Engaging GPs in commissioning: realist evaluation of the early experiences of Clinical Commissioning Groups in the English NHS

Abstract

Objectives

To explore the ‘added value’ that general practitioners (GPs) bring to commissioning in the English NHS. We describe the experience of Clinical Commissioning Groups (CCGs) in the context of previous clinically led commissioning policy initiatives.

Methods

Realist evaluation. We identified the programme theories underlying the claims made about GP ‘added value’ in commissioning from interviews with key informants. We tested these theories against observational data from four case study sites to explore whether and how these claims were borne out in practice.

Results

The complexity of CCG structures means CCGs are quite different from one another with different distributions of responsibilities between the various committees. This makes it difficult to compare CCGs with one another. Greater GP involvement was important but it was not clear where and how GPs could add most value. We identified some of the mechanisms and conditions which enable CCGs to maximize the ‘added value’ that GPs bring to commissioning.

Conclusion

To maximize the value of clinical input, CCGs need to invest time and effort in preparing those involved, ensuring that they systematically gather evidence about service gaps and problems from their members, and engaging members in debate about the future shape of services.

Keywords

GP commissioning GP added value primary care organisation primary care purchasing NHS realist evaluation

Introduction

In 2010, the Coalition Government proposed the transfer of responsibility for commissioning to general practitioners (GPs) working in consortia to be known as Clinical Commissioning Groups (CCGs).¹ The rationale was that closer engagement of GPs in commissioning would ensure: more effective dialogue between primary and secondary care, decision making closer to the patient, and increased efficiency. CCGs were formally established in April 2013 however, there has been little published research about CCGs in general and the ‘added value’ clinicians bring to the commissioning process in particular.

The ‘added value’ that clinicians bring to commissioning is said to include ‘strengthened knowledge of the needs of individuals and local communities …, increased capability to lead clinical redesign and engage other clinicians, …, and greater focus on improving the quality of primary medical care as a key part of a clinically led redesign of care systems”.² An ‘excellent practice’ would have constant clinical focus on improving quality, significant engagement from constituent practices and involvement of the wider clinical community in commissioning.³ However, it is not clear how these aspirations can be or have been manifested in practice.

CCGs were established as statutory bodies in England responsible for commissioning (purchasing) a range of health services from providers. The development of CCGs was deliberately permissive and intended to be ‘bottom up’.¹ CCGs are membership organizations, with individual general practices required to sign up to be members. All CCGs were required to have a governing body with statutory responsibility. Their leadership was to be shared between an Accountable Officer (a GP or a manager) and Chair (clinical or non-clinical). If the Accountable Officer was a clinician, they would need to be supported by another senior manager. CCGs were required to have an audit and a remuneration committee accountable to the governing body. Most CCGs have set up an ‘operational’ body which sits underneath the governing body, a quality committee and other groups. The permissive nature of CCGs’ development has created complexity (Table 1) which is inherent in their internal structure.⁴ This makes it difficult to explore the extent to which GPs could or should be involved.

Table 1.

The complexity and diversity of work done by differing bodies in the CCGs.

	Site A	Site B	Site C	Site D
GB	Assurance and sign-off decisions made elsewhere	Substantive discussion and operational decisions	Assurance and sign-off decisions made elsewhere	Assurance and sign-off decisions made elsewhere
Executive or commissioning committee	Approves strategic and operational issues	Ensures effective commissioning and delivery of the commissioning plan	Have a separate clinical and management team with clearly delineated remit. The remit of the clinical team is to deal with clinical issues while the management team deals with strategic and operational issues	Oversees commissioning activities and review and deliver strategic, operational and financial plans
Quality, safety, finance and/or performance committee	Provides advice and recommendations to the executive group and assurance to the GB on quality of services, clinical effectiveness, safety and patient experience. This committee has a service-specific sub-group which is a joint committee with a neighbouring CCG	Separate the committees into two: 1. Quality and safety committee whose remit is to review and monitor all elements of quality, safety and patient experience 2. Finance and performance committee whose remit is to monitor and review finance and performance plans and achievement	Oversees, understands, reviews and ensures that action is taken for all issues related to quality, finance and performance of services	Monitors various aspects of operations of the providers. They established sub-groups which held meetings with each of the major providers locally. Topics discussed at the sub-group meetings are provider specific
Audit committee	Provides GB with governance, risk management and advice on financial and law compliance
Remuneration committee	Makes recommendations on remuneration, fees and other allowances for employees and for people who provide services to the CCG
Informal groups	A group to delivery of transformation and integration of health and social care services within the district wide area. They have also developed an integrated provider group to enable various providers to engage in service development	A voluntary partnership group which bring together commissioners, local authorities and local providers to deliver a whole system transformation. Under this group, there is an executive level and a programme Board	A forum for commissioners and local providers to exchange their views	A forum to enable high-level clinician-to-clinician discussions

CCG: clinical commissioning group; GB: governing body.

