges.

With respect to prognosis, oncologists reported that they and the PC specialists often had different views for the same patient. Compounding the problem was that each discipline was unaware of the other's differing prognosis. This resulted in mixed messages given to the patient about prognosis and/or treatment options. Disparate prognoses across PC physicians and oncologists were perceived as linked to the largely inpatient nature of PC. Oncologists noted that PC physicians only saw those cancer patients who were the most ill—hence their admission into the hospital—and this colored their perception of all patients in that same cancer stage. These incongruent perceptions about patient prognosis resulted in another point of contention: Oncologists felt that many of their patients were under-treated for conditions such as pneumonia, because PC physicians erroneously believed that the patient was at the end of life.

Regarding care coordination, oncologists reported uncertainty about which specialist was providing specific services. For example, oncologists noted that PC physicians prescribed a certain medication, but it was unclear whether the PC specialist ordered the medication or whether that was the responsibility of the oncologist. Oncologists also reported being left out of communication and treatment decisions for their hospitalized patients, noting that the internal medicine house staff often immediately consulted PC when a patient with metastatic disease was admitted. This resulted in PC making treatment decisions and communicating about prognosis to the patient without the involvement of the oncologists, which was perceived poorly by the latter.

Theme 5: PC specialists are a “team of outsiders” that some oncologists viewed positively and others denoted negatively and that impacts patient care

Oncologists referred to PC specialists as a “team of outsiders” that had both positive and negative implications for patient care. The positive implication of being a “team of outsiders” was that PC specialists provided a neutral source of support for patients to speak openly about their goals without feeling pressured to pursue treatment recommendations made by the oncologist. By virtue of their separate role, some oncologists perceived PC specialists as being better able to elicit patient preferences regarding the termination of anti-neoplastic treatment. The negative implication of being a “team of outsiders” was that PC specialists were described as “strangers” who had few interactions with patients, less rapport, and limited understanding of the patient's medical history, all of which were perceived as having poor consequences for shared decision making.

Theme 6: PC provides too narrow a focus of care

Oncologists perceived PC as having a limited scope, although there was variation reported in what this scope was. Some oncologists noted that PC physicians focused exclusively on sedation or pain medications, which they felt they could adequately handle themselves. Other oncologists noted that they were reticent to send patients to PC specialists, because the latter provided each patient with a goals-of-care conversation without offering additional services, which they perceived as not appropriate or necessary for all patients with metastatic disease.

Theme 7: The evidence base for specialist PC comes from tertiary care centers, and it does not reflect the reality or needs of many other practice settings

Some respondents noted that the practice of referring patients to PC was derived from research conducted in tertiary-care centers and may not be generalizable to other settings. There were two reasons cited for this poor external validity: Oncologists who practice in tertiary care centers are systematically different from those who do not, and patients who seek care at these centers are different from those who do not. Respondents noted that oncologists working in tertiary care centers are more focused on clinical research and have different relationships with their patients than do oncologists who work in public hospitals or in the general community, thus creating a necessity for PC specialists than is not needed in the latter environments. Other respondents noted that the patients who choose to receive care at a tertiary center are more interested in aggressive treatments and, thus, may need different care practices than those who are seen in other environments.

Oncologists identified three ways in which oncologist and PC specialist collaboration can be improved

