Research Paper - (2008) Volume 16, Issue 6
1King’s College London, Department of Management, School of Social Science and Public Policy, London, UK
2Professor of Organisation Development, Leicester Business School, Department of HRM, De Montfort University, Leicester, UK
3Royal Holloway, University of London, The Management School, Egham, UK
Received date: 31 July 2008; Accepted date: 6 October 2008
Background This paper draws on data from five English primary care trust (PCT) case studies whichformed part of a larger research project that explored the roles and relationships of clinical managers and their colleagues in periods of change within different healthcare organisations.Aims This article uses empirical data to further our understanding of how primary care organisations can successfully implement service improvements.Method Qualitative methods were used to compare across multiple cases. Three methods were utilised comprising semi-structured interviews, document analysis and observation at meetings. Through an iterative process of data coding using the Nvivo data analysis software, final conclusions developed and became more explicit. Data were collected between mid-2002 and 2005. Results Our analysis demonstrates the important influence of context on the change process. The case studies provide evidence of the nature of the relationships between context and progress in organisational change. We identified three interrelated dimensions of organisational context that played a crucial role in the progress or otherwise of service improvement. Conclusion We conclude that primary care organisations need to have three key features in combination to successfully implement service improvements. These are (i) the presence of change leaders, at several levels throughout the organisation; (ii) a coherent change strategy; and (iii) a sound foundation of relationships between managers and clinical professional groups.
change capacity, context, organisational change, service improvement
In this paper, we focus on how healthcare managers and clinicians are able to influence clinical and service developments. Our aim is to develop our understanding of change processes as they unfold in complex organisations and to examine what impacts on an organisation’s capacity to manage changes effectively. The paper draws upon five primary care case studies about the implementation of a National Service Framework for the delivery of diabetes care.[1] This formed part of a larger study about role enactment and change in health care.[2] Our analysis of the empirical data yields important substantive findings on the elements of organisational context critical to innovative improvement. We utilise the data to develop the concept of distributed leadership for change and to theorise on the cumulative processes by which the identified contextual characteristics produce successful organisational change.
Based on a comparative study of organisations in different sectors, Pettigrew and Whipp indicate that there are five core, interlinked factors in implementing successful organisational change.[3] These are environmental assessment, coherence of strategy, leading change, linking strategic and operational change, and human resource management (HRM) assets and liabilities. Dawson[4] and Amis et al[5] develop Pettigrew’s work – typifying the change process as untidy and complex, and not a linear series of events. Based on this process perspective, a more specific literature has elaborated key facets of change processes as they occur within health care.[6–10]
In their analysis, Pettigrew and colleagues characterised the features of so-called ‘receptive’ contexts for change within healthcare settings in terms of their ability to progress a strategic change agenda.[6] They identified eight dimensions that included consistency of strategy, continuity of leadership, involvement of professionals in the process and HR capacity. Later work on innovation develops these ideas and proposes that there are many interactions between features of the context and actors in the change process.[11–13] This implies that conceptually the separation of context from actors and the substance of change is artificial. Context is not merely the backcloth to action, but part of the activity.
An important aspect in the context of organisational change is that of change agents. In health care, Locock et al, interestingly, propose that the term ‘opinion leaders’ would be more appropriate, as their data demonstrate that change agents may both facilitate and impede change efforts.[14] It is evident in health care that clinical leaders can play an influential role, both as the promoters and as the inhibitors of change.[6,13–15]
Attention should also be drawn to the growing body of evidence which proposes that effective change agency in health care depends on collective leadership.[6,16–18]
Methodology
We describe here the methodology of the overall research project.
Research approach: comparative case studies
The approach employed was a predominantly qualitative one, comparing across multiple cases. Qualitativemethods are typically utilised in case study research because of the nature of the ‘how’ and ‘why’ questions under consideration, and the need to thoroughly explore concepts in depth.19 Yin also suggests that case studies are particularly suited to the complex nature of health service systems, which are characterised by continual and rapid change.20 A comparative case study design was used to facilitate the construction of a large-scale database with both internal and external validity.21,22 The approach adopted was an iterative one, since there were some indicators of the characteristics of ‘effective’ change management within complex health and social care settings, but these data were embryonic and fragmented.
This paper seeks to explore the following research question: How can we explain a primary care organisation’s capacity to implement changes effectively?
Pettigrew et al suggest that the first step in beginning data analysis should be to choose the key dimensions under consideration.6 Tracer issues were identified as significant service improvement initiatives enacted within the particular clinical area selected. The tracer issue for the primary care trust (PCT) case study sites was diabetes, due to the rising demand of diabetes services and the changing location of these services from secondary to primary care, and as interesting empirical sites in which to explore theory about context.
