How do they manage? A qualitative study of the realities of middle and front-line management work in health care

Article history

The research reported in this issue of the journal was funded by the HS&DR programme or one of its proceeding programmes as project number 08/1808/238. The contractual start date was in January 2009. The final report began editorial review in July 2012 and was accepted for publication in January 2013. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Copyright statement

© Queen's Printer and Controller of HMSO 2013. This work was produced by Buchanan et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

Things can only get different

You'll hear people say ‘management’ in inverted commas. And I'll say, but you all manage. I think that there's always been quite a hierarchy in the NHS. And I do not believe we've done as much as we can to break that down. I want everyone to take accountability and responsibility.

Modern matron, Netherby hospital

How do hospital managers in the NHS handle the demands of a constantly changing service? How do managers affect the quality of patient care and clinical outcomes? Patient safety is a national priority, but changing working practices following serious incidents can be problematic. Why? We know surprisingly little about the work experience, practices and attitudes of hospital managers, who are key to implementing local strategy and national initiatives. However, when things go wrong, this is the group that often takes the blame. This study seeks to build on what we know about management, change and leadership, and relate this to current trends.

This chapter describes the research questions and aims of this study, explains the background to the project and provides an overview of the research process and methods. This project began in 2009. By 2012, however, the service had changed in a number of significant ways. Two events in particular had an impact on acute hospital management roles during this period.

Research questions

Interviewer: What outputs would you like to see from this project?

Respondent: I want to understand what we can do to support and help those in increasingly challenging roles, because we need to get it right. Otherwise, they will do it badly or won't do it. I want to gain a better understanding of their motivation, so that we can help provide meaningful careers.

Senior executive, Greenhill hospital

This project addressed three sets of questions:

1. Realities: What are the pressures and demands facing middle managers in health care? What are the implications of these trends?

2. Changes: What roles do middle managers play in implementing changes? How are changes arising from serious incidents implemented, and how can this process be improved?

3. Contributions: How does management practice affect clinical and organisational outcomes? What factors influence management contributions to performance? How can the components of an ‘enabling environment’ for management work be assembled and sustained?

Why middle management?

The ward leaders – sisters and charge nurses – make or break the hospital's reputation. We have over fifty wards on two sites. Each has a team of twenty to thirty staff operating 24/7. It's where the costs arise, and where patients and visitors make judgements about standards, depending on staff approach and discipline. That's where the key business of patient care is.

Senior executive, Greenhill hospital

The NHS has since the mid-1980s focused attention on chief executives and trust boards. One manifestation of this was the Leadership Qualities Framework. 27 When it was first published, the principal investigator for the current project was working with a hospital in the Midlands. The human resources director was running management development sessions for new clinical directors and business managers – a group of ‘senior middle’ managers. She telephoned the Leadership Centre (part of the NHS Modernisation Agency) and requested 20 copies. She was asked about the use to which these would be put. When she explained, she was told that copies could not be provided. Why not? Because the framework was designed for board-level directors and not for less senior staff.

This anecdote is symptomatic of what appear to be deep-rooted attitudes towards leadership and management in the NHS. Since 2009 there has been a National Leadership Council. There was no National Management Council. There was an elite top leaders programme for those in roles deemed to be ‘business critical’. There was no top managers programme. The NHS had developed several leadership competency frameworks. There were no management competency frameworks. An NHS Leadership Academy was launched in November 2011. There was no Management Academy.

The updated version of the Leadership Qualities Framework28 does suggest that the framework applies to staff at all levels – except for two of the seven domains, ‘creating the vision’ and ‘implementing the strategy’, which are the preserve of ‘a relatively small group of people who hold designated positional roles, and are required to act as leaders in formal hierarchical positions. These two domains focus more on the contribution of individual leaders rather than the general leadership process’ (p. 8). If this ‘general leadership process’ means middle managers, then this implicit division of leadership labour ignores two decades of compelling research evidence. 29,30

One consequence of the focus on senior leadership is that less is known about the roles, experiences, contributions and motives of middle managers. 31 The presumption that their contributions are less ‘business critical’ has passed unchallenged. The Service Delivery and Organisation (SDO) research programme had funded previous studies on leadership, organisation cultures, performance management and service reconfiguration. Those studies, however, did not explore directly the work of middle managers. There was, therefore, a pressing need for research into the challenges facing management in health care and this was recognised as a priority theme for research. 32

Have conditions changed so dramatically since 2007 as to render the findings from this project obsolete? On the contrary. The impact of institutional and organisational context remains central to our understanding of management roles and contributions. For acute hospitals, change agendas are complex and challenging, and middle managers are key to implementation. Although other issues have surfaced and generated additional research questions, the challenges that the NHS has faced since 2007, and will face from 2012 onwards, have only reinforced the importance of this theme.

Research methods

This project used a multimethods collaborative design involving set-up interviews, focus groups, a management survey, documentation from participating sites, serious untoward incident narratives based on documentation and interviews, and management briefings (see details in Appendix 8). The aim throughout was to engage participants as co-researchers, in formulating the original proposal, in survey design and in selecting case incidents. The project stages are summarised in Table 1.

Names, ranks and numbers

We want to turn this into a great organisation, a great place to work. We need to engage people or lose talent, especially in general management roles. Retaining talent in management roles is going to be a problem. They could just walk away to other sectors.

Director, Greenhill

There is a widespread perception that the NHS is bureaucratic and overmanaged by ‘pen-pushers’ and ‘grey suits’. 31,41,42 Management costs were thus a popular target for politicians during the life of this project, particularly approaching the general election in May 2010. 43,44 In exploring the NHS management population, this chapter reaches four conclusions:

1. Evidence suggests that the NHS is undermanaged, despite claims about disproportionate increases in management numbers, and the composition of the NHS workforce is stable.

2. The proportion of staff with managerial responsibilities is much higher than suggested in official statistics, which do not recognise many middle managers or those whose roles combine clinical and managerial responsibilities.

3. Acute trust management includes ‘pure plays’ with managerial roles and ‘hybrids’ with combined clinical–managerial duties; hybrids outnumber pure plays by four to one.

4. The obsession with management numbers and costs overlooks more significant issues concerning management capacity and the contributions that middle and front-line management make to clinical and organisation outcomes.

Pure plays and hybrids

Other evidence shows that the official figures underestimate the numbers of those with either a ‘pure’ management role or a ‘hybrid’ role combining clinical duties with managerial responsibilities:

• pure plays: roles that are wholly managerial

• hybrids: roles that combine clinical and managerial responsibilities.

The NHS Information Centre data do not count as managers any clinical and scientific staff holding hybrid managerial roles; however, if hybrids as well as pure plays are counted, then the proportion of NHS staff in management positions is not 3% but around 30%.

An overmanaged service?

The NHS may be viewed as undermanaged, with only 3% of its 1.4 million employees in management roles. Knowing that at least one-third of staff (in acute trusts) have managerial roles, can we conclude that the service is overmanaged? This raises the question of management capacity.

Management capacity can be defined as the ability of the function as a whole to respond effectively to the demands placed on it at any given time. The management agendas of the trusts collaborating in this study were lengthy, complex and fluid (see Chapter 4). Different profiles of pure plays and hybrids may be equally effective in the context of different local conditions, past histories and current agendas. It is not possible to assess capacity simply by comparing numbers. Fewer pure plays may mean lower management costs, but that could increase the burden on hybrids, reducing time for patients, professional development and innovative service improvements. That may in turn lead to the appointment of more clinical staff to compensate, cancelling the cost advantage of having fewer professional managers.

With regard to the ‘3% versus 30%’ question, at least four other factors impinge on the management capacity of an acute trust. First, ‘the other 30%’ are not all FTEs; most are hybrids. One senior hybrid asked about the mix of clinical and managerial responsibilities in nursing roles replied:

Head nurses are ‘50–50’. They're responsible for the managerial nursing aspects in their directorate. But I hold them accountable for nursing professional issues as well. For a matron, I would expect that to be very much more clinical. And probably it should be 75–25, clinical – managerial. For a ward sister, the ideal is that they should have two days a week where they can deal with the managerial elements of their role, but it's also about making sure standards are adhered to. So probably a 60–40 split.

In your judgement, are some of those staff spending more time on their management responsibilities than they should?

No. I would say it's the other way around. They don't get enough management time.

Director, Netherby

Interviews with modern matrons elsewhere suggest that this balance of responsibilities varies from trust to trust, and fluctuates over time, depending on the management agenda.

Second, despite our previous definition, some ‘pure plays’ have specialist non-managerial responsibilities and so they are not necessarily each a ‘full’ FTE. For example, one manager commented as follows about her managerial specialist responsibilities:

Question: Is my profession ‘manager’ or ‘trainer/coach’? Answer: Trainer/coach. Therefore, when answering what percentage do I ‘manage/lead’ versus what is ‘practical/specialist’, I'd say 70–30. I do have a practical/specialist role I fulfil that is not management. There is clear water between the two, and I suspect that many others will say the same. Few of them will see themselves as NHS managers wholly, rather a specialist first, in a management role, but their loyalties are often to the profession they trained hard in.

Senior manager, Greenhill

Thus, 100 pure plays does not necessarily equate to 100 FTE managers. Although no longer employed full-time in a clinical capacity, many pure plays have a clinical background (such as nursing) that they bring to their relationships with clinical staff, and use to inform decisions and judgements. Some maintain their registration, and ask their workforce information offices to code them with their profession in the annual census figures, and not in the G matrix as managers.

Third, most hybrids have had limited management education, beyond short courses. Pointing to the piecemeal ‘on the job’ management development of medical staff, Gillam53 argues that ‘the idea that all doctors can just “manage” is hopelessly naive’ (p. 1). Short courses include The King's Fund management and leadership for clinicians course, which runs over 5 days, and Leading an Empowered Organisation (LEO), which was a 3-day course (now discontinued) for staff (including nurses) in management and supervisory roles. A master’s degree in business administration takes a year of full-time study to complete; few clinical staff have the desire, time or financial resources to obtain this qualification. During this project, Clearview, Greenhill and South Netley implemented their own leadership and management development programmes. But these still amounted to more ‘short courses’, albeit intensive and tailored. The management capabilities of many hybrids are thus dependent on limited educational input and considerable on-the-job learning:

We don't have any managerial training and lots of work is left to us. I expect that in the future we will be left to make more decisions. There is more workload. You have to do appraisals, you are asked to create a new policy and deliver presentations and you are expected to do more.

Mixed focus group, Greenhill

I've been in post for three years, and I want promotion. Are they developing me into a matron's role? There is no ongoing development. Ward managers are just left to get on with it. You are left on your own. Find your own development. The one-to-one meetings that we have with more senior management are all about operations and finance.

Ward sister, Netherby

Fourth, many hybrids do not view themselves as managers or they use the label reluctantly. In 2010, ward managers at Netherby voted to change their job title – to ward sister. Other examples included:

I've been a consultant nephrologist since 2003, and I've been involved with the emergency department project since 2005. The role of clinical lead is managerial if you want to call it that. In the emergency department project, this has involved structure change. But ‘management’ has connotations among clinicians relating to stick not carrot. Managers are people who withhold resources rather than try to improve services. My role is more of a leadership and organisation development role, much wider than management.

Clinical director, Greenhill

I would describe my role as clinical leadership. I'm not nursing, not hands on. I am not at all managerial. My role is professional leadership. But this does influence behaviour, practice, change, service improvement, so managerial in those senses. The senior staff on the wards have managerial responsibility; ward sister, charge nurse are front-line managers. Matrons and head nurses are middle management. I have a manager role, but I have professional accountability, so the emphasis of my role is with professional leadership.

Senior nurse manager, Netherby

These participants were more comfortable describing themselves as ‘leaders’. Before fieldwork commenced, we were instructed by a senior manager at South Netley to change the subtitle on the project documentation, from ‘middle and front-line managers’, so that this would appear as a study of ‘managerial and clinical leads’. These views reflect a trend in the public sector to value and promote leadership at the expense of mere management. 54,55

Mintzberg4 offers a straightforward resolution to the ‘leadership versus management’ debate. Asking if one would like to be managed by someone who doesn't lead, or led by someone who doesn't manage, he concludes that, ‘We should be seeing managers as leaders, and leadership as management practiced well’ (p. 9, italics in original). With a similar view, and using ‘managers’ and ‘leaders’ interchangeably, Quinn et al. 56 argue that, ‘we want to distinguish between individuals who happen to have management positions and individuals who truly display leadership in their management of others’ (p. 333). The evidence from this project demonstrates that, if the distinguishing features of leadership concern designing, inspiring and driving change, then middle management roles in the NHS are indeed a blend of management and leadership.

To illustrate how pure plays and hybrids populate a hospital management structure, Figure 1 shows the organisation chart for the medicine division at Clearview. This division had 350 beds in 16 wards, employing 1200 staff (headcount). The division's budget in 2010–11 was over £80M, with target savings of around £5M. The organisation chart identifies 21 pure plays and 80 hybrids. In other words, as indicated previously with regard to what appears to be the typical balance of roles at trust level, the ratio of hybrids to pure plays in this division was around 4 to 1.

FIGURE 1.

Clearview Hospital medicine division organisation chart (October 2011). a, Managers with clinical backgrounds. AHP, allied health professional.

Finding the front, defining the middle

For the purposes of this project it proved difficult to define unambiguously ‘the front line’ in relation to ‘the middle’. Definitions provided by various commentators suggest that this should not be a problem. However, three considerations apply. First, in acute care, the management ‘front line’ is mobile. Second, the hybrids who occupy this ‘fluidic space’ have onerous managerial responsibilities that influence quality of patient care, as well as organisational performance. Third, the distinctions between roles in this ambiguous zone are more important than whether they are defined as ‘middle’ or ‘front line’. We have thus dropped the phrase ‘front line’ from discussion. The term ‘middle management’ relates to anyone with managerial responsibilities who does not sit on a hospital board of directors. Distinctions between roles will be drawn with reference to specific job titles.