A review of previous forms of clinically led commissioning found that there is no clear definition of clinical engagement in commissioning.⁵ The roles clinicians play is dependent upon the nature of clinical engagement and how much control and influence clinicians have over commissioning decisions. A study of clinical involvement in CCGs found that initial enthusiasm among some GP leaders had started to wane which was highlighted as a problem.^6,7

We established a longitudinal study to explore the development of CCGs. In the first phase (January 2011–September 2012), we followed the development of CCGs from inception to authorization.⁸ The aim of the second phase (April 2013–March 2015) was to explore the potential ‘added value’ that clinicians, specifically GPs, bring to commissioning.⁹ Using realist evaluation, we asked:

What value do GPs add to the commissioning process (outcome)?

In what ways do GPs add that value (mechanisms)?

Under what conditions do GPs add value in the way described above (context)?

We present our findings in the context of previous policy initiatives in the United Kingdom to involve GPs in commissioning, covered in a published review.⁵ Some schemes (e.g. GP fundholding) provided direct incentives for GPs, others provided indirect incentives in that any savings made could be reinvested in other services for the local population. The most direct comparison is with Practice-based Commissioning (PBC), the immediate precursor to CCGs. We used our data to identify the mechanisms and conditions which enable GP ‘added value’ in the commissioning process.

Theoretical framework

Realist evaluation is an approach grounded in realist tradition which focuses on the mechanics of explanation, i.e., not just on whether or not a programme works but on how and why a programme works.¹⁰ It addresses questions about what works for whom, in what circumstances and in what respects, and how? The basic realist formula is: Context + Mechanisms = Outcome (known as ‘CMO triads’).

The requirement to explain why programmes work or not implies an emphasis on the role of programme theory which involves assumptions about how the programme might or is supposed to work. What matters is the way in which participants responded to the programme. Programme theory deals with the ‘mechanisms that intervene between the delivery of programme service and the occurrences of outcomes of interest’.¹¹ In realist evaluation, the emphasis on causal explanation also engages with the idea of mechanisms at work.

Mechanisms are ‘underlying entities, processes, or structures which operate in particular contexts to generate outcomes of interest’.¹² Hence, mechanisms produce outcomes and are made up of individual reasoning (choices) and resources available. Whether or not mechanisms produce the outcomes expected is dependent on combinations of its contextual conditions which enable or constrain the mechanisms.¹³

Methods

Using interviews with both clinicians and managers (n=42) in seven case study sites and a close reading of policy documents (July 2013–January 2014), we uncovered the programme theories underlying the claims made about GP ‘added value’.⁹ We tested these theories against observational data (n = 48 meetings; 111 hours of observations) from four case study sites (selected based on size, geographical area and examples of good practice or significant problems) to explore whether and how the claims were borne out in practice. Observations (January–September 2014) were recorded in contemporaneous field notes and interviews were audio-recorded (with consent) and fully transcribed. Data were stored and managed using NVivo. We attended a wide range of CCG meetings including the Governing Body, executive groups, membership and informal group meetings.

Field notes and associated documents were read repeatedly by the research team for familiarization and discussed at regular collaborative meetings. The analytical process was not necessarily sequential but data were analysed as a set of Context–Mechanism–Outcome statements, i.e., ‘in this context those mechanisms enacted by that actor generated that outcome’. The process was iterative, moving back and forth between observational evidence from different sites, which lead to refinement of the theories.

Results

We identified four programme theories underlying the claims about GP ‘added value’ in commissioning:

Theory 1: GPs’ frontline knowledge about patient experiences would enable them to identify problems and deal with them promptly.

Theory 2: GPs’ frontline knowledge about services would enable them to improve service design.

Theory 3: GPs’ clinical experience and knowledge would enable them to have the authority to speak to other clinicians in ways which improves commissioning.