Approach to data collection
Three methods were utilised to gather data necessary for this analysis – semi-structured interviews, document analysis and observation at meetings. These multiple data sources addressed a wide range of issues and provided a more convincing and accurate contextual account. Multisite ethical approval for this study was granted, and all standards of confidentiality and anonymity were met. Data were collected between mid- 2002 and 2005.
Semi-structured interviews formed the main source of data in this study and were conducted with 80 representatives across the five PCT case studies. Table 1 provides a breakdown of the number of interviews by case study site and professional categorisation.
Table 1: Interviewees by case study site and professional group
There is some variance between case study sites in the number of interviewees, and this reflects lower representation in some sites and negative responses from some of those who were approached. For instance, there was a greater proportion of non-respondents at the Diabetes 4 site. In this case, the researchers relied more heavily on observational and documentary data sources. Each interview lasted for approximately one hour.
Key organisational documents, such as meeting minutes and terms of reference, strategic planning documents, discussion papers and job descriptions, were analysed to provide a historical narrative of organisational context and a textual indication of the tracer issues and role interpretation.
Sixteen meetings were attended across the five case study sites to gain further insight into role enactment, relationships between professionals and the tracer issues, and to provide further support for the interview and documentary data collected. Table 2 details the range of meetings that were attended.
Table 2: Meetings attended across the PCT case study sites
The variance in the number of meetings attended reflects the disparate focus of each of the sites. Meeting notes were taken regarding the content of the interactions, in conjunction with observations of group dynamics, decision making, attendance and the time devoted to particular agenda items. Observation provided a more ‘authentic’ image of group dynamics compared to one-to-one interviews.6
Approach to data analysis
All of the primary data collected – interview transcripts and observational notes – were coded using the NVivo software package (QSR International). The research team collectively identified a range of preformulated codes, on the basis of the original research questions outlined in the initial research proposal and the subsequent themes for analysis. These analytic themes were reflected in the interview questions, and the initial coding framework generally followed this structure. The coding framework was then developed collaboratively between the principle investigators and the two project researchers, with a continual, iterative discussion regarding the coding framework. Following this coding process, the data were then organised and compressed to draw broad conclusions regarding the research findings. Through an iterative process of data coding, final conclusions developed and became more explicit, and a final report was produced.23
In PCTs, it is the chief executive officer (CEO) who holds formal, executive power and who is directly accountable to the Department of Health, in terms of performance. The CEO sits at the apex of the organisation as amember of the PCT board which is chaired by an independent chairperson. Beneath this board, there is a professional executive committee (PEC) composed of representatives of the clinical professions and management. The PEC chair, who is usually a general practitioner (GP), is now in a more confused and somewhat ambiguous role and their authority is dependent on the PCT’s CEO. Although GPs are not the majority group in a PCT, they continue to be the largest, single clinical group, and the data presented indicate their continued clinical dominance.
Our data display, across the cases, the importance of context for understanding why and under what conditions clinical service improvements may or may not progress. Our argument accounts for the variations in progress on introducing or developing aspects of diabetes services and in meeting the standards and targets set in the Diabetes National Service Framework (D-NSF). Diabetes services were operating in a radically different context, since primary care had undergone a major reorganisation with the creation of PCTs. In addition, demand for diabetic services had grown rapidly, with steep increases in numbers of cases of diabetes, among both the young and the old, and a new set of national standards. Figure 1 provides an overview and tracks the improvement capacity of each of the cases.
Figure 1: Typology of service improvement capacity across diabetes sites. SHA, strategic health authority
Within each box, we attempt to encapsulate the interrelated characteristics that affect the organisational capacity for change and then link them to progress in service improvement. To explain the typology, at the top of each box, the context of each site is given a symbolic descriptor ‘label’. Below this context label is a brief explanation of the title to show the facts that support the symbolic title. Then towards the bottom of each box, we provide a second symbolic title to ‘label’ the nature of the progress that has been made in improving the specific area of care under study in this trust. We stress that this label only refers to progress in diabetes care and does not suggest that sound or appropriate progress has not been made in other areas of care. It should be noted that the horizontal axes (labelled ‘limited change’ to ‘proactive change’) indicate the PCTs’ capacity for change rather than how much change necessarily occurred during the period of field research.
Within a PCT, one key set of relationships is between the chair, the CEO and the medical director (which we have called the triumvirate). We found that in four PCTs these relationships were cohesive and friendly, but in Diabetes 2, this was not the case.Within all five PCTs the other relationships that seemed to be important in the management of change were those between senior managers and hybrids (clinicians including doctors, nurses and professions allied to medicine who are in managerial roles) and those frontline staff providing care within the organisations.