There are particular reasons for viewing ward sisters as middle, not front-line, managers. First, their roles are similar to those of matrons and senior nurses, to whom they report. Second, they are ‘mini-general managers’, controlling a budget, overseeing the development and morale of nursing staff, supporting medical staff, delivering service improvements and supporting and improving the patient experience. 57 Third, the ‘manager–subordinate’ relationship they have with other nurses is blurred, as staff nurses routinely ‘step up’ to cover for ward sisters in their absence. Fourth, with their operational experience, they act as intermediaries between senior management and day-to-day patient care. 58–60 Finally, it has been suggested that management regulation should not stop at board members, but should include ward managers. 61 The ward sister's experience is a middle management experience.

Conclusions and implications

The nature and significance of context

The NHS, like any other organisation, is embedded in a historical, sociocultural, economic and political context which shapes the norms, values and expectations that in turn influence the structures and processes of the health-care system. 65 The nature of, and relationships between, the dimensions of the institutional context are critical to understanding the demands that are placed on the management function, and the changing nature of managerial roles.

Context is not simply a stage on which action takes place. The context of an organisation is fluid and dynamic, influencing and in turn shaped by organisational events. 66 These inter-relationships do not necessarily generate a state of constant flux and indeterminacy. The institutional context is also responsible for reinforcing and perpetuating organisational characteristics, for maintaining patterns of continuity. For example, the so-called postmodern trend for flat, fluid, agile organic structures that was supposed to sweep away traditional, rigid, slow-moving bureaucratic structures has barely touched the NHS in its 60-year history. 67

Six dimensions of the context in which the NHS operates are identified (Figure 3). These concern the history of the service, the contemporary business of health care, governance, regulation, finance and the role of the media. These dimensions overlap; discussion of regulatory regimes could equally belong with ‘governance’, and discussion of health care as a business with ‘finance’. The overarching aim of this discussion, however, is to explore relationships between the institutional context and middle management roles. The main conclusions from this analysis are:

• History dimension. Although subject to constant reorganisations, core features of the hierarchical, centrally controlled professional bureaucracy appear to be relatively stable. Rules, regulations and slow decision processes can be beneficial, offering predictability and consistency, but can also impede innovation and rapid change. The current regulatory regime is a legacy from a series of ‘high-profile’ failures, the most recent concerning events at the Mid Staffordshire NHS Foundation Trust (‘Mid Staffs’). The global financial crisis that began in 2008 led to ongoing national austerity budgeting and to pressure on health care to make major efficiency savings, of approximately one-fifth of the annual spend of the service.

• Business dimension. The private and public health-care sectors are now closely intertwined. In hospitals, clinical services are structured and run as ‘businesses within the business’. Clinical and managerial staff are encouraged to adopt a more commercial orientation and to develop new business models, managerial and process innovations, inventive uses of new technologies, public–private collaborations and competitive strategies. A key challenge concerns balancing efficiency and productivity with quality and safety of patient care. The NHS is also expected to generate revenue from innovations and exports.

• Governance dimension. The government elected in 2010 embarked on a complex, rapid and controversial reorganisation of NHS structures. Some existing bodies were abolished (strategic health authorities and PCTs) and new ones were created, including a National Commissioning Board and local clinical commissioning groups. Policy to cut management costs generated considerable resentment. The promised liberation, autonomy and empowerment were difficult to spot, and ‘micromanagement’ continued.

• Regulatory dimension. Providers answer to many regulators, auditors, inspectorates and accreditation agencies (RAIAs), whose information demands overlap and can be burdensome. Regulation has improved aspects of performance, but has not prevented systemic failures in care. The regulatory regime extends to innovation, where the call for a more creative, commercial approach is accompanied by a compliance framework, standardised efficiency measures, fines and an implementation board. In the wake of the Mid Staffs inquiry, some form of regulation of health-care management was expected to emerge during 2012–13.

• Finance dimension. ‘The Nicholson challenge’ to find £20M efficiency savings between 2010–11 and 2014–15 was described as a ‘NHS recession’, a ‘funding ice age’ and a ‘perfect storm’. Trusts from 2010 had to implement CIPs generating recurrent savings of 3–4% per annum, raising anxiety about job security among all staff including management. Although cutting costs and improving productivity were over-riding issues, these actions had to be balanced with the priorities of improving the quality and safety of patient care. Middle managers in particular felt the need for financial management skills.

• Media dimension. The press ‘rules of production’ mean that sensational, dramatic ‘bad news’ stories are more likely to be reported than good news. The constant ‘bad press’ influences public perceptions of the service and its staff, and also affects government policy. Commentary supportive of management is rare; poor management is typically blamed for system faults and failures, and a negative stereotype of managers as costly pen-pushing bureaucrats prevails, potentially inhibiting motivation and recruitment.

FIGURE 3.

Dimensions of the institutional context of the NHS.

The six dimensions

The management dilemma

The features of the institutional context of NHS management work are summarised in Table 13.

At the heart of this analysis sits a management dilemma. Providers are encouraged to operate as commercial enterprises, and to be entrepreneurial, innovative and revenue generating. But they are expected to do this in the face of a burdensome regulatory regime in which failure to comply with central controls can lead to financial penalties and senior staff job losses. This has not deterred many significant innovations. But this dilemma may be an insurmountable barrier to the ‘disruptive innovation’ that may be required by a traditional service in a rapidly changing world. 112

The long ‘to do’ list

This chapter explores how the local organisational context shapes the nature of management work. First, the six acute trusts are profiled. Second, the implications of service-line management structures are considered. Third, the management agendas of two trusts are described. Finally, the implications for middle management are explored. The analysis of the institutional context in Chapter 3 exposed a tension between innovation and compliance. This chapter exposes a further tension, between policy to reduce management numbers and adding to management agendas. Information for this chapter comes from trust websites, other hospital information and set-up interviews. The conclusions are:

• these are large, complex organisations, often one of the largest employers in their local area, some with histories reaching back to the 18th and 19th centuries

• most have ‘legacy issues’ related to old buildings and computer systems, with the latter frequently being incompatible with one another and inadequately specified to provide appropriate and timely business information

• clinical directorate structures, dating from the 1980s, were being replaced by service-line management, with divisions run as ‘businesses within a business’, bringing autonomy, transparency and accountability, but encouraging ‘silo’ working, discouraging information sharing, generating insecurity and creating a lack of clarity with regard to accountability

• the scale of the financial and operational challenges was such that three trusts felt that it necessary to recruit external management consultancy support for their ‘transformation programmes’

• those challenges focused attention on organisation culture and on leadership and management skills and styles

• acute trust management agendas are long and complex, with most issues involving significant organisational change

• middle managers are subjected to ‘multiloading’ (dealing with many different activities) and ‘perpetual loading’ (working at capacity, always economising), the standard responses to which (halt less essential activity, focus on a small number of key issues) are not available

• capacity (see Chapter 2) is a critical management issue, and a priority for further research.

Trust profiles

To indicate the scale and complexity of the trusts participating in this study, and the scope of the hospital management task, Tables 14 and 15 profile the attributes of the two foundation and four non-foundation trusts respectively.

Changing structures

The regulator, Monitor, was encouraging established and aspiring foundation trusts to adopt service-line management methods,113 and hospitals were introducing new organisation structures, with new management roles and responsibilities. This involved establishing autonomous clinical divisions operating ‘businesses within a business’, with clinical staff taking the lead on service development. Each business unit thus had clearly identified resources, including staff and support services.

In 2009, for example, Netherby had a traditional clinical directorate structure, with a separate management team for each of the 18 directorates. The 10 clinical directorates were each managed by a triumvirate including a clinical director, directorate manager and head nurse. An external management consultant working with Netherby in 2011 was asked about the strengths of this structure: ‘I'm struggling to think of any strengths of the current organisation. Managers seem to be dedicated and the pay is good.’ Turning to weaknesses, he observed:

The hierarchies are too deep. There are no clear lines of responsibility and accountability. The information to inform decisions is poor. Business cases are based on selective information. There's a lot of flag waving around ‘clinical standards’ and ‘patient safety’. One-third of the staff are in the ‘administrative and secretarial’ category. How much energy do we need in those areas? There's a lot of bureaucracy. We need to understand non-value-adding activities. What are the consequences of not doing some of these things? Do we know how much this is costing? At the moment it's all anecdotal; we need evidence. We're short of change agents who possess ‘Rottweiler tenacity’. And there's too much politics.

One director observed that issues were ‘passed up the tree’ and escalated unnecessarily. Accountability and performance management were lacking. There were too many directorate managers and management layers, and structures were not consistent across services. In 2011, Netherby introduced a service-line management structure. Two care groups – medicine and surgery – were created, each with a group director, lead nurse and two general managers. The clinical directors became medical service leads, each supported by a service manager. This management reorganisation would generate savings of between £2M and £3M per annum.

For Monitor, the benefits of this approach included the empowerment of clinicians, efficiency and productivity, local ownership of budgets, ‘seeing the big picture’ and accurate patient-level costing. According to matrons at Greenhill, however, these structures were problematic.

Greenhill introduced a service-line structure just before this project began, and the matrons argued that the new ‘independent’ divisions had entrenched the trust's ‘silo mentality’. An operations manager at Clearview described the now hostile resource discussions between their divisions as ‘Reservoir Dogs’ moments (from the movie in which leading characters threateningly point their guns at each other). The Greenhill matrons argued that divisional silos had other problems:

We used to know what was going on in the other areas or divisions. And we used to share good practice. But not now. As modern matrons, we have to take on the role of duty lead nurse (DLN) once or twice a week. This involves a shift from seven in the morning until three in the afternoon. And we also do this once a month on a weekend. The nature of this role hasn't changed much, but we are now being asked to do this more often. As the organisation structure has become more ‘silo’ based, this means that we don't have such a good understanding of what happens in other areas as we did before. So that makes the DLN role more difficult because you're always having to ask about things, to get the information.

If duty nurses do not fully understand the issues they face, and spend time gathering information, this may adversely affect the quality and safety of care. The advantages and drawbacks of service-line management structures in acute settings have not been well researched.

Management agendas

A central feature of the organisational context concerns the nature of the agendas that management had to deal with. In set-up interviews, middle and senior managers were asked to identify the main issues on the management agendas of their trusts. Given space constraints, this section explores the agendas of two trusts, Netherby and Greenhill: one foundation, the other not.

Much doing to be done

If NHS managers are struggling it is because of the size of the agenda and the lack of respect for the management task involved; and the systematic lack of investment in learning – it is hard to learn together when there is so much ‘doing’ to be done.

p. 19114

What are the management implications of this analysis of organisational context and management agendas? From an operational management perspective, there is a daily challenge to match patient demand with clinical, physical and financial resources. Patient numbers are high and unpredictable, and reduced length of stay means that patient flows are faster paced. Demand has increased in many services while resources have been cut. Management has also had to address radical changes flowing from national, regional and local reorganisations, higher public expectations, an increasingly burdensome regulatory regime and the need to reduce costs, year on year, potentially affecting job security. In this context it is not possible to focus on, say, two or three issues in 1 year and on another set of issues the next. This is an agenda on which all of the items are always priorities, which compete with each other for attention and resources, and which change frequently.

From research in commercial organisations, Bruch and Menges115 describe a similar pattern: increasing market pressures and speed of activities, raised performance goals, new technologies and systems and attempts ‘to make this furious pace the new normal’ (p. 82). This leads to what they call ‘the acceleration trap’, in which focus is scattered, staff are demotivated and customers become confused. In the ‘fully trapped’ companies that they studied, 60% of employees agreed or strongly agreed that they lacked sufficient resources to get their work done.

Are NHS managers experiencing a similar ‘acceleration trap’? Our management survey (see Appendix 3) included two items relevant to this question. One concerned having sufficient resources (reverse wording from that of Bruch and Menges115) and the other concerned improving the service with fewer resources. Responses to those items are summarised in Table 18.

The percentage of managers agreeing or strongly agreeing that they possessed sufficient resources ranged from 19% at Netherby to 42% at Greenhill. The ‘% disagree’ figures are more comparable with the Bruch Menges question, and range from 66% at Netherby to 38% at Greenhill. Experience of the acceleration trap thus appears to be mixed. However, only 30% of the whole sample agreed or strongly agreed that they had sufficient resources, with 70% disagreeing or neutral. Turning to the second of the items, a more consistent picture emerges. In all five trusts, > 90% of respondents agreed or strongly agreed that they were expected to improve the service despite resource cuts.

Having to cover many different kinds of activity simultaneously is what Bruch and Menges115 call ‘multiloading’. An associated pattern is ‘perpetual loading’, in which the organisation operates close to capacity, drives employees hard and deprives them of any hope of retreat: ‘when is the economising going to come to an end?’ (p. 83). The solutions for breaking out of this trap are to prioritise projects, halt less important work and clarify strategy. They cite the chief executive who allowed managers to name only three ‘must win battles’ to focus attention, energy and action.

The evidence from this project suggests that many health-care managers are subjected to multiloading and perpetual loading. The institutional and organisational contexts, however, make it difficult to follow the obvious advice. Recent experience suggests, for example, that an acute trust focusing its management attention on, say, addressing financial problems and the demands of regulatory bodies may find that less attention is paid to the quality and safety of patient care, with disastrous results. This appears to have been part of the explanation for events at Mid Staffs. 70,78

A further answer to the problems of multiloading and perpetual loading lies with management capacity. Our evidence indicates that middle managers are highly motivated and deeply committed, but that they are overstretched and under-resourced. National policy has been to treat health-care managers as wasteful bureaucrats and to reduce management numbers and costs. This has been accompanied by other policies, directives and initiatives that have added to the management agendas of acute trusts, stretching capacity even further. Management capacity should perhaps be a priority issue for further research.

