'We can't get anything done because...': making sense ...

Original research

‘We can’t get anything done because. . .’: making sense of ‘barriers’ to Practice-based Commissioning Kath Checkland, Anna Coleman, Stephen Harrison, Urara Hiroeh National Primary Care Research and Development Centre, University of Manchester, Manchester, UK

Objectives: To investigate the issues raised by participants as ‘barriers’ to the development of Practice-based Commissioning (PBC) in ‘early adopter’ sites in England. Methods: Detailed case studies of five PBC consortia in three Primary Care Trusts (PCTs). Data collection included interviews with a wide range of respondents (46 in total), including general practitioners, PCT employees, Local Authority employees and patient representatives, observation of many different types of meetings (68 in total), and analysis of documents tabled at meetings and circulated at other times. Results: It has been claimed that progress in developing PBC has been slow. Our respondents articulated a number of factors that they felt were holding them back, which could have been identified as ‘barriers’ preventing change. The issues raised were consistent across our sites (lack of time, resources and personnel, and difficult relationships with the PCT), but observation suggested that these issues arose out of very different organizational ‘sensemaking’, and as a result the apparent ‘barriers’ had different meanings in different organizational contexts. Conclusion: Weick’s concept of ‘organizational sensemaking’ provides a useful framework within which to explore the problems encountered when implementing policy. Observational methods are a powerful tool in understanding sensemaking. The variations in sensemaking that we observed suggest that the use of ‘barrier’ metaphors in descriptions of implementation problems risks homogenizing the portrayal of situations that differ greatly in reality. This implies that top-down or centrally driven solutions to such situations will often be inappropriate. Journal of Health Services Research & Policy Vol 14 No 1, 2009: 20 –26

Introduction Practice-based Commissioning (PBC) was introduced by the UK government in 2004 with the stated aim of engaging frontline clinicians in commissioning health care services for their local populations1 in order to achieve diverse official ends such as reducing health inequalities and providing more efficient, more responsive services, closer to patients’ homes. Prior to PBC, these NHS services would have been commissioned by Primary Care Trusts (PCTs). Early studies2,3 report slow progress Kath Checkland PhD, Clinical Lecturer in Primay Care, Anna Coleman PhD, Research Fellow, Stephen Harrison PhD, Professor of Social Policy, Urara Hiroeh PhD, Research Associate, National Primary Care Research and Development Centre, University of Manchester, 5th Floor, Williamson Building, Oxford Road, Manchester M13 9PL, UK. Correspondence to: [email protected]

# The Royal Society of Medicine Press Ltd 2009

in developing PBC. This paper reports selected findings from a study of the early stages of the implementation of PBC. We used observation and interviews to build up a detailed qualitative picture, paying attention to both positive and negative aspects of the implementation process. During the study, numerous fieldnotes and quotes were gathered that expressed the sentiment: ‘we can’t get anything done because. . .’, and it is this aspect of our findings that is addressed in this paper. A wider report of our overall findings relating to the implementation of PBC has been published elsewhere.4 In this paper we do not address wider aspects of PBC or attempt to evaluate it. The conventional approach to our respondents’ complaints would be to collect the quotes and notes together into common themes and identify these as the ‘barriers’ that were preventing change occurring. We have argued previously5 that this is an impoverished approach to

20 J Health Serv Res Policy Vol 14 No 1 January 2009

DOI: 10.1258/jhsrp.2008.008043

Practice-based Commissioning

Original research

studying change in organizations, with the metaphor of ‘barriers’ serving only to reify and simplify problems that were the product of deeper underlying organizational realities. Far from being ‘barriers’ that could be ‘removed’, the problems that actors identified as preventing change were a complex mix of issues and constructions that arose from social interactions in the organizations studied. We draw on Weick’s6 concept of ‘sensemaking’ in organizations as a conceptual framework within which to develop a more nuanced understanding of how health policies are enacted.

