Abstract

Background. Palliative care is an important, complex aspect of primary care, requiring a multidisciplinary approach. The Gold Standards Framework (GSF), a programme used by over 3000 UK practices, aims to facilitate high-quality palliative care through the introduction of systematic clinical and organizational processes. Quality payments for palliative care are available to UK practices which maintain registers and hold multidisciplinary meetings.

Objectives. To explore the effectiveness and sustainability of the implementation of GSF at practice level.

Methods. The study followed a qualitative comparative case study design using in-depth interviews and observational data with 15 practices participating in GSF, from three areas differing in socio-geography. Semi-structured interviews (total 45) with GPs, community nurses and practice managers were supplemented by observation of practice meetings and systems, to provide contextual insights. Transcribed interviews were analysed using a thematic matrix approach and comparisons were made within and between practices. Practices were identified on a continuum of performance (high, medium and minimal) according to the evidence of functioning in palliative care-related activity.

Results. Considerable variation existed between practices in both the extent of palliative care-related processes and the effectiveness of inter-professional communication. High-performing practices displayed a clear-shared purpose for palliative care with effective communication, whereas minimal performing practices demonstrated little utilization of basic GSF processes and deficiencies in inter-professional communication.

Conclusion. Effective palliative care requires good team relationships and robust processes. While GSF can enable such improvements, quality measures focusing on processes alone are inadequate to distinguish good practice, questioning the effectiveness of current quality measures in UK general practice.

Introduction

End of life care is a fundamental aspect of primary care.1–3 Patients and carers have identified a vital role for primary care teams in offering continuity of care and holistic, patient-centred communication throughout the patient's last illness.4 In addition, primary care professionals also have an important role in enabling patients to achieve their preferences for end of life care.5 Primary palliative care has been a central aspect of the development of palliative care within the UK over the last 40 years,6 recently acknowledged by the National Institute for Health and Clinical Excellence7 and reflected in government policy.8,9

Several initiatives aimed at enhancing primary palliative care have emerged including the GP facilitator scheme, developed as a partnership between the Royal College of General Practitioners and the national charity ‘Macmillan Cancer Support’ 10 and the Liverpool Care of the Dying Pathway supported by Marie Curie Cancer Care.11 Arguably, the most notable initiative is the Gold Standards Framework (GSF) for palliative care which emerged from the GP facilitator project and provides a comprehensive programme for primary palliative care covering seven key areas (Box 1).12 Starting as a pilot programme with 12 practices in one locality in 2000,13 GSF has been disseminated nationally through a phased approach.14 Currently, the framework has been implemented by around 3000 practices the UK, covering about a quarter of the population,15 and has attracted considerable international interest.16

Box 1 Aspects of the GSF programme: the seven C's6

C1, Communication: Palliative care register maintained; regular PHCT meetings held to discuss and plan care for patients on the register.

C2, Coordination: A practice coordinator (e.g. GP, DN or manager) oversees the work of GSF, acting as a contact for the local GSF facilitator; record maintained of preferred place of death; named GP and DN for each patient.

C3, Control of symptoms: Effective holistic management of all aspects of care, including the use of physical assessment tools to identify and address physical symptoms and psychosocial and spiritual needs.

C4, Continuity: A key PHCT member identified for each patient and information flow maintained e.g. with out-of-hours providers.

C5, Continued learning: Practice-based education in palliative care, including audits and significant event analysis.

C6, Carer support: Identifying and meeting the needs of informal carers, before and after the patient's death.

C7, Care of the dying: Identifying terminal phase, managing it appropriately, using protocols such as the Liverpool Care Pathway; availability of anticipatory care procedures; assessing need for continued medication in last days of life; informing family when dying phase is approaching.

As with other areas of chronic disease management in general practice, a specific register of palliative patients is fundamental to the achievement of high-quality care.6,17,18 In addition, regular communication between members of the primary health care team (PHCT) to enable a multidisciplinary approach to advance care planning is necessary.19 In recognition of this, since April 2006 practices in the UK that demonstrate maintenance of a register and regular palliative care meetings are financially rewarded through the Quality and Outcomes Framework (QOF).20 The apparent success of GSF in this respect is highlighted in the data from audit questionnaires which practices complete at the outset of GSF and 12 months later, indicating that approximately 90% of practices maintained a palliative care register and 80% had regular PHCT meetings at which palliative care patients are discussed.21 While such practice activity may be a necessary prerequisite for effective primary palliative care, the extent to which these measures are sufficient to enable coherent and coordinated care to be delivered to the patient has not been studied. Practices that have taken part in the GSF provide a context for researching these issues.

The current study aimed to identify factors that may account for different levels of commitment to a multidisciplinary approach to primary palliative care by individual primary care teams. The objectives were to investigate these issues through a comparative case study approach that explored the experiences and attitudes of primary care staff within several primary care teams. In addition, by comparing the emergent insights from this study with data provided in audit questionnaires completed by practices 1–2 years earlier at the time of their initial implementation of GSF, we aimed to assess the extent to which uptake of the programme enabled a sustained commitment to primary palliative care.

Methods

A qualitative case study analysis22 was undertaken with practices from three areas which had commenced GSF implementation between March 2003 and September 2004.

Study design

Participating practices were visited between October 2005 and March 2006 and interviews were undertaken with district nurses (DNs) (normally the GSF coordinator), GPs (normally GSF lead), practice managers and other key informants—e.g. other GPs, DNs, local GSF facilitator and Macmillan nurse. Inspection of practice systems and documents, such as the palliative care register, was undertaken, and practice and locality meetings observed.