Looking at change in the five PCTs overall, the first point tomake is that, during the period of field research, there appears to have been a low priority for all five PCTs in comparison with other more pressing issues, usually driven by the central government agenda and various government targets. Thus, although many of the PCTs had groups dedicated to the D-NSF, and senior managers were conscious of the need to improve diabetes care, change in the area was limited. What change did occur was fragmented and frequently driven bottom-up by dedicated individual hybrid clinicians. To illustrate our argument concerning the importance of clusters of features of context, the remainder of the results will focus on Diabetes 3 and 2 at the extremes of the typology.
Diabetes 3
At the positive end of the spectrum, Diabetes 3 is an example of an organisation with a supportive and inclusive philosophy. It is led by a cohesive triumvirate that provided a model for others who saw themselves as leading change. The CEO stated:
‘... the whole modernisation agenda has dramatically changed what a CEO does because it is no longer about actually trying to get the best out of an established system, it’s now about changing the way the system works.’ (CEO)
The relationships within Diabetes 3 between senior managers and between managers and clinical professionals were sound and robust. This philosophy of facilitation rather than forceful management and of processes of influencing and persuading were evident throughout the organisation. Clinical managers in Diabetes 3 were very much involved in negotiation, mediation and conflict resolution, building and improving relationships. The expectation was that fellow professionals would always respect your point of view:
‘Whatever you do to influence ... [i]t’s got to be evidencebased and credible ... . if it’s a good one and it’s better than all the rest, you don’t really have much of a problem selling it. [... ] Communication, that’s the key thing. And involvement.’ (diabetes specialist nurse)
The PEC chair in Diabetes 3 saw his role as strategic andwas keen that clinicians had a central role in policy and decision making. Equally significantly, individual clinical professionals also acted as important change agents, absorbing best practice and diffusing this among colleagues through networking and discussing how such changes could be integrated within professions. Several allied health professionals (AHPs) saw their role in Diabetes 3 more in terms of ‘leadership’ than management, and characterised the management style within the PCT as democratic and not dictatorial, which might clash with the professional ethos.
The PEC chair in Diabetes 3 saw his role as strategic andwas keen that clinicians had a central role in policy and decision making. Equally significantly, individual clinical professionals also acted as important change agents, absorbing best practice and diffusing this among colleagues through networking and discussing how such changes could be integrated within professions. Several allied health professionals (AHPs) saw their role in Diabetes 3 more in terms of ‘leadership’ than management, and characterised the management style within the PCT as democratic and not dictatorial, which might clash with the professional ethos.
‘... diabetes ... is seen as ... a poor relation to some of the other bigger areas that have been going longer like cardiac care and cancer care.’
As a result of these features of the context, the implementation of the D-NSF got off to a slow start. This was mainly due, as in other PCTs, to other more pressing priorities compounded by a lack of available management expertise and commitment to lead a project of this scale.
Although the PCT senior management team appointed a new project leader in 2003, the pressure of other ongoing projects meant that progress remained slow. Significantly, among all of the respondents in Diabetes 3, this individual project manager was the one who displayed real change expertise. His/her expertise extended beyond core management and included political skills, coupled to a reputation as an effective change agent. For instance, in restructuring the local diabetes implementation team, subgroups of clinical professionalswere created,which were widely supported as a way of involving those professionals. It was agreed that the combined acute/primary care, local implementation team for diabetes should report directly to the PCT board, the only PCT in our study where this was achieved. By mid-2004, the implementation of the D-NSF had become a higher priority.
Diabetes 2
By contrast, at the more negative end of the spectrum, Diabetes 2 is distinctive amongst our cases for its more ‘managerial’ culture and poorer clinical–managerial relations than the other PCTs. Although the chair and the CEO formed a constructive partnership, their working relationship with the PEC chair was more troubled. Indeed, during the period of the fieldwork the PEC chair resigned, due to frustration with the slow pace of change in general and in diabetes care in particular. He described the organisation as ‘a very bureaucratic environment’. Senior managers described their roles as ‘enabling’, but many decisions were taken by the senior management team (SMT) in relative isolation. Thus the CEO described the role as:
‘... a leader and [to] enable colleagues to work together, therefore my leadership will be exercised on getting people to work together, not telling individual elements what to do but my energies go on, come on, let’s do this together.’
The CEO at Diabetes 2 was keen on developing external contacts and put a lot of effort into cultivating contacts with the wider health economy that includes the local council and the other local trusts. This PCT is seen as a strong agency within the local health economy, even though this is challenged by the local acute trust.