Introduction

One of the main aims of this study was to gain a better understanding of the realities of middle management work in acute health-care settings and, in particular, to understand:

• the motives and rewards for middle managers

• the pressures and demands that they face

• how middle management work is changing.

This chapter first reviews published research evidence relating to the nature of middle management work in general and then reviews evidence relating to the nature of middle management work in health care. The evidence from this study concerning motivation, demands and trends in the nature of management work is then summarised, drawing on survey results along with interviews and focus groups. The main conclusions from this analysis are that:

• Middle managers in general hold a wide range of responsibilities, from ‘keeping the show on the road’ to contributing to strategy, innovation and change, with a profile that sharply contradicts the ‘pen-pushing bureaucrats’ imagery.

• Middle managers, since at least the start of the 21st century, have seen their jobs enlarged, their responsibilities widened, the pace and intensity of their work increased, their working hours lengthened and their performance monitored more closely.

• Managers in health care face trends and pressures similar to those affecting middle management elsewhere, but also face the challenges of driving change in a professional bureaucracy, dealing with a negative image and a perceived absence of adequate support for hybrids with managerial responsibilities.

• Surprising survey findings concern the absence of the negative management stereotype (the majority were happy to be seen as managers), the prevalence of ‘extreme job’ characteristics and low levels of both job and organisational satisfaction.

• Survey findings triggering concern relate to unsustainable workloads, inadequate resources, poor work–life balance, the small but significant proportion considering leaving the service and the view that financial pressures have compromised patient safety.

• The primary motivations and rewards for middle managers include making a difference, driving innovation and change, doing a good job, feeling valued, developing others, working in high-performing teams, and personal development.

• A small number said that rewards were lacking, that contributions were not being recognised and that leadership styles were demotivating, and 42% of survey respondents agreed or strongly agreed with the item, ‘I sometimes feel like leaving this employment for good’.

• Many middle managers in health care appear to have jobs with a profile that can be described as ‘extreme’ in terms of pace, intensity, scope, conflicting priorities, long working hours and other characteristics – a pattern that is exciting and fulfilling for some.

• Extreme jobs are also associated with stress, fatigue and poor work–life balance, can generate organisational and domestic difficulties, may not be sustainable and may be especially problematic for those in hybrid managerial roles in which rapid and intense context switching between different kinds of complex tasks could increase human error.

• Given the multiple roles of middle managers in ‘keeping the show on the road’, mediating between the front line and top team and driving change and innovation, it may be useful to consider strategies for ‘releasing time to manage’ in the face of increasing workloads and decreasing resources.

• One research priority arising from this analysis concerns the development of a better theoretical and practical understanding of the concept of management capacity, and of how capacity can be assessed and strengthened in a context of financial constraints.

• A second research priority concerns understanding the nature, incidence and implications of extreme jobs among health-care managers, along with ways to redesign such roles, and/or provide better forms of support for those who hold extreme jobs, including resilience training.

Studying management roles in the 1980s, Scase and Goffee116 described a context in which (conservative) government policies were increasing pressure on all organisations to be more cost-effective and efficient, with corporate restructuring reducing management job security. Expectations of personal development and opportunities to exercise independent judgement were not being met, and middle managers were being asked to achieve goals with limited resources, and with tighter monitoring of standards. Scase and Goffee116 reported that the ‘reluctant managers’ in their study were ‘able to adapt by investing very little of themselves in their work and merely performing their jobs to an acceptable minimum standard’, particularly in bureaucratic settings with more rules and constraints, and with more scope for minimalist role playing (p. 51).

These observations concerning cost pressures, insecurity, regulation and diminishing resources apply to health care in the second decade of the 21st century. However, rather than respond by decreasing personal investment in work, for many health-care middle managers in the current study the issue concerned potentially excessive personal investment, in the form of ‘extreme jobs’.

The management role

Interest in management work dates from 1956 and Whyte's The Organization Man,117 which created the stereotype of the compliant, risk-averse bureaucrat, an image that survives into the 21st century in characters such as David Brent, the manager in the television comedy series The Office. Whyte's account was followed by numerous other studies of management roles. From observation in a chemical plant in the American Midwest, Dalton118 noted that departures from formal organisation processes were required to make the plant run smoothly, emphasising the role of informal networks, office politics and personal influence. Following in Dalton's ethnographic footsteps, Watson6 documents the rewarding and painful realities of managing from his study of a telecommunications factory in the British Midlands. The subtitle of Watson's book is ‘culture, chaos and control’, highlighting dimensions of the management work that he observed.

Mintzberg4,119,120 describes the activities of management on the ‘planes’ of information (communicating, controlling), people (leading, linking) and action (doing, dealing). He also describes the pace, brevity, variety, fragmentation and discontinuity of management work, and the preference for action and informal and oral communication. Defining management as ‘deciding what should be done then getting other people to do it’ (p. 6), Stewart5 is also concerned with management activities – planning, organising, motivating, controlling, co-ordinating, staff development – noting the variation in management roles, and the fragmented, chaotic and hectic nature of the work. From his study of general managers, Kotter7 notes the long working hours (60–90 a week) and the considerable time spent in conversation, asking questions, joking and persuading – but this is intelligence gathering, not time-wasting.

Hales22 portrays the variety in management work, which typically involves activities such as acting as figurehead, monitoring and disseminating information, networking, negotiating, scheduling and monitoring work, allocating resources, human resource management, problem-solving, innovating, and technical tasks relating to professional or functional specialisms. Management thus involves an ‘inescapable preoccupation with routine . . . keeping the show on the road’ (p. 51), reacting to events, focusing on the urgent and unforeseen, accompanied by tension, pressure and conflict in juggling competing demands. There is also considerable choice and negotiation over the nature and boundaries of the management job and how it is done.

Hales22 notes that the focus on what managers do overlooks the impact of those activities. The management contribution is explored in the following chapter. However, it is important to note that evidence concerning middle management contributions to organisational effectiveness departs from the stereotype of bureaucratic blockers. It is now recognised that middle managers play a key intermediating role between front-line operations and senior management. 30,121 Middle managers have also been identified as key change agents,122 leading change ‘quietly’14,123 and ‘below the radar’,124,125 as ‘ideas practitioners’,126 as well as implementing, synthesising, championing and facilitating strategic organisational change. 12,29

How are middle management roles themselves changing? From their study of eight public and private sector organisations in six Western European countries, Dopson and Stewart127 challenged the image of middle managers as frustrated, powerless and disillusioned with dreary jobs in rigid hierarchies. They found middle managers responding to complex and changing pressures, with increased workloads and responsibility, with accountability for a wider range of duties, seen by senior management as pivotal to implementing change. Middle managers had to become more flexible and generalist, and shorter hierarchies had put them closer to the top team.

Hassard et al. 128 interviewed 250 managers in 30 organisations in the USA, the UK and Japan, concluding that middle management roles have become more challenging, with longer working hours, increased performance pressures, expectations of rapid results and a blurring of work–domestic boundaries. This may be offset by higher salaries, greater responsibilities and more interesting work, ‘but the overall feeling of being overwhelmed in work while the traditional promotional ladder has been largely removed was a major and widespread finding’ (p. 228). However, this study also found that middle managers were highly motivated despite the pressures, and they argue that performance could be improved if the pressures were released. In addition, ‘authoritarianism and top-down rule were alive and well’ (p. 13); there was not much sign of the ‘post-bureaucratic’ organisation.

The evidence thus shows that middle managers have an increasingly varied range of responsibilities, with long working hours and rising performance expectations, playing key roles in operational and human resource management, mediating between the front line and top team, contributing to strategy, innovation and change. These features also apply to middle management work in health care, as explained below.

Middle management in health care

Over the past decade, layers of national and regional organisations have accumulated, resulting in excessive bureaucracy, inefficiency and duplication. The Government will therefore impose the largest reduction in administrative costs in NHS history. Over the next four years we will reduce the NHS's management costs by more than 45%.

p. 431

Despite evidence confirming the significance of middle management, the stereotype of the ‘petty bureaucrat’ prevails in health care, where management costs are associated with wasteful bureaucracy. Hyde et al. 129 argue that denigration of the function obscures vital strategic and co-ordinating work, that ‘middle management’ is an identity that nobody wants and that middle managers have become ‘a lost health service tribe’. However, Currie and Procter51 stress the importance of the ‘diplomat’ role of middle managers in a professional bureaucracy, mediating between medical and corporate goals. Carney130 and Pappas et al. 131 argue that ‘the strategic middle manager’ has become more, not less important in health care. National policy has rejected those perspectives.

Hyde et al. 129 observe that ‘little is known about the realities of health management and middle management work in the NHS is obscured’ (p. 18). One reason for the lack of evidence perhaps concerns the distributed nature of the function, which does not comprise a well-defined group;132 literature reviews must rely on search terms such as ‘lead nurse’, ‘ward sister’, ‘modern matron’, ‘service lead’, ‘clinical director’ (see Chapter 2). A second plausible reason relates to the stigma attached to middle management, making this group difficult to identify. In addition, a search for relevant literature reveals more commentary than research, which is more likely to be published in sector- and occupation-specific journals than in general organisation and management studies journals. The current study thus seeks to contribute to our understanding of those management realities.

A review of published evidence leads to eight observations. First, research focuses on nurse management roles and clinical directors, with less attention to other middle managerial levels and non-clinical management roles. Second, hybridisation has meant changing job titles, from ward sister to ward manager (in the UK133), from charge nurse to nurse unit manager (in New Zealand134), reflecting expanding roles, covering larger areas and more staff, and increasing workloads with longer working hours. 135–138 Third, hybridisation involves clinical staff in budgeting, human resource management, planning, change and administration. 139–144 Fourth, middle managers (pure plays and hybrids) have been subjected to increased monitoring and accountability. 145–147 Fifth, the significance of the bridging role of middle management, between front line and top team, has increased. 148 Sixth, these changes have taken place in the absence of adequate preparation for those moving from clinical to hybrid roles. 149 Seventh, hybrid middle managers have problems balancing clinical and organisational priorities and resource allocations. 150,151 Finally, the emphasis in middle management roles has shifted towards change agency and leadership. 152–154

Research also suggests that middle managers face several problems in addition to workload and competing stakeholders. They are often excluded from decision-making and feel isolated and controlled. 141,155,156 They have difficulty influencing and challenging doctors51,157 and work with ambiguous lines of responsibility. 158 Hybrids report tensions between clinical and operational priorities,159,160 along with difficulty balancing professional development with work objectives,161 and the transition from a clinical to a hybrid managerial role raises identity issues. 156,162 Hybrids are also concerned about isolation from their professional peers. 157 Many feel that they lack support and are unprepared for handling human resource management issues. 163 They are also acutely aware of the negative stereotype devaluing management and threatening job security. 1,145,152,164

In sum, middle managers in health care face a range of pressures similar to those in other sectors, including widening roles, increasing responsibility, longer working hours, work intensification and more intrusive performance management. However, the management population of the NHS comprises a heterogeneous combination of pure plays and hybrids, who also have to contend with the challenges of driving change in a professional bureaucracy, a negative stereotype and, for hybrids in particular, a perceived lack of preparation for, and subsequent support in, managerial roles.

Management survey findings

This section summarises findings from our survey, conducted in 2011. Using 5-point Likert-scaled items, the survey had five sections: realities, job characteristics, organisational outcomes, job satisfaction and personal experience. Items also included a widely adopted organisational commitment scale165 and a similarly well-recognised set of stress indicators. 166 The survey was distributed by e-mail at five trusts, covering staff with managerial and hybrid roles. The response rate varied from 77% at Greenhill (distribution was confined to the trust's ‘100 leaders’), to 19% at Clearview and Netherby. The overall response rate was 24%, generating 611 usable responses. Of those, 18% had management roles and 51% had hybrid roles; 31% did not answer this question. Of the total, 42% were female and 19% were male; 39% did not answer this question. Over two-thirds of responses fell into five categories of work: management, administrative and clerical, project management, nursing and finance. Further details of the survey administration, along with sample properties, response rates, frequency distributions for the whole sample and responses for each trust, are reported in Appendix 3.

Some findings were anticipated, others were surprising and some are cause for concern (percentages cited here are the aggregates of either ‘agree and strongly agree’ or ‘disagree and strongly disagree’; Appendix 3 reports the detailed breakdown of responses across the full scale).

Motives and rewards

The stereotype suggests that middle managers are motivated by bureaucratic rules, organisational stability and administrative order. The reality is different. Survey responses suggest that middle managers have other motives and information from set-up interviews supports this observation. Here, findings from two trusts elaborate this point, and evidence from others is briefly presented.

How is the role changing?

I manage the PALS [Patient Advice and Liaison Service] team of three officers across the two sites, four complaints officers, and staff in patient and public involvement who I have also picked up. I have just got them organised and they are working well. My aim is to value and nurture my staff and I've put a lot of emotional investment into developing the team. Now I have to find £27,000 savings over two years; £12,000 last year, which I didn't achieve, and £15,000 this year. Starting from January 2010. I have just got them organised, and now this. I want to weep.

Senior manager, Greenhill

Middle managers in general have seen their jobs enlarged, their responsibilities widened, the pace and intensity of work increased, working hours lengthened and performance monitored more closely. Middle managers in health care face broadly similar issues. From survey data, Table 21 shows the proportions who agreed/agreed strongly with 14 items that sought to capture these trends.

For the first four items, around 80% or more agreed, reflecting pressures stemming from the government policies being pursued during this project, concerning cost-effectiveness, targets, commercial orientation and performance monitoring and regulation. Responses to the next four items capture the consequences for individual management roles: lack of time, insufficient resources, unmanageable workload and unsatisfactory work–life balance. Only one-third or less did not rate those issues as problems. Items 9 and 10 concern the human resource management implications, passing pressure down the organisation, and finding it more difficult to motivate staff respectively.