Recruitment

Three Primary Care Trust (PCT) areas were selected which demonstrated (i) contrasting levels of GSF uptake; (ii) varied socio-demography; (iii) different GSF phases to assess sustainability; and (iv) differences in input from the local GSF facilitator (a health care professional who had taken on the role of recruiting and assisting practices in GSF development). Multi-centre ethics committee and local PCT governance permissions for these areas were obtained.

Area 1 comprised affluent suburban and socially deprived inner city localities of similar size, and had achieved a high practice uptake of GSF. Area 2 centred on a small industrial town with a surrounding rural area, also with a high GSF uptake. Area 3 had low GSF uptake, and was the largest geographical area with two market towns, a city with a manufacturing industrial base and large rural hinterland.

Purposive sampling of practices that had participated in GSF was undertaken, assisted by the local GSF facilitator, to provide a varied mix of practices in terms of location, size, training status and the level of uptake of GSF. Selected practices were contacted by letter and a follow-up telephone call. Two practices in Area 2 declined to take part when initially approached; therefore, two others with similar characteristics were invited to participate.

Each recruited practice was visited with the intention of agreeing interview times with the practice manager, the GP with lead interest in palliative care, the DN who worked most closely with the practice and any other key informants suggested by the practice.

Data generation

Interviews with practice managers followed a relatively structured format focusing on practice demographics, to enable an overall view of the practice to be constructed. Interviews with GPs, DNs and other key informants were semi-structured and explored respondents’ experience of implementing GSF within the practice with particular emphasis on the utility of the palliative care register and communication issues within the practice, following the interview schedule summarized in Box 2. Interviews took 30–60 minutes to complete, and were recorded and transcribed; one GP declined electronic recording but agreed to detailed notes being taken.

BOX 2 Summary of interview guide

  • Establish how the practice became interested in GSF

  • Elicit information on current activity in palliative care

  • Probe as to whether palliative care approach changed as a result of GSF

  • Ascertain how the palliative care register was developed (criteria used)

  • Investigate whether the register is making a difference

  • Enquire as to how it is decided when patients are in last 6-12 months of life

Observations of meetings and the use of GSF processes (such as registers) were made opportunistically during visits to practices and recorded as field notes. Such observations were used to inform interview questions and prompts and to confirm the extent to which processes that were reported as having been implemented were evident (for example the extent to which meetings were well attended and how registers were being used). Such observations also provided an opportunity to make a subjective assessment of the atmosphere within the practice and its overall morale, organizational style and relationships between staff.

Analysis

Each practice was examined as a case unit, and matrix analysis23 was used to code the interviews and record key themes which emerged through the analysis. Particular notice was made of interviewees’ perceptions about improvements that had occurred to identify the attributes that facilitated them. Comparison between interviews was made to test the degree to which accounts at each practice concurred with each other. Emergent themes from interviews were triangulated with observational data to demonstrate how organizational processes, contextual factors and behavioural issues affect the practice of palliative care through the adoption of GSF.

All practices were represented in a chart to allow comparisons of the extent to which delivery of palliative care through GSF principles had been achieved, on the quality of multidisciplinary relationships, and the degree of correspondence in responses from interviewees in the same practice. The identification of a continuum which allowed differentiation and rating of practices according to performance in these areas was discussed with an experienced local GSF facilitator to enable subtle differences to be detected. Consistency in the approach to analysis and interpretation was maintained through a process of iterative assessment conducted by KM and JD.

Following completion of analysis of the case study data, findings were compared with GSF audit questionnaire data which had been completed by these practices at the time they entered the GSF programme and 12 months later. This was to enable assessment of the sustainability of the GSF approach to primary palliative care.

Results

Nine practices were recruited in Area 1 (the area with highest uptake of GSF) and three practices each in Area 2 and 3 (Table 1). A total of 45 interviews were conducted with GPs (17), DN/Macmillan and practice-based nurses (19) and practice managers (9). The number of interviewees per practice varied according to the availability of individuals, including their willingness to participate. Where practice managers were unavailable for interview, demographic details about these practices were obtained from the other interviewees.

TABLE 1

Characteristics of practices participating in the case study analysis

Practices 15 
Number of GP partners in practice (range) 2–11 (part and full time) 
Patient list size (range) 2400–13 500 
Number of training practices 10 
Urban 
Rural 
Semi-rural and mixed 
Practices 15 
Number of GP partners in practice (range) 2–11 (part and full time) 
Patient list size (range) 2400–13 500 
Number of training practices 10 
Urban 
Rural 
Semi-rural and mixed 

Informants identified several benefits from maintaining a register of palliative care patients (Table 2) and holding regular meetings to review care plans for these patients. However, as illustrated in Table 3, there was considerable variation in the extent to which practices had implemented other activities in primary palliative care as exemplified by the GSF. Practices were classified into high, medium and minimal performers in primary palliative care procedures. There was no particular pattern between the size, training status and location of practices and their primary palliative care activity.