Theory 4: GPs have a symbiotic relationship with managers, which together is more than the sums of its parts.

Clinical leadership and decision making

Theory 1 (GPs’ knowledge about patient experiences) and Theory 2 (GPs’ knowledge about services)

Our interviewees suggested that GPs’ knowledge about patients and available services would enable them to identify and deal with problems early on and improve service design. These theories share a common theme with all previous clinical commissioning schemes which also rested upon the claim that GPs are better informed about patients’ needs than managers as they have access to practice-level data and direct feedback from patients.^14–16

We identified some of the mechanisms by which GPs may contribute more fully to the commissioning process. CCGs need to ensure that there is a facilitative environment which assures GPs that it is safe and easy to express their concerns, and contribute to or attend meetings. Important contributing factors included good chairing of meetings and a willingness to vary the meeting format to maximize the opportunity for active engagement. For example, Site D decided to use a ‘select committee enquiry’ format in one of their meetings with hospitals to encourage GPs to question and contribute. Good communication was important, particularly given the complexity of CCGs’ internal structures. There needed to be proactive communication with clinicians to ensure that they understood which forums to address their concerns to, which meetings were happening where, and what topics would be covered on the agenda.

CCGs also need to actively seek out the views and experiences of clinicians not actively involved in the CCG on a day-to-day basis. Whilst GPs’ personal knowledge of patients and services was valuable, it needed to be supported with aggregated qualitative data such as that relied upon by managers. There also needs to be significant preparation prior to meetings, which includes giving GPs task-specific information before the meeting.

Our interviewees also claimed that GPs ‘see the whole system’. However, our observations showed that GPs’ knowledge was often pertinent to a particular service and they did not necessarily have insights into a full range of services, nor about how services work in general. Service reconfiguration and a proliferation of hospital providers make it difficult for individuals to understand the full range of services available locally. Our sites recognized this, with many seeking to establish a searchable database with information on the range of available services. Overall, clinical voices were valuable in providing contextual details and information as to whether services were being delivered as intended but they required additional information from managers in order to understand fully the pattern of available services.

The concept of clinical leadership is central to all models of primary care-led commissioning. Under GP fundholding and PBC, GPs were encouraged to engage in commissioning. GPs were the undisputed leaders in fundholding and assumed most of the responsibility of running the scheme including budgeting, contracting and liaising with external stakeholders such as the Health Authority (HA) and local providers.¹⁷ Similarly, most PBC consortia were dominated by GPs and had a degree of delegated decision making power.¹⁴ Under Primary Care Trusts (PCTs) and HAs, by contrast, only a select number of GPs were engaged, either in an advisory capacity or not perceived as influential, with GPs struggling to exert control over decision making.¹⁸ They were used primarily as a sounding board or source of factual information.¹⁹ Total purchasing pilots had no organizational template so there was substantial variety in terms of the nature of interaction between GPs and their constituent HA.²⁰ In practice, projects that were badged as ‘commissioning total purchasing pilots’ had meaningful clinical input (albeit if the input came from lead GPs) and more autonomy and influence over decision making.²¹

Policy rhetoric presented CCGs as different from previous initiatives because GPs would take full commissioning responsibility.¹ Our study shows that GP involvement in CCGs looks very similar to previous clinically led commissioning. Most GP leaders who are actively engaged in CCGs are the usual suspects, who have previously held a leadership role; CCGs had considerable difficulty in enthusing new GP leaders.⁸ However, while previous clinical involvement was limited to those in leadership positions, in CCGs this can occur at different levels of the organization. Some CCGs had separate clinical and management teams while others had a GP majority overseeing the commissioning function. Some CCGs have localities, neighbourhoods or a Council of Members who may or may not be given devolved budgets and responsibility. This makes it difficult to draw general conclusions about GP involvement as the extent to which the new system has enabled better or more robust GP involvement varies in different CCGs.

Clinical involvement

Theory 3 (clinician-to-clinician discussions)

Whilst initiatives such as PBC had involved GPs in the commissioning process,¹⁴ some argued their lack of statutory power had limited their effectiveness.²² Similarly, only 20% of Primary Care Groups involved clinicians from primary and secondary care.²³ By contrast, our participants claimed that CCGs have enabled GPs to become more involved in contracting and made provider clinicians more likely to engage. This suggests that GPs’ clinical experience gives them knowledge and experience which they can use to speak to clinicians outside the CCG in ways which improve commissioning by, for example, challenging existing ways of doing things. GPs’ clinical experience gives them the authority to talk about clinical issues and to challenge providers, if required, in a way that managers could not. In contrast to Theories 1 and 2, in which CCGs are seeking their members’ frontline knowledge to be used in commissioning process, Theory 3 is about engagement with the wider body of clinicians who are external to the CCG.