The last four items apply to hybrid roles, with clinical staff increasingly taking on managerial roles. However, although just over half of respondents agreed that they had problems balancing managerial and clinical responsibilities and > 40% said that they had reduced clinical duties accordingly, given the pressures on all middle managers it might have been anticipated that the proportion of hybrids experiencing this clash of priorities would have been much higher. Also surprising is the low proportion – less than one-third – expressing resentment at having to reduce clinical responsibilities in favour of managerial work. This again contradicts the negative stereotype of health-care management and is also at odds with the image of health-care professionals who are supposed to place a low value on the management role. This suggests that the traditional tension between ‘white coats and grey suits’ may be eroding, with recognition of the need for closer interprofessional collaboration in the face of the range of complex challenges facing the sector.

Details behind those responses can be seen in the focus group discussion with three lead nurses at Greenhill. Their insights, summarised in Table 22, reflect dimensions of the management agenda discussed earlier and were typical of responses from other groups about changes in management roles.

This is a bleak picture, suggesting that managers lack the time, resources and support to fulfil their responsibilities effectively. The threats and blame, bureaucracy, inspections, tension, firefighting, increasing workloads and lack of trust and autonomy are likely to work against pressures to be more adaptive, creative, innovative and commercially orientated. Nevertheless, focus group participants invariably expressed their commitment to the service, to their profession and to patients. The bleak picture must be set in the context of positive responses under other headings, particularly with regard to motives and contributions. One conclusion is that, if the trends captured in Table 22 were to be reversed, the resultant effect on motivation, commitment and contribution could enhance individual, divisional and corporate effectiveness considerably. This set of issues is discussed in Chapter 6, in terms of building and sustaining an enabling environment for management work, in which most of the appropriate actions are cost neutral, requiring changes in attitudes, behaviours and processes.

These changes appear to have taken some management roles in a different direction, creating for some what have been described as ‘extreme jobs’, which are the subject of the following section.

Extreme jobs, resilience and job crafting

Are the dimensions of the changing roles of health-care managers undesirable? This section offers a more nuanced response, based on the concept of the ‘extreme job’, identified by Hewlett and Luce167 among high-earning professionals in law, finance and consultancy. They identified 10 dimensions of such roles, but four appear to be less relevant in health care: direct reports, entertaining clients out of hours, profit and loss responsibility and international travel. From the preceding discussion of how management roles are changing, the other six dimensions that may apply are:

1. unpredictable work patterns

2. fast pace with tight deadlines

3. broad scope of responsibility

4. ‘24/7 availability’

5. mentoring and coaching other staff

6. long working hours.

Our focus group evidence suggested a further six ‘extreme’ health-care dimensions, which are illustrated here with candid quotes from participants (the sources of the following quotes are omitted to preserve anonymity):

1. 7. Making life or death decisions:

   • ‘We were on the unit while the alarms were going off, and the patient kept going into cardiac arrest. It was disturbing from my perspective, because I'm a nurse by background and I've worked in ITU [Intensive Therapy Unit], and they were flying around, the arrest button was going, alarms were going and it was very hard.’

2. 8. Dealing with conflicting and changing priorities:

   • ‘It's a case of, “what are we going to panic about this week?” We have more to deal with. There are so many competing priorities and edicts, it's easy to get swamped and forget why you're here.’

   • ‘There will be a big push on something, because we're having a visit, or we're non-compliant in an area, so we do that, then forget it, and move on. We're having a visit about safeguarding, so that's in vogue.’

3. 9. Doing more with fewer resources:

   • ‘I was told to “get that racking off the floor”, by screwing it to the wall. But I can't do that; it costs too much. I can't spend £400 on a bit of equipment, an ophthalmoscope, but we could be sued if a symptom is missed. Medical engineering want £35 a year for the maintenance contract. So, no new equipment.’

   • ‘We're facing greater demands. It's hamsters on the wheel. We need head space. It's like Winnie-the-Pooh being dragged backwards down the stairs by Christopher Robin; there must be a better way to come downstairs, if only I could stop bumping and think about it.’

4. 10. Responding to audit, compliance, regulation and inspection agencies:

   • ‘Responding to regulatory bodies, all coming at it from different angles. I spent last week just filling in forms, responding to request for information. I had no time to do anything that would make a difference.’

   • ‘I've been inspected by the SHA, OFSTED, LINKS, CQC, HOSC, HSE. They have different requirements, which overlap, but they have different parameters. They all want information, reports, and action plans, which contradict each other.’

5. 11. ‘Big tent problem-solving’; the need to get many people to agree:

   • ‘I think I'm creative at finding innovative solutions. But implementing anything new takes massive amounts of energy, and you are ground down. You get doors closed in your face repeatedly.’

   • ‘There's so much red tape. Our hands are tied. I have to get four sign-offs before I can recruit a nurse. We're not trusted to manage.’

6. 12. The floggings will stop when morale improves:

   • ‘There are no meetings where you are given credit for success. It doesn't matter what we do, it's never good enough.’

   • ‘My general manager is like “the grim reaper”. No positives, all negatives. We never get any thanks or positives, even when we have made our targets.’

This 12-component profile was used to design a simple diagnostic with items such as, ‘I never know what's going to happen next in this job’ (unpredictable work patterns), ‘I'm always trying to meet another deadline’ (fast pace, tight deadlines) and so on. This was completed by six members of an operations management team at Clearview, who were asked to respond ‘yes’ or ‘no’ to each item on the premise that ‘yes’ to six or more items implied an extreme job. The group commented:

I answered all but one of the questions with a ‘yes’. I have tried to stick to my working hours (minus lunch) and have just seen the work pile up.

Directorate support manager

I can answer yes to all the questions. Some parts I love, but there are many days where I would like space to think. This is becoming more precious. We have made errors through firefighting, but we applaud ourselves when we go ‘phew that was a lucky escape’.

Operations manager

Answers are a mix of yes, no, and sometimes.

Operations manager

This is reasonably balanced, not exaggerated or understated. The greatest frustration in medical management is the challenge of producing change in an environment that is very resistant.

Consultant

The adrenalin rush can occur without the need to rush around. Great colleagues are even greater if you don't get so tired that you snap at them.

Discharge planning manager

There is a great deal to be said for people who manage their workload in a way that means that they work their hours (plus some extra when necessary). They shouldn't feel bad about working their contracted hours.

Deputy operations manager

This pilot suggested that the concept of an extreme job profile for health-care management had face validity, judging by the first four responses, and that jobs vary in substance and over time on those dimensions. Some feel that extreme jobs should not be necessary. This diagnostic became the ‘job characteristics’ section of the survey. Responses to the 12 items are shown in Table 23.

With one exception, agreement with the items was high, ranging from 54% (relating to unpredictability) to > 90% (doing more with less). This response pattern suggests that the incidence of extreme jobs (as defined by these items) in health-care management is widespread. Only around 30% agreed with the last item, a result inconsistent with focus group comments. Either those comments were misinterpreted, or this item was badly worded; an item reading ‘it is never good enough for senior management’ may have produced different results.

As expected, those with high extreme jobs scores also indicated higher stress levels and poorer work–life balance. Interestingly, the positive relationship between having an extreme job and being under stress was stronger in men, despite the fact that women were more stressed generally. There was no relationship between extreme jobs and commitment, the latter being more closely associated with influence, recognition and ‘buzz’.

Hewlett and Luce167 found that ‘extreme jobbers’ were attracted by the challenge, enjoyed the work and felt fulfilled by it. Asked, ‘why do you do it?’, they mentioned adrenaline rush, great colleagues, good pay, power and status and recognition for achievement. Those items were added to our survey design and responses are summarised in Table 24. We know that health-care managers are motivated by the challenge and the ‘buzz’ that make extreme jobs attractive. We also know that working with colleagues in high-performing teams is another key motivator. However, less than half thought that they were well paid, and recognition, power and status appear not to be experienced by around two-thirds of this sample. Other evidence shows that health-care managers are also motivated by making a difference, driving change and innovation, doing a good job and developing others. The health-care variant of the extreme job profile proposed here may thus be accompanied by a health care-specific motivational profile relying at least in part on professional and public sector values and purpose, which may offset or buffer or compensate for the negative aspects of extreme jobs.

The subtitle of the article by Hewlett and Luce is ‘the dangerous allure of the 70-hour workweek’. High-performing ‘adrenaline junkies’, in particular, appear to enjoy this kind of work,168 so where is the danger? Multitasking for long hours across complex roles can lead to fatigue, burnout and mistakes. 169 In such circumstances, family life may also be adversely affected. In a health-care context, patient safety and care quality could potentially be compromised. In addition, working constantly at this pace may not be sustainable. 170 Only one-third of survey respondents said that they were able to maintain a satisfactory work–life balance. But an improved work–life balance could create a climate that is not challenging enough for the ‘A players’, who then become more difficult to recruit and retain.

A varied, intense, fast-paced role with responsibility and long hours can be rewarding. The fact that such jobs can be attractive, however, could be a problem. The original research found extreme jobs among high-earning professionals. This study suggests that some middle and senior managers in acute health care may also have jobs that are extreme, in terms of the profile suggested here. Most hospital managers are ‘hybrids’, covering managerial and clinical responsibilities. If the management component of their roles has become more extreme, this may affect the time and energy that they devote to patients, could increase the potential for errors and may also discourage clinical staff from taking on such jobs in the first place. In addition, some of the extreme health-care job dimensions are potentially demotivating: competing priorities, having to do more with less, the burden of regulation, bureaucratic barriers, a climate of negativity. The assumption that all extreme jobs are exhilarating and motivating is misleading; some are exhausting and frustrating.

There are substantial research traditions with regard to ‘heavy work investment’ topics such as long hours, work intensification and ‘workaholism’. However, research into extreme jobs has been limited, perhaps because of the non-academic source of this concept. A literature search identified only two other studies,171,172 which found that longer hours were associated with higher satisfaction, better career prospects and higher salary, but with higher levels of stress, more psychosomatic symptoms, lower family satisfaction and poorer emotional health.

Conclusions and implications: releasing time to manage

There's no time for extra projects and that sort of thing, which is a shame. So you're just doing your day job, there's no time . . . you don't get a chance on any working parties for progression and that sort of thing. I don't think we do enough forward planning. Because we haven't got the time to do that. While you recognise that, yes, this is a firefighting type of job, and it is reactive as opposed to proactive, I'm sure, well, in fact I know, that if we had more time to plan, we would do things differently, and a bit more efficiently.

Operations manager, Clearview

The combined weight of evidence from past research and from this project suggests that middle managers play key roles in maintaining operations, mediating between the front line and top team, developing innovative service improvements and helping both to shape and to implement strategy. Further evidence concerning middle management contributions is explored in the following chapter. Middle managers appear to be deeply committed to the service and its patients, and highly motivated to make a difference. The motivational profile includes hygiene factors but is dominated by themes that would be cost neutral to strengthen, such as teamwork, recognition and autonomy. However, the evidence also portrays a group that is facing increasing workloads, decreasing resources, tighter regulation and more onerous performance expectations. Although having the ideas, this group lacks not only the resources but also the authority, senior management support and the time to develop them.

One initiative that has seen widespread application in acute trusts is ‘the Productive Ward’, an initiative designed by the NHS Institute for Innovation and Improvement. The aim of this initiative, with regard to nursing staff, concerns ‘releasing time to care’. It would therefore be useful to consider ways of ‘releasing time to manage’. Asked in focus group discussions to identify steps to achieve this, middle managers offered the following suggestions:

• What can I do?

  • walk the floor, talk to people, don't be driven by e-mails

  • make time for reflection, create space where problems or issues can ‘sit’, for example on a whiteboard in an office – people can visit and offer thoughts and suggestions.

• What can we do at directorate level?

  • ensure that staff are prepared in advance for performance review meetings; rehearse

  • rapid improvement events, step back, time out to reflect on what we are doing

  • release time at the mid-levels, ‘office days’, fewer and shorter meetings.

• What corporate actions would help?

  • clinical director and directorate manager make a powerful team – develop this potential

  • give managers timely and understandable financial information

  • empower ward sisters to manage their wards effectively

  • we have potential at all levels that we need to tap into

  • make it safe to challenge; the attitude, ‘I'm only a porter’, needs to be replaced.

These suggestions are offered here for three reasons. First, they demonstrate that the constraints on middle managers are widely recognised. Second, they indicate that there are many, simple, practical steps that could begin to address this issue. Third, they serve to introduce the concept of an ‘enabling environment’ – a supportive context for management work – and this concept is explored in more detail in Chapter 4.

The NHS has a motivated and committed middle management population. Given the motivational profile, most of the actions necessary to sustain and to strengthen motivation and commitment are cost neutral. Middle managers are a key source of new ideas and innovations for service improvement and development. The workload, time and other organisational constraints under which they operate stifle their ability to implement those ideas. The middle management population is thus in danger of becoming an underutilised, misused, wasted and potentially wasting asset.

What's the problem?

The management of extreme events

Figures published yesterday by the National Patient Safety Agency show that more than 5,000 patients died or were seriously injured as a result of NHS safety incidents in the six months to March 2011. A total of 1,313 people in England and 78 in Wales died as a result of medical errors involving the care they received from hospitals, mental health trusts and ambulance services. A further 3,699 patients in England suffered ‘severe harm’ – permanent harm, including disability and scarring.