TABLE 2

Ways in which practices reported making use of the palliative care register

  • To raise awareness of patients' needs among reception staff, GP registrars and locums

  • To record patients’ name, address and treatment

  • To fast track patients for consultations and visits

  • To prioritize provision of prescriptions and emergency medication

  • To prompt medication reviews every 6 weeks, and to screen for side effects

  • To ensure proactive assessment of patients’ needs through home visits

  • To avoid the need for post-mortems through visiting regularly

  • To enable earlier identification of carer support needs

  • To ensure effective out-of-hours arrangements are in place

  • To register patients with hospices early in case respite care becomes needed

 
  • To raise awareness of patients' needs among reception staff, GP registrars and locums

  • To record patients’ name, address and treatment

  • To fast track patients for consultations and visits

  • To prioritize provision of prescriptions and emergency medication

  • To prompt medication reviews every 6 weeks, and to screen for side effects

  • To ensure proactive assessment of patients’ needs through home visits

  • To avoid the need for post-mortems through visiting regularly

  • To enable earlier identification of carer support needs

  • To ensure effective out-of-hours arrangements are in place

  • To register patients with hospices early in case respite care becomes needed

 
TABLE 3

Mechanisms and processes in participating practices: illustrative cases of high, medium and minimal performance

Performance Characteristics Registers and meetings Additional GSF activities Examples of high-functioning relationships Examples of low-functioning relationships 
High Practice A Register developed by GPs and practice manager Liverpool care pathway DNs report more prompt and appropriate palliative referrals No evidence of DN involvement in classifying patients as palliative 
PCT 2, Phase 4 Medication card use 
Patient list size in range 5001–10000 Weekly multidisciplinary discussions Recording of PPOD Educational knowledge shared 
Group, training practice Bereavement visits 
(four interviews) Critical incident meetings Carer information GPs have oversight of all patients 
Anticipatory medication Even distribution of GSF workload between GP partners 
Out-of-hours forms 
Practice E Register developed by GP Emergency medication Macmillan nurse attends PHCT meeting to discuss new diagnoses and palliative cases System appears highly dependent on lead GP: GP prefers to minute meetings rather than delegate; no meetings occur if this GP is absent 
PCT 2, Phase 4 Collectively decide when to contact out-of-hours service to report PPOD 
Patient list size in range 5001–10000 Multidisciplinary meeting every 4–6 weeks Out-of-hours forms DN indicates improved access to GPs DN unaware of criteria for identifying patients for register 
Group practice GPs more willing to attend GSF meeting 
(four interviews) 
Medium Practice H Register developed by DN GSF folder maintained GP perceived informal meetings with DN as ideal for regular interaction and immediate resolution of issues Lack of formal meetings with GP considered unsupportive by DN 
PCT 1, Phase 3 Home assessment visits 
Patient list size in range <5000 Informal’ coffee morning discussion White board  
Two-partner practice Earlier identification and response to patients’ social, spiritual and emotional needs through earlier referral to DN 
Training practice Three monthly PHCT meetings Nominated GP per case DN feels Macmillan team is controlling flow of information 
(two interviews) 
Practice I Register developed by GP White board/GSF folder GP claims DN can establish contact to discuss patients at all times by phone or between consultations, and not necessary to wait for meeting DN reports ad hoc communication with GP (e.g. request for visits) 
PCT 1, Phase 3 Emergency medication 
Patient list size in range <5000 Out-of-hours forms used DN unsatisfied with lack of briefing of recent test results and outpatient/GP appointments 
Two-partner practice No regular meetings 
Training practice Patients prioritized for consultations and visits No discussion by GP and DN of register 
(two interviews) 
Minimal Practice J GP and DN (retired) developed register White board and case file Used GSF to alter ad hoc approach (e.g. encourage GPs to consistently complete DS1500 and inform DN punctually) Current DN not sparing time to meet and discuss patients despite ‘protests’ of GP 
PCT 1, Phase 3 DN had strong awareness of how to identify palliative patients and checks whiteboard and liaises with practice manager GP partners want GSF lead to review their patients and do not attend meetings 