Practice-based Commissioning Under PBC, general practices receive an ‘indicative budget’ with which to commission services for their patients. Although the original official guidance suggested that practices would be provided with individual budgets,7 most practices have joined together into informal or formal groups usually known as ‘consortia’1 which pool these budgets. Official guidance contains little about the development of consortia, resulting in diversity across the country: in a survey, PCTs reported as many as 16 and as few as one consortium in their area, each covering populations from less than 20,000 to more than 250,000 patients.1 PCTs remain legally responsible for the contracting process, but practices (and by implication consortia) are expected to work with their local PCT to re-design services in ways that improve patient experiences and keep them out of hospital. General practitioners (GPs) can put forward business cases to provide services within their practices, or can work with the PCT in wider service re-design involving, for example, the development of outpatient services run by specialist GPs. Incentives to encourage engagement with PBC include a cash payment for agreeing to take part, and the ability to reinvest 70% of any savings made in new services for their patients. Practice-level financial performance data is provided by PCTs, and practices within groups are generally expected to work towards making savings on these by, for example, reducing their prescribing costs and/or hospital referrals. Emergency admissions and outpatient referrals are an obvious target for savings, and consortia across the country are working on schemes to provide alternatives in the community and so cut these costs.4

Sensemaking in organizations Consortia can be thought of as fledgling organizations,6 developing structures and processes to facilitate their work. Weick argues that the central activity within all organizations is what he calls ‘sensemaking’: ‘To talk about sensemaking is to talk about reality as an ongoing accomplishment that takes form when people make

retrospective sense of the situations in which they find themselves and their creations.’8

Thus members of organizations extract cues from the changing environment in which the organization exists. What is seen as significant will vary, and is influenced by previous experiences as well as by underlying values. The action that occurs as a result of these cues both ‘enacts the world’8 and plays a part in determining which cues are noticed. Furthermore, the action that is taken will change the world, influencing which issues are noticed as being important. This will in turn affect how organizations act in the future. Weick posits an ongoing cycle of reaction to the world and action which changes that world, driven by internal factors such as previous experiences, beliefs and values. Hence, the way that the members of a consortium ‘make sense’ of the notion of PBC will not only determine how they act, but will also influence how they see the world around them. If, hypothetically, PBC is seen as a re-invention of the earlier ‘fundholding’ scheme,9 then those involved will concentrate on aspects of the guidance and legislation that facilitate the setting up of small scale practice-level schemes as occurred under fundholding: in Weick’s terms, these are the ‘cues’ that will be ‘noticed’. If these schemes are successful, this view of what PBC is about will be reinforced. If, however, PBC is seen as a more radical tool for the reorganization of services and the re-design of primary care, then groups might think about merging practices, forming companies or competing with the private sector to deliver a wider range of services. Exposure to the wider world of commissioning that this involved would then influence their future beliefs about the nature of PBC. Weick further argues that the process of sensemaking described here is rooted in identity construction. Thus, how GPs act out PBC is not only a technical issue, but also an issue of identity; if ‘we’ see ourselves as a friendly, like-minded group who prioritize shared values, we will behave differently than if we see ourselves as dynamic entrepreneurs. This collective identity is not necessarily a shared identity – complete consensus is unlikely in any organization – but rather represents an identity with which members of the organization feel they can live. Finally, Weick argues that action based upon sensemaking never rests upon a complete understanding of the situation at issue; the best that can be achieved are plausible explanations of the world8 that allow decisions to be made and action to be taken. Thought of in these terms, the issues raised as ‘barriers’ preventing action from occurring can be seen as windows into the underlying sensemaking within PBC consortia. The matters that organizational actors identify as a problem will be partly determined by their overall sensemaking about PBC and their role in the process, and this in turn may allow analysts a more

J Health Serv Res Policy Vol 14 No 1 January 2009 21

Original research


nuanced understanding of the factors affecting implementation and the steps necessary to progress.

Methods The data presented here form part of a wider study investigating PBC in England.1,4 The qualitative part of the study from which the present analysis is drawn comprised detailed data collection in three sites identified from our existing knowledge as ‘early adopters’. A sample of three PCTs (within which five PBC consortia were selected for study) was chosen purposively in order to cover a range of consortia types rather than as representatives of a population. The resulting sample varied widely in terms of size, governance arrangements and ways of working. More details about the sample are given in the project report.4 Within each PCT area, one or more consortia were chosen for study. In the PCT with one consortium which covered the entire PCT area, this was straightforward (Site 1). Elsewhere, recruitment of consortia was guided pragmatically by agreement to participate. In Sites 2 and 3, there were multiple consortia and agreement was eventually reached to study three of them in Site 2 and one in Site 3. Table 1 summarizes the characteristics of the sites and consortia. NHS ethical approval was obtained via the MREC procedure (01/11/06 06/ MRE02/81). NHS research governance approval was sought and obtained from relevant PCTs.