NHS Confederation press summaries, 14 September 2011; www.nhs.confed.org

Following a number of ‘high-profile’ service failures, patient safety has become a national priority. 99 From 2003, a voluntary reporting scheme, the National Reporting and Learning System (NRLS), operated by the National Patient Safety Agency (NPSA) has been in place. This invited health-care organisations to report serious untoward incidents. Serious untoward incidents are events likely to cause significant harm to patients, members of the public or staff and that could also cause disruption to the service and damage the reputation of the organisation and/or its employees. ‘Never events’ (25 are designated) are a category of serious incidents that should not occur if preventative measures are in place (e.g. wrong site surgery, misidentification of patient). In April 2010, reporting became mandatory. The increase in the number of incidents submitted to the NRLS between 2003 and 2010 can be explained by the growing percentage of health-care providers submitting monthly reports (1% in 2004; 81% in 2009); around one million incidents are now reported annually. In 2009, although most incidents caused no harm, 3735 were reported to have caused death (NPSA Quarterly Data Workbook: www.nrls.npsa.nhs.uk/resources). Walshe and Higgins186 date the increase in inquiries into major service failures from the Ely Hospital scandal in 1967. Recent events at Mid Staffs have also been subject to several investigations. 70

Some of these serious untoward incidents lead to compensation claims. Established in 1995, the NHS Litigation Authority (NHSLA) handles claims against the NHS, administering the Clinical Negligence Scheme for Trusts (CNST) and managing non-clinical claims and the ‘liabilities to third parties’ scheme. Table 26 shows the numbers of claims received since 2004 and the payments made against them. 187

The total cost of claims was almost £800M in 2010–11, more than double the cost in 2004–5. By 2012, a typical settlement for a catastrophic injury claim was £6M. The NHLSA attributed these rising costs to the ‘no win no fee’ market whereby claimants can litigate with no financial risk, and to lawyers and agencies ‘farming’ complaints against NHS organisations. 188

Serious incidents and never events harm patients, disrupt services, damage organisational reputations and individual careers and can generate heavy costs. There is now a substantial body of work in this area, triggered by the landmark publication by the US Institute of Medicine, To Err is Human,189 which estimated the annual cost of health-care errors (in the USA) to be between $17B and $29B. That literature is rich in terms of guidelines, diagnostics, protocols, checklists and alerts. The main problem, however, appears to lie not with guidance, but with implementation.

Much is already known about the causes and management of extreme events. Crisis management research has mirrored the stages of the event sequence narrative. This typically begins with the ‘incubation period’190 and associated ‘resident pathogens’,191 followed by a focus on risk management and emergency planning. 192,193 There is practical crisis management advice,194 along with accounts of individual events: the Mann Gulch disaster;195 the loss of the Challenger space shuttle;196 Snook's account of a fatal friendly fire incident;197 Perrow's studies of ‘normal accidents’. 198 Inquiry reports can be seen as proxy research data. 199,200 The development and application of safety cultures,201 ‘resilience engineering’202 and ‘high-reliability organisation’203 have been advocated as ways to reduce the incidence and impact of extreme events.

The dominant perspective in the post-event period views what happens to inquiry recommendations in terms of organisational learning. 204 A number of commentators have noted the absence of a change management perspective,40,205 and this is the lens – focusing on the design and implementation of a change agenda – through which the extreme events summarised in this chapter are viewed.

Root cause analysis (RCA) is the main approach to health-care incident investigation. RCA tends to focus on immediate and well-understood causes, and can overlook wider systemic, contextual and temporal factors. Carroll199 explores the logics driving incident review teams in nuclear power and chemical processing settings, concluding that RCA leads to ‘root cause seduction’, by which he means a reassuring preference for simple and rapid solutions, potentially overlooking the less obvious and more challenging systemic roots of failures. Leveson206 argues that ‘the emphasis in accident analysis needs to shift from “cause” (which has a limiting, blame orientation) to understanding accidents in terms of reasons, i.e., why the events and errors occurred’ (p. 241). Although based on research in other sectors, the advice of these commentators appears useful in examining the reasons behind and the aftermath of the health-care incidents discussed later in this chapter.

Current approaches to accident modelling also rely on systems-theoretical perspectives,207,208 rather than ‘domino’ and ‘Swiss cheese’ models. Rasmussen209,210 observes that, although incidents may be caused by loss of control of work processes, other factors are involved in safety control. Figure 4 depicts what Rasmussen209 calls ‘the problem space’ (p. 185), identifying the multiple layers of control that are often implicated in accident causation, the disciplines associated with their respective study and ‘environmental stressors’ such as changing political climate, public expectations, financial pressures, changes in education and competencies and the pace of technological change.

FIGURE 4.

The problem space: sociotechnical system risk management hierarchy. Based on Rasmussen, 1997. 209

This generic model seems to be relevant to UK health care, particularly with regard to system layers and environmental stressors. An accident scenario can be represented by an ‘accimap’ showing how events and conditions interact, revealing the pattern of antecedents, causality and consequences. Figure 5 shows an accimap for the case of Mrs Mayland (discussed later). Although difficult to read in this format, this illustrates the complex causality behind this patient's death, and the links between different layers of analysis over time. Although this patient died from a prescribing error, the accimap also captures the roles of regulatory bodies, organisation and management, infrastructure and processes and procedures. Rasmussen209 argues that such incidents ‘have not been caused by a coincidence of independent failures and human errors, but by a systemic migration of organisational behaviour toward accident under the influence of pressure toward cost-effectiveness in an aggressive, competitive environment’ (p. 189). Individual decision-makers don't see ‘the complete picture’ and are not able to judge the state of defences, which depend on decisions taken by others elsewhere.

FIGURE 5.

Accimap of the Mrs Mayland incident. BNF, British National Formulary; CT, computed tomography; GCSF, granulocyte colony-stimulating factor.

The systemic nature of serious incidents in health care is reflected in ‘the London protocol’, which identifies the hierarchy of factors influencing clinical practice: institutional context, organisation and management factors, work environment, team factors, individual (staff) factors, task factors and patient characteristics. 211,212 Although this protocol underpins RCA training, the process and outcomes of investigations based on this approach do not always reflect a systems perspective.

Nicolini et al. 213,214 identify barriers to learning from RCA investigations. Investigations are conducted by staff who were not involved in the incident and who continue with their full-time roles. Interrogating witnesses, collating and analysing evidence and drafting recommendations can be a complex and protracted process. Outcomes are often influenced by time constraints and lack of expertise. Tensions are generated by the need to make service improvements in the face of limited resources, competing priorities and conflicting views. They found that investigation reports often focused on local clinical practices and context, and they also note that ‘many potential root causes were discounted in the analytical process, often based upon the assumption that such latent factors were not easily resolved or because the complexity and ambiguity of the issue would not allow it to be resolved with a single, clearly containable countermeasure’ (p. 37). 213 Patterns and trends in incidents were rarely considered and the investigation report was often seen as an end in itself. The need to produce an acceptable plan focused attention on available solutions and minor improvements, rather than on organisational changes with substantial resource implications. This study concluded that risk managers and investigation teams lacked ‘a coherent orientation towards managing change’ (p. 39), for which the typical strategy meant designating a clinician responsible for the action plan, and checking progress some months later. Implementation issues such as facilitating change, addressing resistance and linking to other initiatives were rarely considered.

The four incidents that follow illustrate many of these issues, confirming aspects of the nature of investigation processes and reinforcing the barriers identified by Nicolini et al. This analysis also points to ways in which RCA methods could be complemented by other tools and perspectives, to increase the probability of effective and more rapid changes flowing from such events.

Clostridium difficile at Burnside

In contrast to the incidents that follow, the way in which Burnside managed a serious outbreak of C. difficile was a success. This incident was explored, at the chief executive's suggestion, to explain that success. The resultant approach is termed a ‘maintenance model’, because the actions necessary to prevent a recurrence of a crisis such as this are quite different in nature and timescale from those actions that were required to contain the crisis in the first instance. In other words, the maintenance phase of the event sequence, after the immediate crisis was over, was as important as the initial crisis management and emergency response. How did this incident arise? The factors contributing to the outbreak are summarised in Table 27, which, although not following precisely the model in Figure 4, illustrates the multilayered nature of the problem.

Table 28 extends that pre-crisis context, identifying the subsequent main phases in the event sequence narrative for this incident, and tracking the impact on C. difficile infection rates.

During the pre-crisis ‘incubation period’ the infection rate of up to 30 a month was perceived as ‘normal’ and, although this was a known problem, this did not trigger any special action. This changed when, in June 2007, national tables identified Burnside as having one of the worst C. difficile infection rates in England; as one staff member noted, ‘bloody hell we're in the bottom ten’. This observation did then trigger a rapid emergency response, the details of which are shown in Table 29. The hospital's success appeared to be due to the impact of a combination of actions managed as an evolving programme, constituting a ‘package deal’, in contrast with traditional ‘stepwise’ change implementation guidelines. This six main components of this ‘package’ included:

1. A turnaround team: a cross-departmental clinical and managerial group with authority to act without further permission from senior managers.

2. Appraise and prioritise: rapid decisions on immediate actions, delayed action on more difficult and sensitive issues.

3. Emergency response: quick demonstration that the problem was understood and was being addressed; autocratic, ‘no questions – no negotiations’ style; ‘political fix’ to reassure external stakeholders as well as ‘real fixes’ to resolve the problem.

4. Systemic solution: systemic problems need systemic solutions, including individual, team, organisational, financial, infrastructural and other factors; in addition to many changes in working practice, communications were frequent, authoritative and appealed to professional values rather than external targets.

5. Measure and report progress: infection rates were monitored and published; all staff were made constantly aware of performance on key metrics; continuing success motivated staff to maintain the trajectory.

6. Plan for continuity: crisis over, is the turnaround team redundant? No; the team continues to work, maintain focus on the agenda, maintain and improve reduction in infection rates – the shift from emergency response to maintenance phase was critical to success.

The number of new cases fell below 15 a month in August 2007, and by the end of 2009 the rate was between zero and five new cases a month; that rate has been maintained since (Figure 6).

FIGURE 6.

Clostridium difficile infection rates at Burnside, December 2005–September 2011.

What works in one setting will not always work elsewhere. However, this analysis suggests that changes can be implemented rapidly through a combination of compelling evidence, autocratic management (where appropriate), a powerful cross-functional management team and innovative communications that encourage behaviour change by addressing beliefs and values rather than focusing on externally imposed targets. This combination of actions, including a constantly changing change agenda (to combat the ‘infection control fatigue’ identified by an infection control nurse), accompanied by attention to sustainability was successful both in the short term and in the long run. This was not a ‘quick fix’ approach to crisis management. The Burnside ‘maintenance model’ is consistent with the systems approach outlined above, and offers a practical framework that could be adapted by other provider organisations faced with similar difficulties.

Mrs Mayland

The incident involved the death of a patient . . . a catalogue of errors, but the final element that caused her death was being prescribed two drugs that interact with each other, and the results weren't picked up quickly enough and then acted upon. She died just over a week after the drugs were administered. A catalogue of errors up to that point with people not knowing who's in charge and not being clear. But as with all of the most serious untoward incidents, there was a sort of an opportunity, almost up until the last few hours when it could have been turned around.

Associate director, Clearview

Mrs Mayland attended Clearview for a hysterectomy in 1999 – which will be taken as the starting point for this incident timeline. In 2000, she developed severe rheumatoid arthritis and was enrolled in a trial for a new treatment regime that included methotrexate, a ‘high-risk’ drug. In 2000, another arthritic patient at Clearview died as a result of a methotrexate prescribing error. The inquiry noted that this was a medical patient on a surgical specialty ward, blood tests had been delayed and junior doctors did not notice that a full blood count test result was not available. An audit found that only 23% of nurses knew about restrictions on the use of methotrexate, and only 17% of prescriptions were completely accurate. Training and drugs policy were to be revised.

While Mrs Mayland was taking part in the trial, blood was detected in her urine, and an ultrasound scan detected a large ovarian cyst. In spite of her arthritis, she drove herself to Clearview for elective surgery on 26 June 2005. She never left the hospital alive. She was referred to a consultant gynaecologist who removed the cyst laparosopically. Shortly after that procedure she suffered abdominal pain and failure to pass urine. Bilateral ureteric damage from the surgery was identified and corrected. She was then to be discharged to another hospital, which caused distress as her husband had recently died there, so she remained at Clearview. The subsequent inquiry suggested that she should have been given the option to return home, and one interviewee noted that, at this point, ‘she nearly escaped us’. However, she contracted a C. difficile infection from which her recovery was slow. She was then given a faulty rollator (a wheeled walking frame) and fell and fractured her hip; a total hip replacement had to be performed. For this procedure she received the antibiotic trimethoprim, which interacted with the methotrexate. Her blood count began to show a low white cell and neutrophil count. Clinical and nursing records show delays in conducting blood tests and lack of attention to her deteriorating condition. Mrs Mayland died on 3 September 2005. The incident timeline ran for 11 years, from 1999 to 2010, when a final review of the original inquiry recommendations was conducted.

The post-event segment of the incident timeline lasted for 5 years and is summarised in Table 30. This incident had a profound effect on the trust. It led to the establishment of a Patient Safety Executive and the development of a Patient Safety Unit reporting to the trust quality committee, a ‘safety first’ programme with a new patient safety strategy and a more proactive approach to safety issues with Clinical Area Safety Assessments (CASAs) of clinical processes. Patient safety became an important board agenda item, along with patient stories. One of the roles of the Patient Safety Unit was to pursue recommendations that had not been fully resolved.

The inquiry report in 2006 made 46 recommendations. By the end of the first quarter of 2010, 4.5 years after Mrs Mayland's death, 31 of those recommendations had been ‘completed and closed’, six were ‘completed but with remaining issues’ and nine were ‘unresolved and open’. In other words, two-thirds of the recommendations were considered to have been implemented in full, with one-third outstanding or requiring further work. What is the explanation for that ‘tail’ of issues?

Table 31 summarises the 31 recommendations ‘completed’ by September 2010. Organisational changes often generate problems, but in this instance creating a new unit, hiring more staff, streamlining processes, providing more training and nominating ‘methotrexate days’ could be regarded as ‘low hanging fruit’, and hardly likely to trigger resistance. One interpretation, therefore, is that recommendations that were simpler and quicker to implement were tackled first.