Patient list size in range 10001–15000 Monthly meetings not sustained Out-of-hours forms GP felt team working between GPs and DN has improved with time, though not attributable to GSF GSF implemented ‘under much duress’ following several attempts by facilitator to enlist the practice 
Group practice PHCT could not perceive advantage of GSF approach 
(two interviews)    
Practice O No register No GSF paperwork used Telephone communication with Macmillan nurses was strong, though infrequent meetings with them Practice manager not accessible to DN 
PCT 1, Phase 6 DN reported receiving no GP support 
Patient list size in range 5001–10000 No meetings GSF not a priority due to competing initiatives Significant illness list of patients displayed on board prior to GSF 
Group, training practice Ad hoc care provided 
(one interview) 
Performance Characteristics Registers and meetings Additional GSF activities Examples of high-functioning relationships Examples of low-functioning relationships 
High Practice A Register developed by GPs and practice manager Liverpool care pathway DNs report more prompt and appropriate palliative referrals No evidence of DN involvement in classifying patients as palliative 
PCT 2, Phase 4 Medication card use 
Patient list size in range 5001–10000 Weekly multidisciplinary discussions Recording of PPOD Educational knowledge shared 
Group, training practice Bereavement visits 
(four interviews) Critical incident meetings Carer information GPs have oversight of all patients 
Anticipatory medication Even distribution of GSF workload between GP partners 
Out-of-hours forms 
Practice E Register developed by GP Emergency medication Macmillan nurse attends PHCT meeting to discuss new diagnoses and palliative cases System appears highly dependent on lead GP: GP prefers to minute meetings rather than delegate; no meetings occur if this GP is absent 
PCT 2, Phase 4 Collectively decide when to contact out-of-hours service to report PPOD 
Patient list size in range 5001–10000 Multidisciplinary meeting every 4–6 weeks Out-of-hours forms DN indicates improved access to GPs DN unaware of criteria for identifying patients for register 
Group practice GPs more willing to attend GSF meeting 
(four interviews) 
Medium Practice H Register developed by DN GSF folder maintained GP perceived informal meetings with DN as ideal for regular interaction and immediate resolution of issues Lack of formal meetings with GP considered unsupportive by DN 
PCT 1, Phase 3 Home assessment visits 
Patient list size in range <5000 Informal’ coffee morning discussion White board  
Two-partner practice Earlier identification and response to patients’ social, spiritual and emotional needs through earlier referral to DN 
Training practice Three monthly PHCT meetings Nominated GP per case DN feels Macmillan team is controlling flow of information 
(two interviews) 
Practice I Register developed by GP White board/GSF folder GP claims DN can establish contact to discuss patients at all times by phone or between consultations, and not necessary to wait for meeting DN reports ad hoc communication with GP (e.g. request for visits) 
PCT 1, Phase 3 Emergency medication 
Patient list size in range <5000 Out-of-hours forms used DN unsatisfied with lack of briefing of recent test results and outpatient/GP appointments 
Two-partner practice No regular meetings 
Training practice Patients prioritized for consultations and visits No discussion by GP and DN of register 
(two interviews) 
Minimal Practice J GP and DN (retired) developed register White board and case file Used GSF to alter ad hoc approach (e.g. encourage GPs to consistently complete DS1500 and inform DN punctually) Current DN not sparing time to meet and discuss patients despite ‘protests’ of GP 
PCT 1, Phase 3 DN had strong awareness of how to identify palliative patients and checks whiteboard and liaises with practice manager GP partners want GSF lead to review their patients and do not attend meetings 
Patient list size in range 10001–15000 Monthly meetings not sustained Out-of-hours forms GP felt team working between GPs and DN has improved with time, though not attributable to GSF GSF implemented ‘under much duress’ following several attempts by facilitator to enlist the practice 
Group practice PHCT could not perceive advantage of GSF approach 
(two interviews)    
Practice O No register No GSF paperwork used Telephone communication with Macmillan nurses was strong, though infrequent meetings with them Practice manager not accessible to DN 
PCT 1, Phase 6 DN reported receiving no GP support 
Patient list size in range 5001–10000 No meetings GSF not a priority due to competing initiatives Significant illness list of patients displayed on board prior to GSF 
Group, training practice Ad hoc care provided 
(one interview) 