Data collection comprised observation, interviews and analyses of available documentation. We saw interview responses as constructed objects that arise from the interaction between interviewer and interviewee in a particular context. Comparing interview responses with data collected by observation of formal meetings and informal gatherings of actors allowed the development of a more nuanced understanding of the situation under investigation. Sixty-eight meetings associated with PBC were observed across the sites and recorded in fieldnotes. Efforts were made to include meetings of as many different types (including with secondary care providers, rank-and-file GPs and senior PCT staff ) and addressing as many relevant topics as possible. Observation was guided by an overall framework derived from analysis of official documents relating to PBC, but this was not regarded as an exclusive template, allowing more inductive observation of any interesting issues arising or interactions that occurred.10 A total of 46 audio-recorded semi-structured interviews were undertaken with relevant actors across the selected consortia and PCTs. Respondents included GPs involved in PBC, ‘rank-and-file’ GPs not directly involved, practice managers, PCT chief executives, PCT staff engaged in commissioning and appropriate officers of provider organizations which had gained or lost referrals as a result of local PBC decisions. Broad topic guides for the interviews were developed to reflect our understanding of the main issues surrounding the development of PBC, based upon reading of relevant

Table 1 Summary characteristics of cases studied PCT

Consortium

Site 1

Single consortium

Site 2

Consortium 1

Consortium 2

Site 3

GP practices (n) (bands)

Patients (n) (bands)

Total devolved budget for PBC

.250,000

.£100 million

,20

20,000–50,000

Under £50 million

,20

50,001–100,000

Under £50 million

50 –79

Consortium 3

20 –49

100,000– 250,000

£50–£100 million

Consortium 1

20 –49

100,001– 250,000

Under £50 million


Organization and administration Board with majority of elected GPs plus representatives of PCT, local authority and ‘lay’ person Volunteer board, comprising two GPs and a practice manager; PCT-appointed business manager working with the board Managing board consisting of representative (GP or manager) from each practice plus public/ patient representative, social services representative, pharmacist; PCT-appointed business manager working with the board Volunteer board with majority of GPs plus practice manager plus nurse; PCT-appointed business manager working with the board Elected board, comprising equal numbers of GPs and practice managers/nurses, plus attached PCT-appointed manager and administrative assistant


Original research

Department of Health guidance and subsequently modified to allow testing of emerging analytical themes. Documents relating to meetings were collected, along with any other available relevant documents. More than 100 documents such as business plans, discussion documents and performance reports were collected in this way, and, while not subjected to a formal line-by-line analysis, these provided a further point of comparison between what might be termed ‘public’ information disseminated in written form and more private or contingent information discussed orally. We treated the results of this triangulation as a matter of reflection rather than as a simplistic test of validity.10 Interview transcripts and fieldnote data were analysed together, with the assistance of Atlas.ti software, which helped organize the large amounts of data collected and provided a space within which the team was able to work together on the analysis. First level coding11 involved the categorization of responses and incidents recorded in fieldnotes according to a framework developed from careful reading of PBC policy documents. Subsequent analysis comprised more inductive coding, and it was from this phase that the difficulty associated with ‘getting things done’ was found to recur across all our sites and across all categories of interviewee. Emerging themes and theoretical ideas were discussed and refined at team meetings throughout the research, and memos were written and shared,11 allowing precise definitions and use of codes to evolve.

Results and discussion All interviewees were asked about the problems and successes that they and their organizations had encountered in the early stages of PBC. Taken alone, these accounts provide a relatively impoverished view of the process. The strength of our combined methods was that we were not only able to engage in dialogue with participants about what they thought the ‘problems’ were but also to observe in meetings the development of constructions and interpretations of events that came to be seen as ‘problems’. Issues that were identified as problems fell into two main groups: complaints about lack of time and resources; and problems in relationships with the local PCT. However, our detailed observations revealed that, although the issues were similarly labelled, the underlying sensemaking was context-specific. In the sections that follow, examples from interviews and fieldnotes are presented to illustrate these points.

Time and resources In meetings and in interviews, respondents repeatedly raised the lack of time, personnel and expertise as

problems. Unlike fundholding, PBC provides no specific allocation of management funds. PCTs are supposed to provide the management support necessary, and it is suggested that this money should be claimed back out of any savings that are made.7 A variety of solutions was adopted in our sites, including allocation per consortium (regardless of size), per patient registered with the constituent practices and even lump sum allocation from the PCT’s own management budget. In spite of these differences, the problems raised were the same everywhere: GPs do not have time to do the work involved; PCTs are providing insufficient management support and expertise; and the skills available are inadequate – as shown in the following quotes: ‘So, I think we have to have those, you know, that in the system equally well if we are to be a change management organization, if we are to deliver on the targets we are expected to deliver on, we need enough senior managers, project managers even, who can do the work and deliver and we haven’t got that, so I think, to me, there is a real capacity gap.’ (ID 2, PCT employee) ‘Biggest reservation is time. You know, the involvement of a practice that you know, that’s gonna involve time. It’s you know, people think, [they don’t] understand how busy we are in the daytime. . . the surgery, waiting room might be quiet, but it’s a constant flow. . . time is a big issue.’ (ID 33, rank-and-file GP)