Do the recommendations that had not been implemented fully display any common properties? The six items identified as ‘completed but with issues’ were:

• despite review, the content and structure of medical records were still unsatisfactory

• despite development of new nursing care plans, new forms and plans that were ‘outside the agreed process’ were being added to notes

• the ALERT (acute life-threatening events recognition and treatment) sheet had been revised and, although this was in use for 90% of patients, only 10% of these were completed accurately

• despite the introduction of standardised patient observation charts, there were many different charts in use across the trust, with some areas using an electronic system

• procedures for requesting repeat test samples had been defined but results were not recorded appropriately, and many results were not recorded in the medical records at all

• the management of deteriorating patients was part of the medical early warning system (MEWS) with which the trust was 85% compliant, but care of deteriorating patients on general wards was still a cause for concern.

These items appear to concern the reluctance of staff to abandon the old variety of systems, procedures and working practices in favour of new standardised processes.

There were a further nine items – one-fifth of the recommendations – still ‘open’ at the time of the review in 2010, summarised in Table 32. With the exception of the ward pharmacists' issue, the ‘open’ items concern the non-compliance of clinical (mainly medical) staff with new systems, procedures and other working practices, and with what appear to be attempts to standardise procedures in the interests of consistency and predictability.

The progress pattern can thus be explained as follows:

• issues resolved: ‘easy to implement’ changes such as revised guidelines

• issues partially resolved: staff using old systems and practices rather than switching to new

• issues unresolved: medical staff reluctant to adopt standardised procedures.

There had been considerable progress. The management of patients on methotrexate and multidisciplinary teamworking had been improved, and education packages for doctors in training had been revised. Nevertheless, in 2010, prescribing errors were being reported at a rate of 150 a month, and between January 2005 and May 2012 the trust logged 210 incidents (none serious) specifically involving methotrexate, with some of those incidents also involving trimethoprim. At around £10M for a trust-wide system, the cost of electronic prescribing was prohibitive. The 2010 review noted that, although one feature of Mrs Mayland's case was failure to act on a blood test result, ‘resolving abnormal results has been far more challenging than originally anticipated and recent incidents identify that this remains a significant issue’. It thus appears that the ‘completed with issues’ and ‘open’ recommendations concerned actions that had been mandated, but not managed – managed in the sense of establishing the benefits to staff, gaining agreement and commitment, involving those affected in implementation, tackling resistance, regularly monitoring progress and preventing or inhibiting the use of the previous systems and practices that were being replaced. In sum, an ‘orientation towards managing change’ seems to have been absent in this instance. 213

Mr Torrens

Mr Torrens came to the emergency department at Clearview late in the evening of 30 December 2010 having taken an overdose of co-codamol tablets earlier in the day. A young man, with a long history of depression and anxiety attacks, he had attempted suicide in 2005 with an overdose of paracetamol, and had stopped taking antidepressants 2 months before this latest incident. He was admitted to the Clinical Decisions Unit (CDU). He was co-operative during his initial assessment, based on a ‘prompt for mental health/suicidal patients’ form, which informs clinical judgement but does not provide a risk score. Mr Torrens also said that, although he had left a suicide note, this was an impulsive act which he now regretted. By 0530 the following morning he was considered to be medically stable and was referred for psychiatric assessment. This assessment did not take place. The on-call psychiatrist was working at another hospital and his car broke down at 0800. He decided to hand Mr Torrens over to the liaison psychiatry nurse who was due to arrive in the CDU at 0900.

Mr Torrens tried twice to leave the CDU. On the first attempt the nurse in charge took him back to bed and the on-call emergency department doctor noted in the medical record that security should be called if he attempted to leave again. On his second attempt, security was not called. At 0905 the ward housekeeper noticed that Mr Torrens was not on the ward and informed the nurse in charge who initiated a missing persons search. Attempts were made to contact Mr Torrens by mobile telephone, his parents were advised that he was missing and security and police were informed. However, at 0915, Mr Torrens was struck by a train on the railway line close to the hospital and died from his injuries. The liaison psychiatry nurse had arrived at 0910; she was able to access Mr Torrens' detailed mental health records from a ward terminal, which acute trust staff were unable to do.

The investigation into this incident identified three root causes:

1. The mental health assessment performed by emergency department staff did not assess fully the risks associated with patients liable to self-harm.

2. Training to recognise the triggers and behaviours of self-harming patients was inadequate.

3. Support for patients with mental health needs in the emergency department was insufficient.

Training for emergency department staff in recognising and managing patients with mental health issues was subsequently implemented. However, emergency departments can be busy, frenetic environments, unsuitable for patients who are agitated and suffering from mental health disorders. However, the third recommendation here was key. Clearview did not employ on-site psychiatric cover. Psychiatric support was based at another site, with support on call out of hours. Acute trust emergency department staff did not have electronic access to mental health records, and also lacked skills for dealing with mental health patients. In addition, many patients presenting in an emergency department are inebriated or have taken drugs, and mental health staff will not assess under those conditions, waiting until patients are sober and/or the drugs have cleared their system. This meant that psychiatric assessments were often delayed and patients could be moved to another unit, with staff less skilled in managing mental health patients, in order to meet the emergency department 4-hour target. These delayed assessments may have come to be seen as normal:

I think it was an accepted part of the emergency department and the CDU. I think it was probably the thinking – oh, it has always been done this way, always come here, always been accepted to keep the mental health patients, and there has always been a delay in getting them reviewed or moved to a more appropriate environment. That's how it has been.

Senior nurse, Clearview

The provision of out-of-hours psychiatric support was a long-standing concern, with studies suggesting that at least 5% of patients arriving in emergency departments have mental health problems. Acute hospital patients are coded according to medical condition and not mental state, so that figure may be inaccurate. One interviewee commented:

We have had a massive increase, with 18 per cent of people who self-harm through drugs mainly because of the climate we are living in. People are coming in losing their jobs, so mental health is definitely on the increase. And as a department, we don't really have that infrastructure. The other problem is [psychiatric support] not being on site.

Operations manager, Clearview

There had been a similar incident earlier that year, in July, when a male patient in the emergency department became increasingly agitated but his mental condition was not assessed. At nine o'clock in the evening, the patient tried to abscond, stabbed a member of staff with a pen and injured 11 others before the police arrested him (the mental health trust having refused to admit him to their unit). He was assessed the following morning at the police station and then moved to a secure mental health unit. This incident had also prompted discussion between acute and mental health trust management, but without progress.

Following this more recent incident, the management teams from the acute and mental health trusts met to explore appropriate actions. However, at the time of writing, little had been achieved:

I suppose it was the incident on New Year's Eve that was the absolute organisational wake-up call. Having said that, it has been painfully slow to do anything. I think we have a different view on mental health to the mental health trust. Things that cause us angst they do not see as an issue, and I don't think they understand why we see it as an issue sometimes. So getting an understanding of what the problems are hasn't been easy because, actually, what are problems to them are not problems to us, and vice versa.

Senior nurse, Clearview

There was a shared desire to resolve these issues, but no project lead had taken responsibility. In 2012, the mental health trust funded a new facility on the Clearview site, ‘a better environment’ where their staff could undertake assessments, but no changes were made to staff numbers or other working arrangements. One solution concerned the appointment of a mental health nurse or specialist social worker based at Clearview. From the acute trust's perspective, however, this involved covering a staff shortage for the mental health trust. It was proposed to fund a ‘pilot’ mental health nurse cover in the emergency department during twilight hours, along with a new rapid assessment model. The provision of on-site consultant psychiatric cover was considered to be more difficult to achieve.

The main impediments to further change in this case thus appear to concern:

• two organisations with differing views on the conditions that should trigger a psychiatric assessment and the circumstances under which such an assessment should take place

• lack of funding for additional staff and the unwillingness of one organisation to contribute to the staff costs of the other

• mental health trust concern about triggering conflict were they to suggest evening and weekend rotas to staff

• two organisations with different information systems and separate secure access codes.

Patients requiring psychiatric assessment continued to present in the emergency department at Clearview. In 2012, another incident similar to the case of Mr Torrens occurred, but this was still under investigation at the time of writing.

Mr Mitcham

I hate walking the floor now. I see bombs waiting to explode everywhere.

Extreme events workshop participant, Clearview

Mr Mitcham, 66 years old, came to the Clearview eye unit in October 2011 for cataract treatment involving the implant of an intraocular lens under local anaesthetic. He was the last patient having this treatment on the afternoon theatre list and he was patient number 7. However, patient number 6 arrived late and was not ready for surgery. Mr Mitcham was thus taken into the theatre by the anaesthetic practitioner as the sixth patient. The operating surgeon was not made aware of this change in running order. Believing that she was operating on patient six, she implanted in Mr Mitcham the lens for that patient, only realising the error when completing the operation note after the procedure. The consultant was informed immediately and Mr Mitcham was taken back into theatre where the incorrect lens was replaced without problem. He was discharged that day and suffered no permanent harm. The hospital serious incident management group decided that this met the criteria for a ‘never event’ and an investigation was conducted.

Why did this happen? The anaesthetic practitioner (senior nurse and team leader) had arrived in the theatre half-way through the list and had missed the team brief at the start. Although aware that patient number 6 was not ready, she did not explain the change in the running order to the other theatre staff. The operating surgeon then failed to check the patient's wristband, assuming that she had the correct patient. The surgeon addressed Mr Mitcham as patient number 6 when he arrived in theatre, but he did not correct her. It had become normal practice for operating surgeons to identify patients on the ward, at the consent stage, and not to repeat this identity check in theatre. However, patients listed for cataract surgery were placed on a generic list, which meant that the operating surgeon may not have seen a particular patient before the day of surgery, which is what happened in Mr Mitcham's case.

The investigation concluded that the root cause of the incident was a failure to follow normal checking procedures, which included the World Health Organization (WHO) theatre checklist. The surgeon was experienced but had been working at Clearview only since the beginning of that week, and this was her second full theatre list. During her induction, guidance on the use of the WHO checklist, and Clearview theatre procedures, had not been clear (hospitals apply this checklist in different ways). The first five patients on the list that afternoon were operated on without incident.

The main recommendation following this incident concerned the correct use of the WHO checklist. Induction processes in ophthalmology were reviewed and the importance of following checking and handover procedures was reinforced. As normal, to provide an impartial assessment, the incident investigation was conducted by the hospital's assistant director of risk and patient safety and a patient safety manager. They were advised by a Serious Incident Management Group, which included a consultant ophthalmologist, the ophthalmology unit leader, two operations managers, a divisional lead nurse, the head of medical staffing and the Patient Advice and Liaison Service (PALS) manager.

There had been four ‘never events’ in Clearview theatres over as many weeks in 2011: a retained clamp in a patient's neck following surgery, cardiac arrest during sight-saving surgery leading to an adverse neurological outcome, a retained ‘fish’ following abdominal surgery and the wrong implant for Mr Mitcham. Following a long period in which no such incidents had occurred, the theatre teams had a ‘high performance – high reliability’ reputation and were surprised and shocked that these events had occurred and were concerned that this should not happen again. Three members of the research team were thus invited to run a workshop in March 2012 on ‘managing change after extreme events’ for 11 members of the theatre management group. The workshop aim was to provide fresh perspectives and tools to help participants develop solutions. One of these tools concerned the development of ‘mess maps’ to help explain why these events occurred. 215

Workshop participants were asked to develop a mess map for each of the four never events, using flip charts and ‘Post-it’ notes, identifying immediate causes, along with the role of underpinning issues and contributory factors, the external context, local conditions and other reasons. 206 The mess map for Mr Mitcham is shown in Figure 7. This echoes the findings of the investigation: new surgeon, inadequate induction, poor communications, lax checking procedures. The mess map also identifies four other sets of contributory factors, not picked up by the incident investigation:

• External pressures. The map notes the 18-week RTT target and the need for external reporting of never events, even though in this case no patient was harmed.

• Internal pressures. The morning clinic was busy, surgeons were under pressure ‘to get a move on’, surgeons arrived late, staff missed their lunch break, surgeons had complained that afternoon lists were too long.

• Personal preferences. Medical staff behaviour could be idiosyncratic, ‘everybody does their own thing’, the new surgeon had a different lens-matching practice, the consultant was doing paperwork at the time.

• Smooth running. Paradoxically, this incident may have occurred in part because things were going well – Friday afternoon, experienced staff, mundane process, five cases routine, changing list order is commonplace; did a successful theatre team ‘drop their guard’ under these conditions?

FIGURE 7.

Mess map of the Mr Mitcham never event. EPR, electronic patient record.

This mess map, produced relatively quickly during a workshop by staff who were involved in the incident, appears to offer a richer explanation for the incident than the carefully researched investigation report, and reflects more closely the systems-theoretical model of incident causality explained earlier.

In the aftermath of future incidents of this kind, one option would therefore be to complement impartial RCA with a mess mapping process involving the staff concerned, to develop action plans to which those staff have themselves contributed. There was further internal support for the conclusions of this mess mapping exercise. The trust had formed a Never Events Group, to identify actions to prevent further incidents and to give staff ‘the opportunity to comment on their view of life in the theatres and any factors which might impact on their ability to deliver the highest standards of care to our patients’. This second aim was achieved through ‘listening events’ in November 2011, with 10 senior staff acting as facilitators and which involved over 100 staff, who welcomed this opportunity to share their views. The listening events report categorised the feedback as follows:

• procedures and systems: not enough time to spend with patients

• culture in theatre: bullying, racism, spinning out procedures to fill the time

• communication and team interaction issues: inconsistent and contradictory messages

• leadership: lack of engagement with staff, watching not experiencing

• training: non-existent induction processes, people moving too quickly

• staffing issues: lack of porters, ratio of trained to untrained staff cut to critical

• time pressures: many issues only apparent on morning of operation, WHO checklist

• equipment and instrumentation: no systematic review of sets, wrong instruments

• other: incident handling criticised, safety events held at times when most cannot attend.

Despite evidence pointing to organisational conditions and system properties, the report concluded:

It should be remembered that the four never events were, to varying degrees, the result of individuals failing to follow procedures and good practice. However, if any of the factors identified in the listening events acted as contributory factors then action needs to be taken to address these issues.