In spite of attending many pathway development and contracting meetings, in which hospital clinicians were present, we did not observe many instances in which GPs brought their clinical knowledge to bear in challenging their hospital colleagues. In fact, one of our study CCGs had made the decision to only send managers to these meetings as it was not regarded as a good use of GP time. We only observed one instance where hospital clinicians’ behaviour was apparently influenced by the presence of GPs. In one of the quality and performance meetings we attended, it was noted by those present that the relevant hospital’s Medical Director had not been attending the meeting regularly. GPs in the group expressed their disquiet about his absence, suggesting that they regarded it as unacceptable, as there were a number of important issues on the agenda such as serious safety incidents. This challenge has led the Medical Director to attend subsequent meetings.

The fact that we did not observe many instances of GPs engaging with and challenging their hospital colleagues does not mean that these conversations are not happening. It could be that the conversations are happening informally, outside of the formal meetings that we observed. That we rarely observed GPs in contracting meetings suggests that the claims made by those espousing this theory – that clinicians bring a unique and important focus to meetings with providers – was not much experienced in practice.

We identified some of the mechanisms which underlie the successful attempts made by CCGs to achieving clinician to clinician discussions. Site A had been trying to get their members to own the CCG since they were authorized. To achieve this, the CCG decided to invite a hospital providing urgent care to attend the membership meeting to enable GPs to challenge the hospital’s failing performance. However, when the meeting took place, no such challenge occurred. In contrast, in Site D where a similar meeting was convened, GPs had a challenging discussion with the hospital staff. The CCG decided, as a trial, to organize the meeting in a ‘select committee’ style, with GPs as committee members and the hospital representatives as witnesses providing an account of their services. Before the meeting, the chair had primed some GPs in the room with questions to ask. These GPs read out the questions when prompted to do so by the chair, and this was followed by questions from other GPs from the floor. The panel was then given the chance to answer. The meeting turned out to be quite a challenging session, with the panel becoming somewhat defensive in their answers at times. By the end of the meeting, they identified a list of issues to work on as a group (including hospital clinicians, GPs and managers). Hence, careful preparation prior to the meeting and the role of the chair enabled the GPs to have challenging discussions with clinicians from the hospital.

Our study shows that GPs can contribute significantly to commissioning by using their clinical experience to engage effectively with other clinicians but that this requires careful preparation and management and probably only occurs on a limited scale. To engage effectively with the wider clinical community outside the CCG, the formal architecture of the CCG and the operation of statutory authority were not necessary. Some of our sites had established informal groups across the health economy which focussed on a variety of issues including high-level strategy and service development ideas. These groups have no formal or statutory role. The key mechanism underpinning this is that those present should be senior enough within their own organization to make binding commitments on their behalf.

Theory 4 (clinician–manager symbiosis)

Our participants claimed that GPs and managers work together more effectively than they would be able to alone. Managers would formulate and write strategies and business plans while GPs would assist in clinical input and engage with other clinicians. Mechanisms which enable this theory to work include a history of cooperation. Where this history did not exist, careful appointment procedures in which GPs were fully engaged could support the development of new close working relationships. GPs and managers should also recognize that they had different skills and contributions and that they felt able to challenge one another. The status of the CCG as a membership organization was crucial and as the confidence that GP members have towards the GP–manager team working on their behalf. The experience of success and having joint responsibility for programme delivery was important both in developing the close and supportive relationship between the two individuals and in bringing the wider membership along with the process.

This GP–manager pairing was evident under PBC. Most consortia had some kind of managerial support provided by the primary care trust and the extent to which GPs truly led the agenda depended on the dynamics of the relationship between assigned managers and GPs.¹⁴ The assigned managers were facilitative and supportive, acting as enablers to promote clinical leadership. These pairings were also evident in PCTs, although in that case, it was between the Chief Executive and Professional Executive Committee (PEC) chair. The PEC chair was felt to be the most influential member.²⁴ Hence clinical leadership was limited predominantly to one GP who held considerable power within PCTs. The distinctive feature of some CCGs was that these clinician–manager pairings were evident at multiple levels, from the Governing Body to Localities and within different workstreams. It was clear that many of those involved had a long history of working in a particular locality in different roles. A sense of a shared and enduring local commitment seemed to be important in enabling trust. However, in cases where a preexisting relationship did not exist, having the clinicians involved in the recruitment process had been important, as it enabled them to feel that they had chosen their managers.

