Introduction

Integration of palliative care into cancer care is highly endorsed by the World Health Organization, the Institute of Medicine, the European Society of Medical Oncology, and the American Society of Clinical Oncology [1, 2]. Palliative Care has increasing relevance for the care of advanced cancer patients [3, 4]. By improving quality of life with aggressive symptom management, reducing the cost of care, and possibly improving survival, palliative care can no longer be confused with care at the end of life [57].

Elements of palliative care include standard definitions, component parts, models of care delivery, standardized assessment tools, educational programs, fellowship training, and accreditation. Very little is known about the structure of palliative care programs internationally. In light of this, the Multinational Association of Supportive Care on Cancer (MASCC), the European Society of Medical Oncology (ESMO), and the European Association of Palliative Care (EAPC) surveyed palliative programs through each society’s membership.

Methods

Research design

This study used an analysis from a cross-sectional survey conducted through the MASCC website using Survey Monkey (www.surveymonkey.com/). The survey was in two parts and involved palliative care program leaders (Fig. 1). The first part centered on palliative care integration into oncologic care and is reported separately. This project was reviewed by the Cleveland Clinic Institutional Review Board and was granted exemption from consent. The main method of this survey was a questionnaire which was modified from the original questionnaire by Dr. Hui and colleagues [8].

Fig. 1
figure 1figure 1figure 1figure 1figure 1figure 1figure 1

Palliative care leadership survey

Sample

This data was generated from members of MASCC, ESMO, and the EAPC, who completed the survey on the website. Members were informed of the survey and invited to complete the survey by entering the MASCC website. Repeated invitations were made until no further surveys were received. The survey was sent November 15, 2011, to 676 MASCC members and made available on the MASCC website for non-MASCC members. The survey was removed from the website April 20, 2012.

Questionnaire

For the purposes of this study, a survey questionnaire was designed using a modified version of the survey tool developed by Dr. Hui and colleagues with their permission [8]. Modifications involved a question about designated ESMO centers. Exemption from written or verbal informed consent was obtained from the Institutional Review Board of the Cleveland Clinic. Palliative Care program leaders entered the MASCC website to complete the survey using Survey Monkey (www.surveymonkey.com/). The questionnaire queried palliative care program leaders on the name of the program, how long the program was in existence, and the type of services provided (Fig. 1). Leaders were asked if their care included pediatric patients. Questions involved professional backgrounds of the program leader, number of full-time equivalence, certification and patient continuity. In regards to inpatient palliative care, leaders were asked about inpatient palliative care beds, duration of stay in hospital, number of inpatient discharges per month, median survival after admission, reasons for admission, number of family conferences, and advanced directives. Questions regarding consultative services involved the presence of a consultative service, availability, number of referrals per month, survival, and referral sources. Regarding outpatient care, leaders were asked about outpatient clinics, number of days per week a clinic is held, referrals per month, and survival. A number of other questions were asked about institution-operated hospices, education services, fellowship programs, mandatory palliative care rotations for oncology fellows and other trainees, grand rounds, length of training for fellows, accreditation requirements, research and palliative care funding. Responses are reported for all leaders combined and broken down by whether the program is relatively new (less than 5 years old) or mature (greater than 5 years old).

Statistics

Findings were described in number and percentages. Inferential statistics involved the Fisher’s exact test for factors with two levels, chi-square test for unordered categorical factors with greater than two levels, Cochran-Armitage trend test for ordered categorical factors and the Wilcoxon rank sum test for measured factors. A p < 0.05 was considered statistically significant.

Results

Sixty-two program leaders completed the survey (Table 1). One responder indicated that his or her institution did not provide palliative care services and was therefore excluded from analysis. Program names were often described using a single phrase. Thirty-six percent (22/61) used a single descriptor “Palliative Care”, 39 % were described in two or more phrases (Table 1). The majority of the palliative programs completing the survey were >5 years (43/61, 70 %), 3 (5 %) were <1 year, 4 (7 %) 1–2 years, and 11 (18 %) 3–5 years old. Forty-nine (80 %) of programs had mobile/consultative teams, 80 (82 %) has palliative care/supportive clinics, and 38 (62 %) has inpatient beds dedicated to palliative care. Only half (30 or 49 %) had all three services and only 25 % (15) had an institutionally operated hospice. There was a trend for older programs to have outpatient clinics (p = 0.09). Types of services offered included assessment and management of psychiatric disorders, discussions regarding advanced directives, resulting ethical complex issues, assisting with hospice referral, developing comprehensive care plans, pain and symptom management, and psychosocial support. Most programs (59 %) offered seven to eight services. Approximately 1/3 of programs cared for pediatric patients. Approximately half (48 %) had >5 inpatient nurses dedicated to palliative care; 10 % had >5 outpatient clinic nurses dedicated to palliative care.