GSF-initiated phases: 3 (March 2003–2004); 4 (June 2003–2004); and 6 (June 2004–2005); PPOD, preferred place of death.

Levels of activity in palliative care by GSF implementation

Views and opinions of respondents regarding the extent to which their communication relationships had experienced change or improvement are given in Table 4. The quotations reflect a spectrum of performance related to GSF procedures, which we defined as high, medium and low.

Table 4

Perceptions of practice palliative care activity and the GSF programme

High (strong mechanisms and relationships) • ‘We're discussing the palliative care patients with our nurses once a week …. So we're all very much more aware of each other's value to the care of patients … (Use of registers and meetings) is really cast in stone. They’ve become a very fundamental part of the way we work …. I’m sure that will continue.’ (Practice A, GP) 
• … ‘From the GP meetings, we all know within days of a diagnosis usually … if a patient is … coming on our GSF list …. The register has been a godsend bit of a really … it did take a while for it to filter in to us how important it is but, now it's part of the picture …. In a few years it will just be the norm, the standard … part of the furniture really.’ (Practice A, Staff Nurse) 
• ‘Our skills have improved, we're more competent in what we do … there's a keen response from the doctors that we do give good care …. We're all a resource for each other …. We’re braver at networking. We will look for solutions and assistance more now …. The actual practicality of giving the care is better’ (Practice A, DN) 
Medium (moderate mechanisms and relationships) • We have a folder in which we keep all our patients diagnosed as palliative …. Being proactive has improved …. If I was to ring a Doctor … to say one of the patients in palliative care have a problem … they don't always feedback very effectively to me … it's that “I'm a Doctor and you're the Nurse and I make the decisions” (Practice G, DN) 
• ‘We bought a whiteboard, and … some education books …. We also sent that front sheet to the out of hours …. We agreed not to have meetings every month because there were not so many patients …. I couldn’t discuss my fears … the support wasn’t as great though I could see any of them (GPs) at any time …. We never discussed a patient after they died …. I don’t feel we learnt from each other’ (Practice H, DN) 
• ‘When you’ve got one GP trying to cover all patients it doesn’t quite work as well …. That’s why we don’t get the regular meetings …. I’ve never been invited to a meeting. I lie, I was invited once. It got cancelled. We use the register … the implementation has basically been left to me …. Even if I request the out of hours forms to be sent off, I have to complete them (Practice I, DN) 
Minimal (lower adoption of mechanisms) • ‘Our very good district nurse … has retired and the person that has taken over her post doesn’t have the time to implement GSF fully …. Rather than having our formalised once a month meetings … we've not had a meeting since July and it's now October, despite protests. (GSF) allows me to run searches so that people are not slipping through the net’ (Practice J, GP) 
• ‘We started out and were full of the joys of spring in actually doing it for the first month or two …. We are following the principles but not doing the documentation …. They are not writing in the newly diagnosed cancer patients …. We're discussing the patients … not … putting it in this file in reception …’ (Practice M, DN) 
• ‘One of our assumptions was that there'd actually be a wider range of information shared … the Macmillan Service … work very hard to be as accessible … but there's clearly limits to that … the Macmillan Nurses weren't available for all the meetings we had …. It certainly would be helpful …. This business about how you cooperate and work in teams …. I don't think we have a model of that … the meetings around it might work if there was sufficient capacity … it's work in progress’ (Practice N, GP) 
High (strong mechanisms and relationships) • ‘We're discussing the palliative care patients with our nurses once a week …. So we're all very much more aware of each other's value to the care of patients … (Use of registers and meetings) is really cast in stone. They’ve become a very fundamental part of the way we work …. I’m sure that will continue.’ (Practice A, GP) 
• … ‘From the GP meetings, we all know within days of a diagnosis usually … if a patient is … coming on our GSF list …. The register has been a godsend bit of a really … it did take a while for it to filter in to us how important it is but, now it's part of the picture …. In a few years it will just be the norm, the standard … part of the furniture really.’ (Practice A, Staff Nurse) 
• ‘Our skills have improved, we're more competent in what we do … there's a keen response from the doctors that we do give good care …. We're all a resource for each other …. We’re braver at networking. We will look for solutions and assistance more now …. The actual practicality of giving the care is better’ (Practice A, DN) 
Medium (moderate mechanisms and relationships) • We have a folder in which we keep all our patients diagnosed as palliative …. Being proactive has improved …. If I was to ring a Doctor … to say one of the patients in palliative care have a problem … they don't always feedback very effectively to me … it's that “I'm a Doctor and you're the Nurse and I make the decisions” (Practice G, DN) 
• ‘We bought a whiteboard, and … some education books …. We also sent that front sheet to the out of hours …. We agreed not to have meetings every month because there were not so many patients …. I couldn’t discuss my fears … the support wasn’t as great though I could see any of them (GPs) at any time …. We never discussed a patient after they died …. I don’t feel we learnt from each other’ (Practice H, DN) 
• ‘When you’ve got one GP trying to cover all patients it doesn’t quite work as well …. That’s why we don’t get the regular meetings …. I’ve never been invited to a meeting. I lie, I was invited once. It got cancelled. We use the register … the implementation has basically been left to me …. Even if I request the out of hours forms to be sent off, I have to complete them (Practice I, DN) 
Minimal (lower adoption of mechanisms) • ‘Our very good district nurse … has retired and the person that has taken over her post doesn’t have the time to implement GSF fully …. Rather than having our formalised once a month meetings … we've not had a meeting since July and it's now October, despite protests. (GSF) allows me to run searches so that people are not slipping through the net’ (Practice J, GP) 
• ‘We started out and were full of the joys of spring in actually doing it for the first month or two …. We are following the principles but not doing the documentation …. They are not writing in the newly diagnosed cancer patients …. We're discussing the patients … not … putting it in this file in reception …’ (Practice M, DN) 
• ‘One of our assumptions was that there'd actually be a wider range of information shared … the Macmillan Service … work very hard to be as accessible … but there's clearly limits to that … the Macmillan Nurses weren't available for all the meetings we had …. It certainly would be helpful …. This business about how you cooperate and work in teams …. I don't think we have a model of that … the meetings around it might work if there was sufficient capacity … it's work in progress’ (Practice N, GP) 

High performers

These were practices that displayed both efficient formal processes and supportive relationships between PHCT members. They demonstrated shared vision, mutual respect, commitment, consistency, inclusive decision-making and effective organization. They maintained registers to actively and systematically identify patients, and made use of them at regular multidisciplinary meetings to facilitate discussion and care planning. These practices often listed patients on a white board to aid daily communication, made use of a range of GSF materials to aid record maintenance and informed out-of-hours services about palliative patients to support continuity of care.

However, even within this group, there was wide variation in the frequency of meetings, ranging from weekly to bimonthly, levels of attendance and the degree to which meetings were inclusive of PHCT members. In some instances, meetings were organized specifically to discuss palliative patients and on others it constituted an agenda item within a more wide-ranging practice meeting.

The decision to include practices as high performers was swayed by the level of consistency conveyed by interviewees about the regular attendance of GP partners and DNs at meetings, the degree and quality of information exchanged and the impact it was deemed to have on the care provided by the team. Practice A, for example, demonstrated that the whole PHCT was actively involved in palliative care service delivery. In addition to meetings and registers, it had introduced a variety of processes ranging from recording preferred place of death to providing carer information and making bereavement visits. All interviewees associated with this practice independently concurred in their descriptions, reflecting high levels of shared understanding and commitment.

Medium performers

These practices demonstrated relatively well-established processes, but supportive team relationships were less evident in comparison to the high category. Not all practices had formal meetings; for example, their absence in Practice H was identified by the DN as preventing discussion of critical incidents, sharing anxieties and inhibiting joint learning. GP interviewees at these practices were generally more satisfied with the quality of communication than were DNs.