Superficially, these ‘barriers’ seem straightforward – any organization requires sufficient people with the right skills to do the work required. However, Weick reminds us that what the ‘work required’ is thought to be depends upon the underlying sensemaking in the organization. ‘Who or what we need’ depends upon what actors think they are trying to do and upon what counts as valid ‘work’ within the organization. While the complaints about insufficient management resources were similar across our sites, the underlying sensemaking differed, and any solution that was to be put in place would need to take account of this.

Example 1: What counts as ‘work’ for whom? In Site 1, the sensemaking framework that evolved among the consortium members saw being a PBC representative on the consortium board as ‘a job’. This sensemaking may have been influenced by the prior experience of two board members working on aspects of commissioning with the PCT during a sabbatical period. As a result, it became an unspoken assumption that PBC was an activity that took time, and that it should be treated as a real ‘job’. During meetings, the possibility of GP board members reducing their commitment to their practices was discussed, although no firm decision was made on this during the fieldwork period.


Original research

This would have involved being paid to work one day a week on PBC, freeing up resources within their practices to employ cover. As Weick suggests, sensemaking that sees PBC as ‘a job’ generates consequences: particularly, it is likely to precipitate an analysis of whether or not one has the necessary skills, and Site 1 spent time discussing their own development needs. This in turn further fed the sensemaking process and generated a belief that in order to achieve their aims they required ‘real’ additional management resources rather than resources that would be reclaimed out of savings. Out of this, some hard-headed negotiation with the PCT was undertaken. In Site 3, by contrast, GPs did not see their PBC role as a serious ‘job’ and tried to fit the work in around their other commitments. This doctor was considering whether or not to continue with the role of board member; tellingly, his partner did not think it was a ‘real job’: (In answer to a question about ongoing involvement in PBC): ‘It’s primarily to do with the practice workload, I would have said was the main reason, but probably a second frustration that I feel that, it is workload I don’t think that, um, my partner sees it as productive work, outside of the work. . . and if you were just going to a meeting once a month, fine, but then you’re asked to go and see this meeting, that meeting, read all these emails, read all this paper work and I just think it is ballooning out of control really. . .’ (ID 63, GP board member)

Furthermore, we found evidence that, rather than simply needing more staff or more time, consortia wishing to make progress needed to consider the nature of the work that staff were being asked to do. For example, in Site 1 the board GPs were aware that engaging the interest and enthusiasm of ‘rank- and-file’ GPs was important. However, they failed to recognize that decisions about who should undertake which task might take into account more than simply who was available, and that important symbolic messages flow from apparently simple task allocations. Thus the board’s sensemaking about the centrality of its strategic role and the need for more staff to do the tactical work caused it to delegate the task of liaising with practices to a junior PCT manager. Several local meetings set up and led by this manager were observed, from which it was clear that delegating liaison to a junior manager caused the rank-and-file GPs to feel that their concerns were not taken seriously. Meetings were awkward and initiatives introduced met with opposition. These initiatives had been discussed and agreed at board level, and the sense made that ‘implementation’ was a routine task that could be delegated. The problem in this case was not that there were not enough staff, but that the staff were asked to do the wrong type of work. In Site 2 it had been agreed that business managers from the PCT would be attached to each consortium.


However, the nature of these roles was not clearly defined. Thus, the three consortia had the same allocation of staff time, even though they were very different in size. The managers were told that theirs was a strategic role, but this was not agreed with the consortia in advance and, without administrative help, the managers ended up performing tasks such as photocopying. They also were unclear about where their responsibilities and loyalties should lie: ‘Um, I think there is a pull for us to be more PCT, um, but there’s also a pull for us to be more PBC, er, so it is a balancing act to be corporate, but you know to try to put forward the ideas and move thoughts of the PBC in PCT meetings, so it is a balancing act and it is a struggle sometimes.’ (ID 37, PCT employee)

In summary, while our respondents all raised issues described in similar language – too little time, too few staff to do the work – it is not particularly helpful to see this as a ‘barrier’ that could be simply resolved by providing more staff. The meaning of the barrier ‘time’ depended upon the sensemaking in each individual context, and solutions would also need to take that sensemaking into account. For example, in Site 1, how the staff were used was as much of an issue as the number of staff and, in Site 3, additional progress might have been more likely if the GPs had re-assessed the nature of their role, starting to see it as a meaningful ‘job’. In Site 2, the problem was as much to do with a failure to think in advance about lines of authority and accountability as it was to do with time or staff numbers.