‘... at the moment the PCT is very much trying to be the lead organisation in the health system [ ... ] to an extent it’s been accepted by the mental health trust but not necessarily by the hospital ... so shifting the balance of power hasn’t really actually made it a reality.’ (service director, modernisation)
A key part of the issue facing hybrid managers is how, and through what forums, to engage in decision making in the PCT. In Diabetes 2, one can see that the SMT is the central decision-making body within the PCT, and a high priority is placed on directors attending these twice-monthly meetings. The only link between the SMT and the PEC is the director of public health who sits on both committees. In this PCT, the PEC can be seen as an advisory committee, while the trust board is basically a rubber-stamping forum for decisions that are made at the SMT.
Diabetes 2 was the only PCT of our five PCT cases where there was a marked divide between clinicians (particularly GPs) and PCT management which the PEC chair was unable to bridge. It may be because management was more overt and forceful in Diabetes 2 that relationships with GPs involved more conflict. This resulted in fewer hybrids, which made clinical– managerial bridges fragile in nature. Managers and clinical managers in Diabetes 2 were more likely to acknowledge their managerial roles than in other PCTs, while members of the PEC remained firmly wedded to their clinical identities. Significantly, there were fewer staff, particularly clinical staff, who saw themselves as change agents. This demonstrates the importance of key individuals in bridging roles, but also that the effectiveness of such hybrids will be dependent upon a receptive context that is often established historically.
These data demonstrate the important influence of context on the change process, but more than this, the examples provide evidence of the nature of the relationships between context and progress in organisational change. In all our sites, three interrelated dimensions of organisational context played a crucial role in the progress or otherwise of service improvement. These three features were (i) the presence or absence of change leaders, at several levels throughout the organisation; (ii) a coherent change strategy; and (iii) a sound foundation of relationships between managers and clinical, professional groups. We will take each of these in turn.
We have termed the presence of change leaders across levels of the organisation, ‘distributed change leadership’. The meaning of this term is particular and includes several dimensions. Firstly, it relates to the presence of senior (general and clinical) managers supporting change. This necessitates the presence of senior managers who have both the capacity and the interest to support change. Secondly, distributed change leadership includes the presence of credible opinion leaders at middle range levels in the organisation. These individuals may hold managerial or hybrid roles. Their presence is a crucial element in building bridges between professional groups, across professional boundaries but also in linking the senior management and national priorities to the local workface. This finding strongly reinforces prior research on the pivotal importance of opinion leaders in the change process in health care.13,14 The third component to distributed change leadership is the presence of individual ‘workers’ with a keen interest in the changes and a willingness to engage in change efforts. In our data, these individuals were most frequently individual clinicians, of all professions, doctors, nurses and AHPs who had clinical expertise and acted as change agents. Our results suggest that collective change leadership or a ‘leadership constellation’ 18 at the apex of the organisation is insufficient to implement changes to services. Indeed we suggest that knowledgeable change leaders are necessary at several levels throughout the organisation.
The second and interrelated dimension to emerge from our research findings is the need for a coherent change strategy, communicated throughout the organisation (and sometimes also to interlinked organisations) and debated with staff.
The third core dimension we identified was the need for sound interprofessional relationships to exist as a foundation for service improvements. Without this foundation, the data suggest it is difficult to make any headway.
Therefore, as we have stated, the first critical aspect of our analysis suggests that it is this specific cluster or configuration of characteristics that, in combination and mutually reinforcing each other, form the core facets of receptivity. Wedo not seek to argue that there are no other influential factors identified within the organisational contexts studied. Indeed it may be noted that organisational structures can facilitate or hinder change; finance can also help or hinder the delivery of organisational change, but these factors can be compensated for, if our three contextual features are present. We do not claim that the identified factors are an exhaustive list of all the potential connections, since our current findings will need replication and further testing in other healthcare settings.
Our data suggest that for primary care organisations to succeed in making service improvements, they need to work on a number of interrelated organisational issues. Firstly, they need to explicitly develop and identify change leaders throughout the organisation, who are then well placed to disperse change initiatives throughout what are structurally highly diffused organisations. These individuals must be respected clinicians and managers who are also skilled and capable and can garner support for the service improvement initiatives. Secondly, the organisation must develop a coherent change strategy that the change leaders can then promote to colleagues by a process of discussion and debate. Lastly, a sound foundation of relationships between managerial and clinical groups must either exist or be developed on which to build service improvements. This is a challenging agenda but one that is grounded in good empirical data.
Commissioned, based on an idea fromthe author; not externally peer reviewed.
None.