Although oncologists described various challenges in their interactions with PC specialists, they also provided suggestions about how to improve collaboration and the delivery of PC services. Their suggestions fell into three categories. Category 1 consisted of clear divisions of responsibility. Oncologists noted that role clarity between oncologists and PC specialists was important to support collaboration, especially whether the approach to patient care was a team-based or “divide-and-conquer” approach. Oncologists also noted that specifying the services desired during a PC consult (e.g., pain management, psychosocial support) was a way to reduce friction with PC providers. Some oncologists also noted that it was helpful to specify the patient's prognosis and whether the patient had been informed about his prognosis as a way to avoid presenting contradictory or surprising information about terminal status. Category 2 consisted of in-person collaboration between oncologists and PC specialists, which was identified as having more positive effects than communicating via the electronic medical record or via telephone. In-person collaboration could take many different forms, including: joint patient visits with the oncology team; co-locating PC clinics alongside oncologist clinics; in-person meetings between the oncology and PC teams to discuss the treatment of multiple patients; or including PC specialists in cancer committee meetings. Category 3 consisted of PC and oncology sharing support staff. This approach differs from Category 2 in that it does not require the direct communication or collaboration of oncologists and PC physicians. Rather, it relies on social workers, nurses, or other nonphysician PC team member staff serving as members of both the oncology and PC teams and, thus, functioning as a liaison between the two.

Discussion

Our interviews with oncologists practicing in the VA, AMCs, and the general community revealed that respondents were generally supportive of the use of PC in the treatment of patients with advanced cancer. Although this may be at odds with common views of oncologists' attitudes, it is consistent with the sparse previous research conducted in AMCs.^16,17 However, we identified a number of potential explanations for lower-than-recommended rates of referral to specialist PC for advanced cancer patients.

First, we found that many of our respondents reported providing PC themselves (primary PC), with primary PC ranging from pain or symptom management to the more complex goals-of-care discussions. Primary PC did not necessarily obviate the need for specialist PC, but it did relegate specialist PC to be utilized in the cases of more difficult patients. Conversely, other oncologists reported the rapidly changing and growing complexity of the cancer treatment landscape as necessitating more PC physician involvement due to the opportunity costs of oncologists' time. Strategies to improve patient receipt of PC will differ across these oncologists' preferences. For oncologists who prefer to refer all PC needs to specialists, automatic referrals may be an effective strategy. For oncologists who prefer to provide all or much PC themselves, automatic triggers to assess goals of care or patient symptoms may be a strategy that is successful. This may be especially useful in community practices, as respondents noted that the division of cancer care responsibility between oncologists and PC physicians stems from AMC-based evidence and did not generalize to these settings. This finding has important implications for future research not only regarding PC but regarding other medical conditions as well.

The predominant mode of delivery of PC in the United States is inpatient.¹⁸ The heavy concentration of PC in the inpatient setting and the under-availability of outpatient PC created a host of sequelae that appear to negatively affect interactions between oncologists and PC specialists. The resulting triaging of only the most sick patients for PC may be distorting PC physicians' views of the severity of illness of patients with metastatic disease. This, in turn, may also be biasing oncologists to not refer to PC, because they perceive that PC will focus on comfort care and potentially under-treat their patients. Greater availability of outpatient PC and its earlier integration into standard oncology care may address these challenges. They may also mitigate the negative perception of PC as a “team of outsiders” making crucial decisions without having a rapport with the patient or an understanding of the patient's history. Greater availability of outpatient PC may also allow for better in-person collaboration with oncologists, which respondents noted as an important aspect of a positive working relationship.

Collaboration among multiple physicians often results in fragmented and poor-quality care.¹⁹ Care fragmentation and communication difficulties also appear to plague interactions between PC specialists and oncologists, as found in this study as well as in other works.²⁰ In this study, oncologists reported poor communication with PC physicians as well as poor communication between patients and providers as a barrier to referral and effective co-management. Oncologists in our sample also reported ways they address these challenges, including being clear in the consult order as to what specific services were requested, what the oncologist's prognosis was for the patient, and whether the patient was aware of his prognosis. Other work evaluating provider perceptions of PC on heart failure has also found that clarity regarding PC physicians' responsibilities is crucial to physician acceptance of PC providers.²¹ Oncologists in our sample also noted in-person (rather than phone or written) communication between oncologists and PC physicians facilitated collaboration.