Minimal performers

These practices lacked both the formal processes and supportive relationships important in providing palliative care. A number of issues, such as conflicting organizational priorities, strained relationships and other contextual factors were apparent. For example, Practices M and N perceived GSF largely as an administrative task of enhanced record keeping rather than being a facilitator of palliative care. Both relied on informal communication rather than formal meetings. Even so, for most of these practices, there was evidence that GSF participation had increased awareness of the importance of coordination and communication in relation to primary palliative care; for example, by maintaining out-of-hours continuity and providing anticipatory drugs. Practice O represented an extreme of the sample, being resistant to the whole concept of GSF, feeling that implementation had been forced on them externally and having neither instituted meetings nor registers. They had scarcely adopted any procedures recommended in the GSF, reflecting aspects of the individualistic culture and ethos of this practice.

Comparison of observed adoption and self-reported audit data

Table 5 summarizes the audit questionnaire data that practices provided at baseline and at 12 months (final) after initiating GSF implementation, and compares this to the level of activity in palliative care that emerged from the qualitative analysis. Individual yes/no questions were summated for each subscale of the GSF questionnaire21 to give eight subscores for each of the ‘seven Cs’ (Box 1), plus extension of GSF to non-cancer patients from point of diagnisis. Total scores were produced by adding together all subscores.

TABLE 5

Comparison of audit questionnaire results at 12-months post-implementation, with qualitative rating of palliative care performance conducted in 2005/2006

Practice code, PCT and phase Baseline score Final score Difference between final and base score Rating of GSF adoption in 2005/2006 
F, PCT 3 and Phase 5 21 32 11 High 
E, PCT 2 and Phase 4 13 31 18 High 
A, PCT 2 and Phase 6 14 30 16 High 
C, PCT 1 and Phase 3 15 30 15 High 
B, PCT 1 and Phase 6 22 28 High 
H, PCT 1 and Phase 3 16 28 12 Medium 
G, PCT 2 and Phase 4 13 16 Medium 
I, PCT 1 and Phase 6 13 19 Medium 
D, PCT 1 and Phase 3 19 13 Medium 
K, PCT 3 and Phase 5 —a 24 —a Minimal 
N, PCT 1 and Phase 3 20 24 Minimal 
J, PCT 1 and Phase 3 15 21 Medium 
L, PCT 3 and Phase 5 —a —a Minimal 
M, PCT 1 and Phase 3 —a —a —a Minimal 
O, PCT 1 and Phase 6 —a —a —a Minimal 
Practice code, PCT and phase Baseline score Final score Difference between final and base score Rating of GSF adoption in 2005/2006 
F, PCT 3 and Phase 5 21 32 11 High 
E, PCT 2 and Phase 4 13 31 18 High 
A, PCT 2 and Phase 6 14 30 16 High 
C, PCT 1 and Phase 3 15 30 15 High 
B, PCT 1 and Phase 6 22 28 High 
H, PCT 1 and Phase 3 16 28 12 Medium 
G, PCT 2 and Phase 4 13 16 Medium 
I, PCT 1 and Phase 6 13 19 Medium 
D, PCT 1 and Phase 3 19 13 Medium 
K, PCT 3 and Phase 5 —a 24 —a Minimal 
N, PCT 1 and Phase 3 20 24 Minimal 
J, PCT 1 and Phase 3 15 21 Medium 
L, PCT 3 and Phase 5 —a —a Minimal 
M, PCT 1 and Phase 3 —a —a —a Minimal 
O, PCT 1 and Phase 6 —a —a —a Minimal 

GSF-initiated phases: 3 (March 2003–2004); 4 (June 2003–2004); 5 (February 2004–2005); 6 (June 2004–2005).

a

Scores could not be calculated due to insufficient or missing data.

Practices that had achieved a total final score of >24 (maximum total score = 35), with the exception of one practice, corresponded with high performance in primary palliative care. Practice H was the exception as although it had regular meetings at the time the questionnaire had been completed, these had become ad hoc and less consistent over time. Interviews at Practice H revealed communication difficulties which were felt to be hampering the delivery of palliative care. For example, although the GP interviewee was very satisfied with informal daily discussions, the DN described feelings unsupported in her palliative care work.

The boundary between GSF questionnaire scores for medium and low performers was less clear-cut, suggesting that these categories were more porous with greater movement between these groups over time. For example, three practices (D, G and I) had total final scores of 16–19 in the audit questionnaire and were placed in the medium band. Some practices had evidently improved their palliative care practice beyond the level achieved at the end of the first year as reported in the final questionnaire. For instance in Practice D meetings had initially been ad hoc with no formal discussion of patients on the register, but more recently a formalized approach to discussing patients on a bimonthly basis had been introduced. This practice had a very low baseline score (6), but made improvements at 12 months follow (score 19), which were continued and built upon over the following 2 years. However, Practices G and I had similar final scores of 16 and 19 points, but unlike practice D, appeared to have made little further progress.

On the other hand, three practices (K, J and N) had achieved moderate final scores (range: 21–24 points) but were classified as minimal performers as the improvements had not been sustained. For Practice N, the palliative care register had not been maintained, no meetings were organized and the GP informant was disgruntled with the lack of leadership and responsibility in the team and the fluctuating involvement of specialist palliative care nurses. Reasons given for failure to maintain palliative care processes included a lack of cooperation from team members after a change in personnel (Practice J), very poor processes and relationship problems (Practice N) or the identification of too few palliative patients to justify the implementation of GSF to the PHCT (Practice K).

As Table 5 shows, the audit scores of practices that were classified as high performers tended to indicate greater levels of improvement than those of other practices. High performers achieved a mean improvement in their GSF audit questionnaire scores of 13.2 (range 6–18) between baseline and 12-month follow-up and medium performers achieved a mean of 8.5 (range 3–13). Low performers had a mean of 5 (range 4–6); however, this figure needs to be treated with caution since the change between baseline and final scores could not be calculated for four of the low adopting practices due to incomplete or missing data.