Relationship with the PCT There was a consistent discourse across our sites that the local PCT was obstructing progress: ‘The PCT itself has not been, has been, um, you know, been many, many and still are, many, many views. Opposition I have to say and I can’t, I think that’s not too strong a word really, but there are many, many people want to see it fail. Lots and lots of people want to see it fail . . . [they’d be] very, very pleased and unless it goes forward they wouldn’t be displeased if it fails.’ (ID 1, PCT employee) ‘Um, yes, it’s got to be to oversee. Er, I would like to see it; it’s got to help facilitate rather than put roadblocks, er, and impede Practice-based Commissioning, I think um, that’s perhaps been a problem to date. . . and I would like to see PCTs being more enthusiastic about Practice-based Commissioning, but I suspect they are concerned about their futures as well and it’s threatening times.’ (ID 52, non-board GP)

However, this unanimity masks quite different problems relating to ongoing sensemaking within PCTs as well as in PBC consortia.



Original research

Example 2: Official support vs action in practice In Site 1, the PCT was officially supportive and had allowed the PBC group a degree of autonomy. However, the management hierarchy ( possibly understandably, given that it retains formal accountability for the decisions made in the name of PBC) retained a degree of concern about this, and felt the need to informally ‘keep an eye’ on activity. This respondent describes the problems that could ensue as a result: ‘So [for example] if you’ve got an initiative which, um, that is being contracted for under this, under PBC and that could be a demand management initiative. The PBC board would find out what it is for and say “right, we will monitor that and make sure it’s delivered”. We will bring the managers in, we’ll check progress and we’ll ensure, we’ll unblock the blockages and will ensure that it’s lead effectively. Now because it’s being managed by a manager who is employed by the PCT, that manager obviously has ultimately a direct line through to the chief executive of the organization. The executive liked the idea of this initiative, wants to find more about it and that finding out more sends an accountability line back and therefore a whole series of meetings is set up completely outside the PBC, which actually finds out about that initiative, checks the progress and effectively does it. So you end up doing the job twice.’ (ID 1, PCT employee)

The senior management described here apparently intended to be supportive of the PBC initiative – ‘keeping an eye’ in order to make sure things ran smoothly. However, the action generated by this sensemaking – an additional series of meetings and possibly divided loyalties for the junior manager concerned – is likely to have impaired the progress of PBC.

Example 3: Failure to align sensemaking between the two organizations The need for adequate information was constantly cited as a problem:

bringing pages of data in graphs, tables and figures. These were discussed at length, and it was frequently agreed that they were either unsatisfactory or inadequate or not in a useable form. The information staff were then asked to return to the next meeting with the data in a different format. This too was often dismissed as ‘not what we want’. In one site, this cycle was observed occurring at eight successive meetings. It was never considered that PCT employees’ sensemaking about the information and its use might have been different from that of the PBC consortium. Thus asking for ‘more’ or ‘better’ information missed the point that unless those using the information and those providing the information had generated a collective ‘sense’ about its use and purpose, it was unlikely that anything would be provided that was meaningful to both groups. In summary, the ‘barrier’ identified by respondents was ‘inadequate PCT support’ and a superficially attractive solution might be for the Department of Health to issue a directive to PCTs to be more supportive of PBC, or for senior PCT staff to do the same to their junior employees. However, our examples demonstrate that the real problem was the failure to align sensemaking across the two different organizations (the PCT and the PBC consortium). The PCT senior manager in the second example was trying to be supportive, but his sensemaking generated additional layers of reporting and governance that were likely to be perceived by GPs as unhelpful. In example 3, the PCT information staff were also trying to be supportive, but the failure to align sensemaking about the purpose of the information they were providing militated against progress. To generate progress, the two different organizations involved should have worked together to align their sensemaking about the overall nature of the task, rather than concentrating on issues that appeared on the surface to be ‘barriers’.