Table 1 Survey results

Most palliative care program leaders who completed the survey were oncologists (74 %) though 69 % of respondents recorded more than one specialty. The majority of programs have a chaplain, dietitian, mid-level provider, rehabilitation personal, psychologist, and social worker dedicated to palliative care. Less than half (27 of 62 programs, 46 %) had a dedicated pharmacist. The median number of full-time equivalent physicians was 2 (range 0–15) with half (55 %) of programs providing physicians with 20 % academic protected time. Thirty-five (58 %) programs required physician certification in palliative care or to complete a palliative care fellowship; 47 % (28 /62) of program required nurse certification in palliative care.

Most programs (75 %) provided continuity (followed patients for greater than 4 weeks or throughout the course of disease). Seventy-four percent (43) had dedicated acute care beds within a unit with the median number of 10 beds (range 0–43), median length of stay was 10 days (range 3–98), and median discharges per month was 24 (range 2–250). The primary reason for admission to the inpatient palliative unit was symptom management (60 %). Admission for reason of imminent death was 10 %. However, the median inpatient palliative care mortality rate was 40 % (range 2–99 %). The median number of days from admission to death was 10.5 (range 3–55 days). Major sources of referral to the inpatient unit were transfers from non-ICU acute care beds (20 %), emergency departments (10 %), and outpatient clinics (20 %), and only 5 % coming from intensive care units. In half of programs more than 75 % of patients in the hospital on palliative units had psychosocial assessments during admission, had family conferences during the inpatient stay, and most had standing DNR orders. However, only 36 % of programs had the oncologist attend >75 % of family meetings, while 45 % of programs (N = 19) stated that the oncologist attended 15 % or fewer family conferences. Although palliative inpatient consultative services were uniformly provided, most (57 %) were not 24 h/7 days a week services. On average, services had 25 referrals per month (range 3–400), and most sources of referral were medical and radiation oncology and surgery. The median time from inpatient consultation to death was 23 days (range 7–97).

Outpatient clinics were held a median of 5 days/week (range 0.5–7) and saw a median of 30 new referrals per month (range 3–250). Most referrals were from medical and radiation oncology. Approximately half (45 %) had dedicated palliative care outpatient clinics only, 19 % saw patients in oncology clinics only and an additional 19 % saw patients both in palliative care outpatient clinics and oncology clinics.

Approximately 1/3 of palliative care programs (37 %) had active fellowship training, half (52 %) had one to two clinical fellows per year and half (48 %) had more than two clinical fellows in their program. Length of training for fellows to be certified was less than 1 year for 67 % of programs. Half of programs (56 %) had one to two research fellows per year. Excluding programs without ties to cancer care, half (56 %, 27/48)) of programs provided palliative care rotations for oncology fellows, 33 % (14/43) for radiation oncology fellows, 51 % (23/45) for other fellows/residents, 35 % (15/43) for medical students, 9 % (3/31) for pediatric oncology fellows, and 61 % (34/56) trained mid-level providers. Half of programs incorporated Grand Rounds at least once per week (68 %) or two to three times per week (32 %).

Approximately 2/3 of programs (64 %) had research programs, usually consisting of a physician (100 %), data analyst (75 %), research nurse (72 %), and psychologist (56 %). Less than half of these programs (44 %) were fully staffed and almost (62 %) had outside funding, usually from private foundations and philanthropy. Most research programs conducted prospective trials (86 %); half were involved in retrospective studies, case reports/series, and qualitative studies. For the most part, research programs reported their findings in palliative care and oncology journals. Seventy percent of these had one publication in a palliative care journal over the last year, 68 % had at least one publication in an oncology journal, and 49 % had at least one in a general medical journal.

The number of young palliative care programs was small so results need to be taken with caution when compared to more mature programs. More mature programs had shorter hospital stays (median of 9.5; range 3–96) versus younger programs (median 14.5; range 9–98; p = 0.007). Mature programs had a greater number of referrals (median 30, range 3–400) versus younger programs (median 15; range 4–40; p = 0.04). Mature programs provided palliative care rotations for non-oncology fellows and residents more frequently (60 vs. 20 %; p = 0.04).

Discussion

A picture emerges from this survey which characterizes palliative care programs. The average program has been in existence greater than 5 years and has multiple services (inpatient, inpatient consultative, and outpatient). These programs have a limited number of nurses dedicated to palliative care and their leaders have an oncology background plus an addition specialty. Programs usually consist of two full-time physicians who follow patients through the trajectory of illness, manage 10 inpatients, see inpatient and outpatient consults daily, perform a large variety of tasks, and provide teaching in the form of Grand Rounds. A subset will also have fellowship programs with extra responsibilities. Subsets also do research, usually with less than fully staffed research services. More mature programs see more patients, have shorter hospital length of stays and interact with greater frequency with other specialties than oncology. Most programs do publish in peer review journals. This survey did not include administration time, local meetings, and national or international organization responsibilities.

Standardized definitions of palliative care are lacking and the term palliative care has little or no meaning to patients [1]. The use of the term palliative care may put patients off from accepting services as discussed in recent publications from the USA [9, 10]. Symptom management in this setting is largely supportive.