Discussion

Palliative care is a central component of general practice24,25 and the findings of this study support the contention that GSF offers an effective approach to a systematic and high-quality service. This study, however, illustrates the diversity of ways by which practices organize palliative care and the extent to which the developments promoted by GSF are sustainable.

Although the qualitative data were obtained up to 2 years after the 12-month GSF follow-up questionnaire data, there was a broad level of consistency between the two sets of data, particularly for practices with high questionnaire scores. This gives support to the assertion that the changes associated with high levels of palliative care performance in practices implementing GSF are sustainable. Practices that were assessed qualitatively as being high performers in palliative care were, in general, those with the highest questionnaire scores at 1-year post-implementation, and who had also demonstrated the greatest degree of relative change during the first year of implementation.

While establishing palliative care registers and meetings was widely achieved by practices, there was considerable variation in what this entailed. Some practices viewed GSF as formalizing existing high standards of care, while for others it involved new ways of working. Supportive team relationships emerged as critically influencing effective and sustainable palliative care activity within practices. While GSF was often championed by key individuals with palliative care interests, sustainability and more extensive adoption rested on wider participation within the PHCT. Marginal performance by contrast occurred where commitment was less pervasive, practice priorities conflicted and/or where there was changing personnel.

GPs and DNs often had differing perceptions of the quality of relationships and effectiveness of palliative care processes within PHCT teams. Consistent with the findings of King et al.,26 DNs saw the main benefit of GSF as facilitating access to GPs and being provided with timely information and receiving appropriate referrals from them. Conversely, changes in DN staff attached to the team could lead to the failure of previously effective practices and relationships. These factors illustrate how effective PHCT practice in palliative care can be affected by different organizational arrangements, together with the inherent inequalities in power between GPs and DNs.27

Methodological considerations

A significant limitation of this study is the lack of a patient–carer perspective. Patients and carers have identified a consistent and holistic approach from primary care teams as being central to high-quality palliative care4; however, it was beyond the scope of this study to verify the extent to which patients and carers experience this as occurring. We intend to investigate this in a future study.

Even though, GSF has had very high uptake across the UK,15 it is still only a minority of practices that have participated in the programme. Practices in this study were self-selecting, and had adopted GSF prior to there being any specific financial incentive for so doing. These practices may therefore be seen as ‘early adopters’ who have higher levels of interest in palliative care, and therefore may not be representative.28 Further insights into the uptake and sustainability of GSF from the perspective of local GSF facilitators have been obtained through another study which will be reported elsewhere.

It was beyond the scope of this study to interview more than a small sample of informants at selected practices, and the views of interviewees may not have been shared by others within these teams. We tended to interview the GP who was most involved with GSF, and inevitably this individual will have been better informed than colleagues, possibly with a more positive view of the functioning of GSF; conversely, they may have been more critical of deficiencies. Assessing relationships is more subjective than quantifying process measures, and it is difficult to generalize from a small number of interviews and observations.

Conclusion

The evidence from this study supports the view that good primary palliative care requires a confluence of strong process and effective professional relationships. Practices which had effective organizational processes and good relationships appeared much better placed to provide a consistently high-quality multidisciplinary approach to care than those that lack these attributes. Developing the quality of primary palliative care emerged as a complex organizational process with an array of factors influencing performance and sustainability. Shared commitment and purpose, together with adequate resources, leadership and stability in the PHCT are key factors to success. The study presents evidence that GSF is an effective means of facilitating practices in delivering high-quality primary palliative care.

Given the complexity of palliative care and the varying contexts of individual practices and their patients, a variety of methods of delivering primary palliative care are likely to be appropriate. This study has highlighted some of the challenges which practices face in achieving high-quality care. While the GSF audit questionnaires give insights into the palliative care activity of practices, this study has allowed a deeper understanding of the process and has indicated potential facilitators and barriers to adoption. High performance appeared to occur when the implementation of processes promoted by GSF fitted well with the internal environment of the practice. Processes, such as palliative care registers, may be sophisticated or superficial; however, it is the way that they are utilized which affects the quality of care. Similarly, meetings may achieve little unless they are supported by good PHCT relationships. More research is needed to elucidate the leavers and barriers to achieving consistently high-quality primary palliative care.

The insights from this study questions the validity of the current approach being taken in the UK of rewarding processes, such as the introduction of registers and meetings, with financial payments associated with achieving QOF scores. It suggests the need for developing more sensitive measures of quality, both for primary palliative care as well as for other long-term conditions where the need for strong process and effective organizational relationships are likely to be fundamental to achieving quality care.

Declaration

Funding: Macmillan Cancer Support.

Ethical approval: Multicentre Research Ethics Committee approval for this study was obtained: Ref 05/Q2803/48; Appropriate PCT research governance permissions were obtained.

Conflicts of interest: No known conflict of interest exists.

We acknowledge the assistance given in the design of this study and helpful comments on drafts of this paper by members of the Macmillan Primary Palliative Care Research and Evaluation Collaborative. We are grateful to Shona Agarwal who assisted with data collection and to the GSF facilitators and practice staff who participated.