Although the perception of PC as end-of-life care has been mentioned in the literature as a barrier to PC referral in U.S. oncology,^17,22 respondents in our sample viewed PC as appropriate at multiple stages in the disease trajectory. This is in line with other empirical evidence²³ and may represent a change in how the role of PC is viewed more generally. Despite this attitude, oncologists we interviewed reported that the provision of PC remained in practice predominately relegated to end of life, not due to oncologists' perceptions but rather due to structural factors of limited availability and subsequent need to triage patients. As the demand for PC outpaces its supply,^9,24 further work is needed to understand whether the components of PC assessment and intervention can be disaggregated and appropriately provided by other nonphysician practitioners, including nurses and social workers.

This is a qualitative study and as such, our findings are not meant to be generalizable to the population of U.S. oncologists. Rather, the aim of this work is to provide a rich and nuanced understanding of the range of ways in which oncologists think about primary and specialist PC and the integration of PC in oncology. Our sample consisted of 31 oncologists, an appropriately large sample size for reaching data saturation.^25,26 However, it is possible that we did not fully capture the range of opinions of all oncologists, especially those who were not supportive of PC. This qualitative study provides the foundation on which further work can be conducted to explore the extent to which these attitudes hold across different care environments and provider types.

Oncologists are key stakeholders in improving how PC is delivered to advanced cancer patients. They serve as both gatekeepers to referral and the most common primary PC providers. Their views of the major barriers and potential solutions to gaps in PC are thus important in their own right and in guiding quality improvement efforts. Our results suggest that those efforts should include expanding the availability of specialist outpatient PC services to provide patients with longitudinal care that is practiced in a collaborative manner alongside standard oncology care. They also suggest more structured communication about the goals of consultation and existing prognostic information to avert duplication and fragmentation of services. Both these strategies deserve further examination in efforts to foster the natural alliance of oncologists and PC providers in serving their common patients.

Footnotes

Acknowledgments

Support for this research study is provided by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, and Health Services Research and Development (Project Number PPO 14-371). This work was presented as a poster at the American Society of Clinical Oncology 2016 annual meeting, June 3–7, 2016 in Chicago, IL. Support for Dr. C.T. is provided by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, and Health Services Research and Development (VA HSR&D RCS 00-001).

Author Disclosure Statement

No competing financial interests exist for Drs. A.N., P.A.K., C.T., S.M.A., and R.G. During the past two years, Dr. D..W.B. has had stock or other ownership interest in Clinical Oncology Advisory Group and PRM Pharmaceutical. He has had a consulting or advisory role with Clinical Oncology Advisory Group, Bristol-Meyers Squibb, Physician Resource Management, OnQ, Pfizer, MedScape, and Cascadian. His institution has received research funding from Amgen. Dr. M.P. has had consulting or advisory roles with Celgene and Castlight. Dr. K.R. has had a consulting or advisory role with Clovis.

A ppendix 1. Interview Questions

A ppendix 2. Qualitative Analysis Process

References

Institute of Medicine: Delivering High-Quality Cancer Care: Charting a New Course for a System in Crisis. Washington, DC: The National Academies Press, 2013.

Peppercorn

, Smith

, Helft

, et al.: American society of clinical oncology statement: Toward individualized care for patients with advanced cancer. J Clin Oncol, 2011; 29:755–760.

National Comprehensive Cancer Network: NCCN Guidelines®. 2016. www.nccn.org/professionals/physician_gls/f_guidelines.asp (Last accessed August 8, 2016).

Brumley

, Enguidanos

, Jamison

, et al.: Increased satisfaction with care and lower costs: Results of a randomized trial of in-home palliative care. J Am Geriatr Soc, 2007; 55:993–1000.

Temel

, Greer

, Muzikansky

, et al.: Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med, 2010; 363:733–742.

Bakitas

, Lyons

, Hegel

, et al.: Effects of a palliative care intervention on clinical outcomes in patients with advanced cancer: The Project ENABLE II randomized controlled trial. JAMA, 2009; 302:741–749.