Different palliative care service models are present throughout Europe. Day care centers are largely present in the UK and hence were excluded from the survey. Palliative care inpatient beds per million population range from >5 to >1 depending on the region or country. The number of palliative care specialists range from 1 per 100,000 to 1 per 1.6 million population [11].

Challenges to palliative care structures and service are prioritization based on both the oncologist and patient perspective. Since the greatest advantage to palliative care services occurs when utilized early in the course of advanced cancer as reported in the USA, outpatient clinics and availability are key [4, 12]. Shared care is primarily achieved through outpatient clinics. Inpatient consultation services and inpatient palliative care units are largely centered on crisis intervention. Multidisciplinary palliative care team values are largely measured by indirect cost savings.

Palliative care as an element of cancer care requires a medical oncologist dedicated to palliative care [13]. The key to success is collaboration, commitment, continuity, communication, and patient centeredness. To achieve this goal, palliative care programs need to be available with multiple service lines in order to provide continuity. Oncologists need to continue to be involved with patients on palliative care units to provide input as to disease course, availability, and appropriateness of further anti-cancer therapy and to communicate with families about goals of care within family conferences.

Palliative care services are underfunded for education and research. Only 1/3 of programs have fellowship training programs. Grant support is largely from private foundations or philanthropic donations rather than governmental sources.

Although there are no universally embraced quality indicators for the organization of palliative care, a number of quality indicators were generated from a Europall project systematic review [14]. These quality indicators include 24/7 availability of the consultative team, a staff that includes a pharmacist, and a single caregiver that can coordinate care to provide continuity across service lines. All team members should be accredited and a psychosocial assessment along with symptom assessment should occur within 24 h of admission to a palliative care unit. Palliative care programs should have quality assurance, research and fellowship programs [14]. In this regard, many programs fall short in one or more areas.

Educational exchanges between adult and pediatric palliative specialties are very limited. This may be largely due to different cancers and in types of terminal diseases and perhaps different symptom management strategies. The psychosocial stressors and clinical problems are distinctly different. There is probably limited value for pediatric palliative fellows to rotate on adult services.

There are differences between the present study and the survey reported by Dr. Hui [8]. The majority of programs in the present survey have been in existence for 5 years compared with programs surveyed by Dr. Hui (43 of 62, 70 % vs. 39 of 96 or 41 %). More programs surveyed in the present study followed patients in continuity (4 weeks or greater) than in the study published by Dr. Hui (45 of 62.73 % vs. 26 of 96.27 %). Half of programs surveyed in both studies required physician certification in palliative medicine. Most leaders of palliative programs in the present study were oncologists with additional specialty training (74 %) whereas in the Hui study only 17 % had an oncology background. The great majority of palliative programs in both surveys had consultative services. However, only a minority provided a 24-h–7 day-a-week consultative service. The median time from inpatient consultative referral to death was 23 days in the present study and 7 days in the Hui study. Most programs in the present study had inpatient palliative care beds (38/62, 61 %) whereas most programs in the Hui study did not have dedicated beds (27/96, 28 %). Median length of stay in hospital was 10 days in both studies. The death rate on the inpatient palliative care units was 40 % in both studies. Dedicated palliative outpatient clinics were available in approximately half of programs reported in both studies. Approximately 20 % of programs had palliative specialists participate in oncology clinics. However, the time from referral to death in the present study was 40 days and in the Hui study 90 days. Palliative care clinics were available 5 days a week in the present study and 2 days a week in the Hui study. Referrals to outpatient palliative care clinics were greater the present study then programs surveyed by Dr. Hui (30 per month vs. 7–14 per month). Less than half of programs had palliative fellowships. Mandatory oncology fellow rotations in palliative care were greater in the programs surveyed in the present study (46 %) than in the Hui study (22–27 %). This was also true of medical student rotations (26 % vs. 12–13 %). Nearly 2/3 of programs (64 %) in the present study had research as part of the program structure whereas less than half (46 %) of programs in the Hui study were involved with research. The difference between programs may be the result of the ESMO designated palliative programs surveyed in the present study which are required to have a well-developed palliative program integrated into oncology for ESMO recognition. Differences also may reflect differences in palliative care philosophy, service structure and referral between Europe and the USA.

There are several weaknesses of this study. Palliative programs surveyed were largely western European. Responders were likely to be from well-established palliative programs probably embedded in cancer centers. The results would likely to be different if rural and community hospital programs were surveyed. Survey results were influenced by those willing to complete them on line and will likely reflect well-developed programs. We did not query program leaders about financial support of their programs. Questions 76–81 were completed by programs with research structures, though the questions were intended for all completing the survey. This would have skewed the results. The program activity data was largely based on approximate estimates as it is not known how many of these programs maintain ongoing quality assurance data to provide more accurate responses.

Conclusions

Palliative care consists of multiple service lines which provide continuity throughout the course of cancer. Patients are usually seen late in the course of their illness as demonstrated by a high death rate on inpatient palliative care units and short survival in patients seen as outpatients. In general, palliative care is under resourced.