References

1
Mitchell
G
Reymond
E
McGrath
B
Palliative care: promoting general practice participation
Med J Aust
 , 
2004
, vol. 
180
 (pg. 
207
-
208
)
2
Borgsteede
S
Graafland-Riedstra
C
Deliens
L
Francke
A
van Eijk
J
Willems
D
Good end-of-life care according to patients and their GPs
Br J Gen Pract
 , 
2006
, vol. 
56
 (pg. 
20
-
26
)
3
Aabom
B
Kragstrup
J
Vondeling
H
Bakketeig
L
Stovring
H
Population-based study of place of death of patients with cancer: implications for GPs
Br J Gen Pract
 , 
2005
, vol. 
55
 (pg. 
684
-
689
)
4
Kendall
M
Boyd
K
Campbell
C
, et al.  . 
How do people with cancer wish to be cared for in primary care? Serial discussion groups of patients and carers
Fam Pract
 , 
2006
, vol. 
23
 (pg. 
644
-
650
)
5
Munday
D
Dale
J
Murray
S
Choice and place of death: individual preferences, uncertainty, and the availability of care
J R Soc Med
 , 
2007
, vol. 
100
 (pg. 
211
-
215
)
6
Murray
S
Boyd
K
Sheikh
A
Thomas
K
Higginson
I
Developing primary palliative care
Br Med J
 , 
2004
, vol. 
329
 (pg. 
1056
-
1057
)
7
National Institute for Clinical Excellence
Guidance on Cancer Services: Improving Supportive and Palliative Care for Adults
 , 
2004
London
National Institute for Clinical Excellence
8
Department of Health
The NHS Cancer Plan: A Plan for Investment, A Plan for Reform
 , 
2000
London
Department of Health
9
Department of Health
Building on the Best: Choice, Responsiveness and Equity in the NHS
 , 
2003
London
Department of Health
10
Royal College of General Practitioners
The GP Palliative Care Facilitator Project
 , 
1995
London
Royal College of General Practitioners
11
Ellershaw
J
Wilkinson
S
Care of the Dying: A Pathway to Excellence
 , 
2003
Oxford
Oxford University Press
12
Thomas
K
Caring for the Dying at Home: Companions on the Journey
 , 
2003
Oxford
Radcliffe Medical
13
Thomas
K
Noble
B
Improving the delivery of palliative care in general practice: an evaluation of the first phase of the Gold Standards Framework
Palliat Med
 , 
2007
, vol. 
21
 (pg. 
49
-
53
)
14
Munday
D
Dale
J
Palliative care in the community
BMJ
 , 
2007
, vol. 
334
 (pg. 
809
-
810
)
15
EoLCP
The Gold Standards Framework [Web Page] 2006
  
http://www.goldstandardsframework.nhs.uk (accessed on July 26, 2006)
16
Thomas
K
Developing the Gold Standards Framework. 2nd Meeting of the International Primary Palliative Care Research Group. Society of Academic Primary Care. Keene University July
2006
17
Mant
J
MacManus
R
Hare
R
Mayer
P
Identification of stroke in the community: a comparison of three methods
Br J Gen Pract
 , 
2003
, vol. 
53
 (pg. 
520
-
524
)
18
Moher
M
Yudkin
P
Turner
R
Schofield
T
Mant
D
An assessment of morbidity registers for coronary heart disease in primary care. ASSIST (ASSessment of Implementation STrategy) trial collaborative group
Br J Gen Pract
 , 
2000
, vol. 
50
 (pg. 
706
-
709
)
19
Murray
S
Sheikh
A
Thomas
K
Advance care planning in primary care
BMJ
 , 
2006
, vol. 
333
 (pg. 
868
-
869
)
20
BMA
Focus on the Quality and Outcomes Framework [Web Page] 2006
  
21
Munday
D
Mahmood
K
Agarwal
S
, et al.  . 
Evaluation of the Macmillan Gold Standards Framework for palliative care (GSF) phases 3-6. 2003-2005
Palliat Med
  
2006; 20: 134
22
Crabtree
B
Miller
W
Doing Qualitative Research
 , 
1999
2nd edn
Thousand Oaks, CA
Sage
23
Nadin
S
Cassell
C
Cassell
C
Symon
G
Using data matrices
Essential Guide to Qualitative Research Methods in Organizational Research
 , 
2004
London
Sage publications
(pg. 
271
-
287
)
24
Field
D
Special not different: general practitioners accounts of their care of dying people
Soc Sci Med
 , 
1998
, vol. 
46
 (pg. 
1111
-
1120
)
25
Charlton
R
Palliative care is integral to practice
BMJ
 , 
1995
, vol. 
311
 pg. 
1503
 
26
King
N
Thomas
K
Martin
N
Bell
D
Farrell
S
“Now nobody falls through the net”: practitioners’ perspectives on the Gold Standards Framework for community palliative care
Palliat Med
 , 
2005
, vol. 
19
 (pg. 
619
-
627
)
27
Speed
S
Luker
K
Getting a visit: how district nurses and general practitioners ‘organise’ each other in primary care
Sociol Health Illn
 , 
2006
, vol. 
28
 (pg. 
883
-
902
)
28
Grol
R
Baker
R
Moss
F
Quality improvement research: understanding the science of change in health care
Qual Saf Health Care
 , 
2002
, vol. 
11
 (pg. 
110
-
111
)

Author notes

Munday D, Mahmood K, Dale J, King N. Facilitating good process in primary palliative care: does the Gold Standards Framework enable quality performance? Family Practice 2007; 24: 486–494.