Individual human failures were thus seen as the primary explanation for these never events. This conclusion makes ‘the fundamental attribution error’ of blaming individuals and overlooking the context in which they work. 216 The conclusion also overlooks the trust's own evidence concerning ‘life in the theatres’. A more comprehensive approach to understanding why this incident happened is illustrated in Table 33, which locates the combined findings from the investigation, the theatre team's mess map and the listening exercise in Rasmussen's ‘problem space’. 209

Although thorough induction training and verification processes are important, to reduce the future incidence of never events in theatres this analysis points to a wider change agenda, addressing organisation, management, staff, infrastructure and process issues (some of which were subsequently addressed). Viewed from a systemic perspective, it is also possible to reach beyond one incident, and consider what can be learned from incidents like this.

Managing change in wicked situations

We are good at reporting, we're good at investigating, we're good at making recommendations. But it all falls apart in the implementation. Either things don't get implemented at all, or they get implemented, and if you go back one year, two years, three years later, it's slowly tailed off, to a point where nobody quite knew why we were doing it anyway, and something else has come along.

Assistant director, Clearview

The aim of this ‘small-n’ analysis of four idiosyncratic cases is not to generalise to the population of such events, but to inform theory and practice, particularly with regard to the implementation of changes aimed at preventing or reducing the incidence and impact of such events in the future. This involves a combination of analytical refinement and naturalistic generalisation. 35,37 The former concerns developing understanding of theory, refining current assumptions and concepts, whereas the latter concerns the transference of ideas and solutions to similar contexts. (A fifth case, a norovirus outbreak at Wattle Park in which the incident definition was ambiguous, is described in Appendix 4.)

The management of extreme events concerns the linkages between an understanding of why an event happened (causality), the changes that need to be implemented to defend against a recurrence (agenda) and the implementation of those changes (process). Table 34 outlines the findings from the four events reported here with regard to those issues and to the outcomes.

Managing change in such settings is problematic. In breach of conventional change management guidelines, staff who are to be affected are excluded from contributing to the change agenda, other than as witnesses giving evidence, and the implementation process is more likely to be directive than participative. Borrowing the concept of ‘wicked problems’,217 this could be seen as managing change in ‘wicked situations’ with the properties outlined in Table 35.

This can become a ‘super wicked situation’ if:

• There is more than one organisation involved and their perspectives differ.

• This was a repeat of at least one previous incident, so past solutions have not worked.

• The media picked up the story; they are looking for a scapegoat and demanding action.

None of the events reported here attracted significant media attention. The death of Mr Torrens appears to have created a super wicked situation, given the organisational tensions involved. Drawing on the successes and problems arising in those incidents, the management of change in wicked situations may involve developing context-sensitive approaches with these properties:

• Agenda. Go beyond quick fixes and address underlying system issues; use this as an opportunity to address long-standing organisational problems; develop a progressive change agenda, linked if necessary to a defensive one; restore the sense of urgency that accompanied the incident in the first place; receptiveness to change may not be high, and may require stimulation; establish meaningful metrics to track progress.

• Process. Secure long-term senior management support and resourcing; structure a project team, steering group or board that can act on their decisions; engage colleagues who were excluded by the investigation process; resolve competing views of event causality and appropriate responses; address resistance to the ‘difficult’ changes to roles and working practices; design a communication strategy that will inspire as well as inform; aim to sustain changes beyond initial implementation and success.

Two environmental issues may also be important in some settings. One concerns handling the media reporting the story, who can influence the change agenda. The other, as in Mr Torrens' case, concerns managing interorganisational tensions and differences of perspective. In wicked situations such as those found here, it may therefore not be adequate, as some commentators have suggested,205,213 simply to provide risk management personnel with development in change management capabilities, or to have change management specialists on investigation teams. Investigation processes are currently driven by specially formed, impartial groups, using prescribed templates, aiming to produce a ‘quality report’, including practical recommendations, within a set timescale. What may also be helpful, instead of or perhaps in parallel with that formal approach, is a process driven by the staff involved in an incident, using relatively simple and flexible ‘mess mapping’ and other appropriate tools, to identify immediate and systemic changes that will have a sustainable impact.

Serious incidents, never events and other types of accident and failure can be regarded as inevitable features of complex systems. Appropriate responses to this observation thus include constant scanning for warning signs, developing flexible, context-sensitive working practices and helping those involved to deal with ambiguity and to make decisions that are often based on unclear and evolving information. Is it possible to organise in ways that allow problems to be anticipated and caught before they become crises, and to develop systems for containment as well as for effective responses? These are the goals of researchers and commentators advocating high-reliability organisation characteristics. 203,218

Given the complex, systemic, organisation-wide antecedents of these kinds of incidents, it is perhaps not surprising that it has already been suggested that high-reliability concepts can be adapted to health-care contexts. 219 Originally based on studies of aircraft carriers and nuclear power installations, the qualities of high-reliability organisations include mindful preoccupation with failure, reluctance to simplify interpretations, sensitivity to operations, commitment to resilience and deference to expertise. Expertise often resides with front-line staff who are close to the flow of events, and more senior staff must be prepared, when appropriate, to relinquish command. Developing a high-reliability approach requires a sustained organisation-wide programme, rather than managed fixes following specific incidents. With the priority attached to improving the quality and safety of patient care while reducing costs, the systematic tailoring and evaluation of high-reliability methods in acute health-care settings would constitute a valuable research agenda.

Introduction

The focus of my role lies with clinical and managerial strategy and service business plans. The aim is to clarify thinking for the future. I work closely with the medical staff, to bring another set of skills. A lot of my work is commissioned in the corridor. Every service has to develop a formal vision and strategy, with investment objectives. They are my customers, and I help them to develop these plans. With new developments, I ask them, ‘what's the story?’. Often it's about bringing people into the room, exploring their positions, and reconciling these. With one of our services, it took a year to get a consensual story; there were big issues on which different people had different views. So this is a mediating and facilitating role. Sometimes this even involves ‘proximity talks’ when relationships are poor. I'm a quiet spoken pushy bastard. One consultant calls me ‘the spin doctor’. I'm good with words, reframing, convincing. I like getting results. It doesn't always work.

Assistant director, Clearview

This study sought to develop a profile of middle management contributions and to understand:

• how management work affects clinical and organisational outcomes

• factors impeding management contributions to performance

• how an ‘enabling environment’ for management work can be assembled and sustained.

This chapter first reviews past research concerning middle management contributions in general and in health care, before exploring the combined evidence from this project. The main conclusions are:

• Previous research shows middle managers in general contributing to organisation strategy and performance, through mediating, boundary spanning, innovating, championing, influencing and change implementation roles. Past research offers a similar account of middle managers in health care, with human resource, operational and performance management practices being linked to improved patient outcomes and organisational performance.

• From this study, the profile of middle management contributions includes maintaining day-to-day performance, firefighting, ensuring a focus on the patient experience, identifying and ‘selling’ new ideas, translating ideas into working initiatives, process and pathway redesign, facilitating change, troubleshooting, leveraging targets to improve performance, developing infrastructure, developing others and managing external partnerships.

• Past research indicates that supportive cultures combine flexible roles and structures, freedom to innovate, resources and senior management support. Conditions that constrain middle managers include systems and policies that encourage conservative behaviour, complex approval cycles, micromanagement, lack of autonomy and unclear direction from the top.

• The components of an ‘enabling environment’ for management work include top team communications, business intelligence, cutting non-value-adding activity, autonomy to innovate, organisation structures, organisational norms, performance management, interprofessional work, support services, personal development, teamwork and resources; investment in information systems, management development and additional resources can be costly, but action to improve the enabling environment under other headings is cost neutral.

• Evidence from this study also indicates that encouraging co-management – shared leadership of change – by clinical and managerial staff working collaboratively could have major benefits for clinical and organisational outcomes. Several other low-cost, low-risk, high-impact suggestions for building an enabling environment were generated. At a time of rising demand and diminishing resources, it is not clear why these were not more common in practice.

The strategic middle manager

Managers make a huge contribution to patient care, developing services, measuring quality. One of the key things that I find myself doing is drawing ideas together and implementing them. I go to all the clinical meetings, and I'm heavily involved. I also attend consultants' meetings. I act as a linchpin. If I'm not there, the discussion and outcomes can become more fragmented.

General manager, South Netley

Evidence from this study

This section first explores the pressures and demands on middle management and then considers contributions to clinical and organisational outcomes.

What has to change?

Interviewer: Red tape, not allowed to use your experience, not trusted to manage – I thought the trust was developing a culture that encouraged engagement and empowerment?

Respondent: Well it's the usual kind of conflict between the two. Politically correct, and good management processes, encouraging empowerment, engagement, and so on. But, when money is tight, you start getting micro-managed. So things like duty rosters – I can't trust my ward managers now, my ward sisters, to do a duty roster. It's got to come from me, and I've got to sign it off. Because I'm told that I must sign off rotas.

Lead nurse, Greenhill

As organisational problems mount, and new ideas and rapid solutions are at a premium, the senior management reflex is often to centralise decisions, to tighten controls, to limit autonomy. As an associate director at Greenhill observed, ‘We need to “let go”, and that's incredibly difficult at a time when we feel the need to tighten the reins’. However, those responses stifle innovation and demotivate those who are close to the problems but who are constrained in the actions they can take. In this ‘disabling’ setting, middle managers in this study were asked how their contributions to clinical and organisational outcomes could be strengthened. There was no shortage of answers. This section summarises those answers and outlines a model of the ‘enabling environment’ (Table 39).

Most actions to support an enabling environment for management work are cost neutral. Only three of these dimensions are cost generating: business intelligence, personal development and resources.

To represent the richness of the evidence, we will first summarise the content analyses of interview responses from two trusts and then present typical responses from focus group discussions. Suggestions for strengthening management contributions displayed similar patterns irrespective of trust or respondents' roles. So, for example, content analysis of responses from 17 interviewees at Netherby generated 45 ideas, highlighting suggestions for appropriate action. Most suggestions concerned top team communications, autonomy to innovate, organisational norms and performance management. Table 40 shows an illustrative selection.

The 12 interviewees at Greenhill generated 56 ideas. Similar to the Netherby responses, this pattern is illustrated in Table 41, again emphasising the importance of top team communications, autonomy to innovate, organisation norms, performance management, organisation structure, personal development and adequate resources.

Focus group participants were asked the same question: ‘what would have to change in order to strengthen the contributions that you as a manager can make to quality of patient care, and to clinical and organisational outcomes?’ The pattern of responses was consistent with those from interviews. Here are just three examples from the 40 focus groups that were conducted during this project.

One focus group at South Netley included matron neurosciences, assistant general manager medicine, quality manager, human resources partner, assistant general manager, physiotherapy team lead trauma and orthopaedics, neurosciences physiotherapy team lead, assistant general manager surgical services, advanced nurse practitioner neuroscience, physiotherapy manager and leadership development consultant. From a long list of ideas for strengthening their contributions, their priority areas were:

• ‘clarity and priority around goals; clear strategic direction’

• ‘senior management speak transformational then act transactional’

• ‘give us space to achieve; freedom to try things out, without the bureaucracy’

• ‘clarifying expectations for ward staff, particularly responsibilities of band 5 nurses’

• ‘norms and culture get in the way, but we know these are hard to change’

• ‘performance management – the NHS is not accustomed to this – so when people don't deliver, there is little or no accountability; but performance management can be supportive’

• ‘human resource policies for proper sickness, absenteeism and performance management’

• ‘need to develop coping strategies, survival skills’.

Once again, top team communication, autonomy to innovate, organisation norms, performance management and personal development were significant. This focus group also suggested ‘more information, fewer rumours’, ‘less whizzing around operationally’, ‘change in culture away from command and control’ and ‘less struggling to understand the analytical information we're given’. They also suggested that senior management ‘come and work with us, to understand what's going on’, and wanted to see the ‘executive team modelling the behaviour that we are expected to demonstrate’ (they were no longer allowed to use bank staff whereas the executive team hired external management consultants). Reinforcing earlier discussion of clinical–managerial collaboration, this focus group also expressed the desire for ‘an increased culture of clinical–managerial joint working’.

A second example concerns a focus group with four modern matrons at Greenhill who identified the suggestions in Figure 8, indicating what they felt there should be more of and less of. Supporting the final comment in Figure 8, a ward sister in another Greenhill focus group said:

I tackled the CEO in the meeting and got nothing. I e-mailed him, and I got a wishy-washy response. And when I challenged him on the issue he was rude to me. Communication from the top down has to improve. It's almost as if they have a hidden agenda.

FIGURE 8.

What has to change: modern matrons at Greenhill.

Our third focus group example concerns the 10-member operations management team in the medicine division at Clearview, including operations managers, deputy operations managers, division support managers, divisional lead nurse and senior clinical nurses. Their suggestions for strengthening the management contribution covered individual, divisional and corporate actions.

At an individual level, time management, developing self and others, ‘being in the know’, role clarity and being more proactive and risk-taking were seen as important actions: ‘I get involved in the nitty gritty, I should not have to’, ‘more time with line manager to emulate skills’, ‘keep up with new programmes, be aware’, ‘review job roles and tasks and how they need to change’ and ‘move from being reactive to proactive and take more risks, more power to act with patients, use judgement’.

At a divisional level, ‘being in the know’ was again important, along with meetings management, role clarity, learning and sharing practice, and human resource policies: ‘we should all take responsibility to stay informed’, ‘ensure that groups, forums, workstreams have consistent attendance from the same people, not different people each time, which reduces engagement’, ‘clear roles and responsibilities; understanding of roles of others’, ‘learn from the firefighting incidents, and share best practice more across the division’ and ‘enable staff to do the job and tighten up human resources processes’.