Discussion

Main findings

The complexity of CCG structures means they are often different from one another with different distributions of responsibilities between the various committees. This makes it difficult to be sure where responsibilities lie within any particular CCG without detailed investigation, and difficult to compare CCGs with one another. The claims made about the value that GPs bring to these processes were broad and idealized. There was a consensus that greater GP involvement was important but it was not clear where and how GPs could add most value in the complex myriad of committees, groups and forums existing in each CCG. It is thus difficult to determine the extent to which GPs could or should be involved. A significant amount of time and money is being spent involving GPs in commissioning at a time when general practice is under extreme pressure. It is important that GPs’ time is utilized wisely hence, it may be appropriate to reduce GPs’ involvement in some forums depending on the functions of that forum. The complexity of a CCG’s internal structure means that different CCGs should adopt different approaches to maximize their GPs’ time.

Our study supported Theories 1, 2, and 4. GPs knowledge adds value to commissioning. However, it needs to be contextualized and supported by proper analytical data. GP involvement in CCGs looks similar to previous clinically led commissioning as most GP leaders who are actively engaged are the usual suspects. However, there is some evidence that active GP engagement runs more deeply within CCGs than it did in previous clinical commissioning initiatives, with GPs involved in a range of locality-based groups. This makes it difficult to tease out the extent and impact of such involvement. We also observed close and effective GP–manager relationships, which enabled GPs’ time to be better utilized. We did not find many instances of Theory 3, with little evidence of regular or routine engagement of GPs with their secondary care colleagues. When clinician-to-clinician discussions did happen, we found some evidence that such discussions are valuable in the way the theory described.

The mechanisms and conditions which enable CCGs to maximize the ‘added value’ that GPs bring to the commissioning process include the following:

The complexity of CCG structures does not necessarily bring many GPs any closer to the decision-making process, and it was unclear where particular decisions would be made. For GPs to bring useful clinical knowledge, CCGs need to ensure that GPs understand the wider context, purpose and expected outcomes of the discussion. There also needs to be clarity at all levels over decision-making responsibilities. GPs also need to be proactive in asking for task-specific briefings.

CCGs have enabled the potential involvement of a greater number of GPs in commissioning processes. However, many CCGs are struggling to ensure that their local GPs feel ownership of the work that is done in their name.

To maximize GP added value, it may be appropriate to reduce GPs’ involvement in some forums depending on the function of that forum.

CCGs are an excellent vehicle for engagement across organizational boundaries throughout the local health economy. However, the formal architecture of the CCG and the operation of statutory authority were not necessary. What is required is senior-level representation with decision-making power from all groups present.

Mutual dependence between GPs and managers commonly observed in PBC was also observed in CCGs. The pertinent difference is that under PBC, these supportive relationships were generally limited to middle managers and the managers involved carried a dual identity, working both for the PBC group and for the PCT,²⁵ whereas with CCGs, this relationship can be seen throughout the levels of the organizations as they are working as part of the same organization.

Much of what is described under CCGs could have been achieved using PBC as a vehicle. However, the scope of activity under CCGs is significantly greater than was the case under PBC (or, indeed, previous clinically led commissioning initiatives), enabling the application of GPs’ knowledge to a broader range of service areas. Our study shows that CCGs do attempt to bring the knowledge and views of frontline GPs into their work but this requires explicit attention to processes and considerable preparatory effort. Online Appendix summarizes the programme theories and the CMO which enable the theories to work in practice.

Implications

This study has identified some of the mechanisms and conditions which enable CCGs to maximize the ‘added value’ of involving GPs in the commissioning process. CCGs face a very challenging future with the ongoing shift of responsibilities for primary care co-commissioning from NHS England which started in April 2015. This brings with it significant challenges for CCGs, First, CCGs taking on responsibility for commissioning primary care services will be effectively commissioning themselves, raising issues of conflicts of interest. To address this, CCGs have established primary care co-commissioning committees which exclude the majority of GP members. This begins to dilute the principle under which CCGs were established, of bringing GPs into the centre of commissioning activity.