Conclusions ‘So we’d like to be able to think, to plan for savings and freed up resources, but it’s incredibly difficult when we don’t have any budget statements from this year, or an end of year report from last year, and you know these are the frustrations that we are working with at the moment, we just don’t, we don’t have any information that we can manage our budgets on, you know if at the end of the year we are overspent, um, we only get told six months after the event, you know. . . has PBC failed or have the PCT failed? Because at what point can you start to take action to change that and make some remedial actions?’ (ID 34, practice manager)

However, longitudinal observation provided evidence that ‘lack of information’ or ‘wrong information’, while undoubtedly an issue, fails to grasp the complexity of the problem. In all three sites, staff from PCT Information Departments attended board meetings,

This study aimed to follow in depth the development of PBC among early adopters. In all three sites, progress had been made in generating new organizations. However, others have commented on the fact that overall PBC progress has been slow3 and it would have been easy to use our data to generate a list of ‘barriers’ that needed to be ‘removed’ to improve the situation. Weick’s concept of organizational sensemaking6,8 and our previous work5 implied, however, that such lists fail to take into account important facets of organizational behaviour. While ‘problems’ tended to be expressed in similar terms across the sites, the underlying sensemaking leading to those problems emerging was generally different across sites, implying that the solutions required would also differ. The common


Original research

approach to organizational change which focuses on removing ‘barriers to change’12,13 tends both to reify problems and to conceal their deeper nature. In the current study, participants in several different sites used similar labels to describe the problems they encountered, but these labels concealed the fact that the ‘problems’ arose from quite different sensemaking, and, as a result, carried different meanings in different organizational contexts. This suggests that centralized or ‘top-down’ solutions will not work unless local context can be taken into account. Thus, for example, a superficial reading of the data would have suggested that ‘better information’ or ‘more management resources’ were all that was needed to bring about change, whereas detailed attention to sensemaking in each individual organization would be required. It seems likely that these findings are capable of wider application in relation to policy implementation more generally. This study also shows that longitudinal observation provides a powerful method for developing an understanding of how NHS organizations behave in a climate of rapid change. Constraints include time and expertise, but this study suggests that it is a method worth pursuing. Attending meetings over a period of many months gave an opportunity to observe the development of issues over time. As the same items came up on the agenda repeatedly, it was possible to observe how responses changed over time, and to compare responses in interviews with observations. It is unlikely that an understanding of the real nature of the ‘information problem’ could have been obtained in any other way.

Acknowledgements This study was funded from the Department of Health core grant to the National Primary Care Research and Development Centre. We are grateful to all our participants for generously

Practice-based Commissioning sharing their time with us and for making us welcome at their meetings and events.

References 1 Coleman A, Harrison S, Checkland K, Hiroeh U. Practice-based Commissioning: report of a survey of Primary Care Trusts. University of Manchester: NPCRDC, 2007 2 Putting Commissioning into Practice: Implementing Practice-based Commissioning Through Good Financial Management. London: Audit Commission, 2007 3 Lewis R, Curry N, Dixon M. Practice-based Commissioning: from good idea to effective practice. London: The Kings Fund, 2007 4 Checkland K, Coleman A, Harrison S, Hiroeh U. Practice Based Commissioning in the National Health Service: Interim Report of a Qualitative Study. University of Manchester. University of Manchester: National Primary Care Research and Development Centre, 2008 5 Checkland K, Harrison S, Marshall M. Is the metaphor ‘barriers to change’ useful in understanding implementation? Evidence from case studies in general medical practice. J Health Serv Res Policy 2007;12:95–100 6 Weick KE. The Social Psychology of Organizing. Reading, MA: Addison-Wesley Publishing Company, 1979 7 Department of Health. Practice-Based Commissioning: Practical Implementation. London: The Stationery Office, 2006 8 Weick KE. Sensemaking in Organizations. Thousand Oaks, CA: Sage, 1995 9 Glennerster H, Matsaganis M, Owens P, Hancock S. Implementing GP Fundholding. Wild Card or Winning Hand? Buckingham: Open University Press, 1994 10 Murphy E, Dingwall R, Greatbatch D, Parker S, Watson P. Qualitative research methods in health technology assessment: a review of the literature. Health Technol Assess 1998;2:1 11 Miles M, Huberman A. Qualitative Data Analysis: An Expanded Sourcebook. Thousand Oaks, CA: Sage, 1994 12 Grol R, Grimshaw J. Evidence-based implementation of evidence-based medicine. Jt Comm J Qual Improv 1999;25: 503–13 13 Hulscher ME, Wensing M, van Der Weijden T, Grol R. Interventions to implement prevention in primary care. Cochrane Database Syst Rev 2001;(1):CD000362