Bakitas

, Tosteson

, Li

, et al.: Early versus delayed initiation of concurrent palliative oncology care: Patient outcomes in the ENABLE III randomized controlled trial. J Clin Oncol, 2015; 33:1438–1445.

Gidwani

, Joyce

, Kinosian

, et al.: Gap between recommendations and practice of palliative care and hospice in cancer patients. J Palliat Med, 2016; 19:957–963.

Lupu

, American Academy of Hospice and Palliative Medicine Workforce Task Force: Estimate of current hospice and palliative medicine physician workforce shortage. J Pain Symptom Manage, 2010; 40:899–911.

10.

Obermeyer

, Powers

, Makar

, et al.: Physician characteristics strongly predict patient enrollment in hospice. Health Aff (Millwood), 2015; 34:993–1000.

11.

Cutler

, Skinner

, Sterling

, et al.: Physician beliefs and patient preferences: A new look at regional variation in health care spending. National Bureau of Economic Research, 2013. www.nber.org.laneproxy.stanford.edu/papers/w19320 (Last accessed August 5, 2016).

12.

Palinkas

, Horwitz

, Green

, et al.: Purposeful sampling for qualitative data collection and analysis in mixed method implementation research. Adm Policy Ment Health, 2015; 42:533–544.

13.

Averill

: Matrix analysis as a complementary analytic strategy in qualitative inquiry. Qual Health Res, 2002; 12:855–866.

14.

Braun

, Clarke

: Using thematic analysis in psychology. Qual Res Psychol, 2006; 3:77–101.

15.

Saldaña

: The Coding Manual for Qualitative Researchers. Thousand Oaks, CA: Sage Publications, Inc, 2015.

16.

Bakitas

, Lyons

, Hegel

, et al.: Oncologists' perspectives on concurrent palliative care in a National Cancer Institute-designated comprehensive cancer center. Palliat Support Care, 2013; 11:415–423.

17.

Fadul

, Elsayem

, Palmer

, et al.: Supportive versus palliative care: What's in a name?: A survey of medical oncologists and midlevel providers at a comprehensive cancer center. Cancer, 2009; 115:2013–2021.

18.

Morrison

, Maroney-Galin

, Kralovec

, et al.: The growth of palliative care programs in United States hospitals. J Palliat Med, 2005; 8:1127–1134.

19.

Kripalani

, LeFevre

, Phillips

, et al.: Deficits in communication and information transfer between hospital-based and primary care physicians: Implications for patient safety and continuity of care. JAMA, 2007; 297:831–841.

20.

, Mileshkin

, Doan

, et al.: Acceptability of early integration of palliative care in patients with incurable lung cancer. J Palliat Med, 2014; 17:553–558.

21.

Bekelman

, Rabin

, Nowels

, et al.: Barriers and facilitators to scaling up outpatient palliative care. J Palliat Med, 2016; 19:456–459.

22.

Von Roenn

, Voltz

, Serrie

: Barriers and approaches to the successful integration of palliative care and oncology practice. J Natl Compr Canc Netw, 2013; 11 Suppl 1:S11–S16.

23.

LeBlanc

, O'Donnell

, Crowley-Matoka

, et al.: Perceptions of palliative care among hematologic malignancy specialists: A mixed-methods study. J Oncol Pract, 2015; 11:e230–e238.

24.

Schenker

, Arnold

: The next era of palliative care. JAMA, 2015; 314:1565–1566.

25.

Namey

, Guest

, McKenna

, et al.: Evaluating bang for the buck: A cost-effectiveness comparison between individual interviews and focus groups based on thematic saturation levels. Am J Eval, 2016;OnlineFirst:1–16.

26.

Guest

, Bunce

, Johnson

: How many interviews are enough?: An experiment with data saturation and variability. Field Methods, 2006; 18:59–82.

27.

Miles

, Huberman

, Saldaña

: Qualitative Data Analysis: A Methods Sourcebook. Thousand Oaks, CA: Sage Publications, Inc., 2013.