Second, with all CCGs encouraged to take on full responsibility for primary care co-commissioning, the performance management of member practices that they will be required to undertake could potentially threaten buy-in and engagement with wider commissioning activity.

Third, there is a growing policy momentum in favour of a population-based approach, with budgets pooled across a geographical area. This will require commissioners, GPs and secondary care provides to work closely together. Current moves towards the devolution of responsibilities to geographical areas bring further complications, with blurring of lines of accountability and responsibility. These challenges point to a need for CCGs to be: adaptable and flexible, alert to the changing environment and deeply engaged with their colleagues across organizational boundaries. Co-operative service redesign across a health and care economy is likely to occupy much managerial and clinical time and effort. Our evidence suggests that maximizing the value of clinical input into this will require CCGs to invest time and effort in preparing those involved, to ensure that they systematically gather evidence about service gaps and problems from their members, and to engage members in debate about the future shape of services. Doing this whilst simultaneously performance managing member practices will be a complex and difficult task.

Footnotes

Acknowledgements

The study formed part of the programme of the Policy Research Unit in Commissioning and the Healthcare System (). The views expressed here are those of the researchers and do not reflect the position of the Department of Health.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Department of Health Policy Research Programme.

Research ethics

The study received ethical approval from the University of Manchester Research Ethics Committee.

References

Department of Health. Equity and excellence: liberating the NHS (White Paper), London: Stationary Office, 2010.

NHS Commissioning Board. Clinical commissioning group authorisation. London: Stationary Office, 2012.

NHS England. Draft framework of excellence in clinical commissioning: for CCGs. Leeds: NHS England, 2013.

Checkland

Mcdermott

. Complexity in the new NHS: longitudinal case studies of CCGs in England. BMJ Open 2016; 6: e010199–e010199.

Miller

Peckham

. What happens when GPs engage in commissioning? J Health Serv Res Policy 2016; 21: 126–133.

Naylor C, Curry N, et al. Clinical commissioning groups: supporting improvement in general practice? London: King's Fund, 2013.

Holder

Robertson

. Risk or reward?, London: Random House, 2015.

McDermott I, Checkland K, et al. Exploring the GP “added value” in commissioning: what works, in what circumstances, and how? 2015, Manchester: PRUComm, 2015.

Checkland K, Coleman A, McDermott I, et al. Exploring the early workings of emerging Clinical Commissioning Groups. Manchester: PRUComm, 2012.

10.

Pawson

Tilley

. Realistic evaluation, London: SAGE, 1997.

11.

Weiss

. Theory-based evaluation: past, present, and future. New Dir Eval 2007; 114: 68–81.

12.

Astbury

Leeuw

. Unpacking black boxes. Am J Eval 2010; 31: 363–383.

13.

Tilley

. Demonstration, exemplification, duplication and replication in evaluation research. Evaluation 1996; 2: 35–50.

14.

Coleman A, Checkland K, et al. Practice-based commissioning: final report. Manchester: National Primary Care Research and Development Centre, 2009.

15.

Wyke

Abbott

. Developing primary care in the new NHS, London: King’s Fund Publishing, 1999.

16.

Dixon

Goodwin

. Accountability and total purchasing pilots, London: King’s Fund, 1998.

17.

Surender

Fitzpatrick

. Will doctors manage? Policy Polit 1999; 27: 491–502.

18.

Locock

Regen

. Managing or managed? Experience of general practitioners in English Primary Care Groups and Trusts. Health Serv Manage Res 2004; 17: 24–35.

19.

Bate

Donaldson

. Managing to manage healthcare resources in the English NHS? Health Policy 2007; 84: 249–261.

20.

Mays

. Total purchasing: a profile of national pilot projects, London: King’s Fund, 1996.

21.

Malbon

Goodwin

. What were the achievements of Total Purchasing Pilots in their second year (1997/98) and how can they be explained?, London: King’s Fund Publishing, 1999.

22.

Curry

Goodwin

. Practice-based commissioning, London: King’s Fund Publishing, 2008.

23.

Audit Commission. The PCG agenda. London: Audit Commission, 2000.

24.

Smith

Regen

. National evaluation of general practitioner commissioning pilots. Br J Gen Pract 2000; 50: 469–472.

25.

McDermott

Checkland

. Who do we think we are? J Health Organ Manag 2013; 27: 4–23.