Corporate actions recommended by this focus group included improved central human resource support for recruitment, more opportunities for formal and informal networking with peers, better communications between divisions, better clarity of roles, the celebration of errors as learning opportunities rather than a blame culture, support services with a ‘customer orientation’, dissemination of best practice – a ‘learning and sharing organisation’ – stronger trust values with regard to respect for and engagement of staff, and reduced conflict between finance and patient safety.

It is significant to note that these ideas concern aspects of personal style, approach to the job, time allocation, the clarity with which roles are defined and understood, responsibilities, communications, sharing, learning, support, networking, values. None of these suggestions imply the need for additional resources. Other than the potential opportunity costs involved in acting differently, these suggestions for strengthening management contributions would be free to implement, and would have potentially significant positive implications for clinical and organisational outcomes.

Space limitations do not allow a full reporting of the responses from all 40 focus groups, which raised similar themes and issues. Shown against the same 12 themes, Table 42 lists the 24 ‘ways to strengthen the management contribution’ codes from the combined focus group content analysis. Translating this information from interviews and focus groups into guidelines for management practice, corporate issues that appear to be of particular significance and concern include:

• clarity and consistency with regard to trust goals and priorities

• clear and consistent communications that keep staff well informed

• support services with a customer orientation providing effective and timely support

• updated IT systems that provide appropriate financial and other management information

• strengthened performance management systems

• investment in leadership and management development

• encouragement and support for clinical–managerial collaboration.

Board-level executives may in addition wish to consider:

• not interfering in operations; focus instead on planning and prioritising

• avoiding a ‘panic of the week’ approach

• listening to middle managers who have a better understanding of operational issues

• involving middle managers in key decisions

• empowering middle managers to make decisions, giving them ‘space to achieve’

• emphasising positives rather than negatives.

Steps that individual managers said that they found useful included ‘get out and about’, ‘make time for reflection’, ‘look at what processes and practices could be done differently to reduce the pressure’, ‘avoid perfectionism’, ‘find support in colleagues’ and ‘develop persistence, political skill and a thick skin’. In sum, be more proactive, reflective, self-aware and resilient.

A contributions-based model of management

This study aimed to develop a profile of middle management contributions, to understand how clinical and organisational outcomes are affected, to identify factors impeding management contributions and to establish the dimensions of an ‘enabling environment’ for management work. In this field, one research tradition focuses on management roles and activities. A second tradition concentrates on linking aspects of management work to organisational outcomes. In other words, an interest in what managers do sits alongside a focus on what managers achieve – the added value. This study aimed to contribute in particular to the latter tradition, not by attempting to correlate specific management practices with performance metrics, but by profiling the contributions of the middle management function, and demonstrating how these are shaped by institutional and organisational contexts.

Figure 9 summarises the overall argument of this project, drawing as indicated on the conclusions of previous chapters. This is loosely based on the model developed by Wooldridge et al. 30 Their ‘organising framework’ is designed to inform research rather than to explain how the management contribution is enabled and constrained, and indicates the significance of institutional and organisational contexts without detailing their properties. Figure 9 in contrast shows how middle management contributions in health care are shaped by national institutional and local organisational contexts, by the composition of the management function and by the enabling (or disabling) properties of the environment in which middle managers operate.

Previous cross-sector research portrays middle managers contributing to organisational outcomes through influencing strategy, by nurturing and championing initiatives and through co-ordinating, mediating, interpreting and negotiating roles. The advantages that middle managers have in driving innovation and change are based on the combination of access to top management and knowledge of operational capabilities. Research also indicates that middle management involvement in strategy leads to improved decision-making, higher degrees of consensus, improved implementation and better organisational performance. There is no reason why these observations concerning middle management work in general should not apply in health-care settings.

Factors inhibiting ‘entrepreneurial’ middle managers include policies that encourage conservative behaviour, complex approval cycles, micromanagement and lack of delegated authority – all of which were observed in this study. Factors fostering innovation include overlapping territories, cross-functional contacts, free flows of information, excess in budgets, loosely defined roles and contact with other organisations. These ‘propellants’ were less evident, and their absence may be sustained by the divisional silos that accompany service-line management structures.

FIGURE 9.

A contributions-based model of the acute trust middle management function. The arrows between the components of this model indicate influence and dependencies (one way or mutual) and are not intended to represent causal relationships.

The concept of the enabling environment is not novel. For example, Burns and Stalker240 showed that ‘organic management systems’ operated more effectively in turbulent environments than bureaucratic or ‘mechanistic’ systems. Kanter241 argued that ‘integrative’ cultures encouraged innovation and adaptability, whereas ‘segmentalist’ (bureaucratic) cultures did not.

Several commentators have examined the related concept of ‘high-performance’ work systems,242,243 which involve clusters of practices designed to increase capabilities, commitment, motivation and individual and organisational performance. Huselid244 linked high-performance work practices to higher productivity, lower staff turnover and higher corporate financial performance, such practices contributing US$18,500 per employee in shareholder value and $4000 per employee in additional profits. Pfeffer245 claimed that high-performance practices could raise an organisation's stock market value by US$20,000–40,000 per employee through policies such as job security, careful recruitment, decentralisation and self-managing teams, remuneration linked to organisational performance, investment in training, reduced status differentials, and information sharing across the organisation. Popular in textbooks, high-performance practices have been less common in practice. 246

High-performance perspectives have been applied mainly to shop floor operatives in manufacturing. Thorlby and Mabin247 ask whether or not the NHS has become a high-performing system and, although pointing to improvements, their analysis focuses on conditions and metrics (cancer, cardiovascular disease, drug costs, survival rates) and not on the organisation conditions and management practices that produce those results. However, Michie and West236 develop a ‘high-performance’ framework for health care, linking patient outcomes and organisational performance to people management practices that include job design, teamwork, involvement, control over work, leadership, and training and development. The ‘enabling environment’ concept is also related to the notion of organisation culture. The latter has attracted multiple definitions and is typically operationalised in terms of organisational properties. In contrast, the notion of enabling environment in the context of this discussion is grounded in specific policies, behaviours and practices. From their study of organisation cultures in the NHS, Mannion et al. 248 found that the service was becoming more hierarchical and competitive, but that the links between culture and performance were difficult to disentangle. Haq249 reports the results of a survey of 132 managers attending NHS leadership programmes, concluding that only 9% were creating a high-performance environment in which teams felt motivated and focused, whereas 77% were creating a ‘tolerable or demotivating’ climate, resulting in disengagement, minimal discretionary effort and underperformance.

Davenport and Harding250 argue that a key senior management responsibility concerns the creation and maintenance of conditions in which others can be successful, establishing ‘the context for performance’. Pointing to ‘the competitive advantage in the middle of your organisation’, they argue that middle managers are repositories of experience, insight and influence, with accumulated operational and organisational knowledge, managing operations, developing others, keeping the show on the road, energising change – consistent with middle management roles in health care. Our concern in this project lies with an organisational environment that enables middle managers to carry out those operational, problem-solving, service improvement and change leadership roles more effectively. The evidence indicates that organisational environments in acute trusts could be more enabling and supportive. Many of the actions required to strengthen those environments are cost neutral, which is significant given the expectation that the service faces a continuing ‘funding ice age’. Hurst and Williams251 therefore ask if NHS hospitals can do more with less, pointing to the importance of leadership, management, staff engagement, improved operational processes and new technology adoption. The ‘enabling environment’ in Table 39, therefore, can be read as a ‘high-performance profile’ for middle management work. To strengthen contributions to clinical and organisational outcomes, senior executives may wish to consider the performance dividend that would result from implementing cost-neutral policies and practices such as:

• ‘high-visibility’ board members setting and communicating clear goals and priorities

• listening to and acting on ideas from staff at all levels

• remembering to praise the positives

• ensuring that management roles are seen to be recognised and valued

• streamlining internal governance arrangements and information requirements

• where possible simplifying the provision of information to external bodies

• providing appropriate and timely business and management information (activity levels, capacity, costs, outcomes), perhaps exploring simple low-cost IT solutions

• granting middle managers authority to fix problems within their own budgets

• allowing middle managers to act on their own initiative to pilot innovations

• creating opportunities for the exchange of information and ideas with other organisations

• eliminating or at least reducing the sign-offs, which generate frustration and delays

• designing a rigorous and supportive performance management system with appropriate professional support for managers who operate the system

• ensuring cross-divisional information-sharing and improvement initiatives

• encouraging clinical–managerial collaboration, to foster trust, respect and mutual understanding, and to drive major change initiatives

• insisting that support services – human resources, finance, IT, estates – adopt a customer orientation

• opening up opportunities for leadership and management development for those with or aspiring to hold managerial responsibilities

• implementing team-building activities; good teamwork does not always develop naturally

• considering ‘invest to save’ projects in which spending will improve future cost-effectiveness.

These are surely ‘common sense’ actions, but the evidence suggests that they are not common in practice. In a context of rapid, radical and ongoing change, with demand rising and funding falling, in which new ways of working are at a premium, why would an organisation not consider taking such low-cost, low-risk, high-return steps? Senior managers, perhaps, in restricting the autonomy and voice of middle managers, are simply passing on down the hierarchy the pressures that they experience from national bodies. Another explanation perhaps rests with the composition of the middle management function, dominated by hybrids with little management training, and who may not be trusted by more experienced executives to make appropriate management decisions.

As well as building an enabling environment, these suggestions have other benefits. First, capable, motivated and committed managers are more likely to stay with an organisation that listens to their ideas, allows them to be innovative, to make a difference. In other words, these guidelines constitute an effective retention strategy. Second, these steps generate low-cost, in-house, tailored development opportunities. Many middle managers participating in this project indicated that their development opportunities had been curtailed by budget cuts that prevented attendance on generic external programmes. Third, clinical staff may be more willing to assume managerial responsibilities, and to discharge these enthusiastically, if they know that they will be allowed ‘space to achieve’. Finally, all of these factors contribute to management capacity, an issue raised in Chapter 2 and a research priority arising from this study. Building and sustaining an environment that enables management work contributes to capacity, motivation, retention and development, and increases the attractiveness of pure and hybrid middle managerial roles.

Management in the plural

Our board-level contact at one trust was asked to comment on the enabling environment concept. He replied that this model was helpful, but requested a simpler, more focused approach. His executive team would want to know ‘the two or three things that would make a big difference, the small number of high-impact changes’. In considering how to respond to that request, one theme appears to be particularly prominent: the comanagement, or shared leadership of change, by clinical and managerial staff, hybrids and pure plays, working in collaboration.

Management, as with leadership, can be defined as a solution to the problem of collective effort. 252 As we have seen, hospital management is a widely distributed function, dominated by hybrids who outnumber pure plays by four to one. The capacity of that function to address effectively the multitude of operational and strategic demands placed on it is therefore crucial. This appears to require what might be termed ‘management in the plural’, drawing on the collective resources of managers with different backgrounds, skills and views.

The concept of distributed leadership, or ‘leadership in the plural’,64 recognises that leadership behaviours can be seen at all levels of an organisation. The traditional concept of focused leadership emphasises the attributes and capabilities of senior individuals. Distributed leadership concerns the achievement of goals through collective action, which may be formally designated, but can also develop spontaneously. 253–255 The notion of distributed leadership is endorsed in the NHS Leadership Framework28 and in the review of leadership and management conducted by The King's Fund. 42 Denis et al. 256 argue that a ‘plurality of leaders’ may be necessary, as no one individual acting alone can combine the influence, expertise and legitimacy needed to drive change in complex social systems. In their review of research in this area, Denis et al. 64 also explore the merits of conceptualising leadership as a collective phenomenon:

It is the common experience of life within and beyond organisations that leadership and leaders are to be found in many places. In a shared power world, plural forms of leadership where different people bring different resources, capabilities, and sources of legitimacy (and yes, followers) to the table offer a path to getting things done: not an idealized path, and not a path that always succeeds, but a key component of organising nonetheless that needs attention from organisational scholars.

p. 6464

Terminology in this field is also distributed. Fitzsimons et al. 257 identify shared, distributed, dispersed, devolved, democratic, distributive, collaborative, collective, co-operative, concurrent, co-ordinated, relational and coleadership, and then distinguish between leadership that is shared (among those holding formal leadership roles) and leadership that is distributed (performed by many across the organisation). Martin et al. 258 distinguish between ‘quiet distributed’ and ‘quiet dispersed’ forms of leadership, the former being within the control of the project in hand, the latter being more diffuse. Currie and Lockett259 explore concertive, conjoint and collective forms of distributed leadership, identifying a spectrum between top team models at one extreme and the spontaneous collaboration of actors pooling their expertise and skills at the other.

What of shared or distributed management? Mintzberg4 introduces these terms in a section on ‘managing beyond the manager’ (p. 147). By ‘shared managing’, he means ‘one managerial job shared among several people’, which is not how the concept of shared leadership is understood. ‘Distributed managing’ concerns the diffusion of responsibility for some managerial roles to various non-managers in a unit (p. 152). Both concepts appear to work more closely with formal organisational positions and hierarchies, and do not address the more fluid, relational, interactional and processual ‘leadership relays’ with which Denis et al. 64 and others have been concerned.

Denis et al. 64 cite examples of distributed leadership, including in health care. 11 They note that these cases ‘do not provide evidence that such forms necessarily lead to positive outcomes or that success might not have been obtained in other ways’ (p. 39). We have reported in this chapter, however, several instances in which change and service improvements may not have occurred, or would not have been implemented so rapidly, had some form of shared leadership – or shared management – not been invoked to make that happen.

Management in the plural is a ‘high-impact, high-performance’ response to our executive. This concerns two components of the capacity model: requisite variety and clinical–managerial relationships. Pure plays and hybrids bring contrasting views to the complex problems that they face and share. There have been several reported experiments in joint working. These include the NHS Institute Duality Leadership Programme,260 the NHS London Prepare to Lead scheme, the North Western Deanery leadership development programme and the London Paired Learning initiative. 53,261 Can shared management, in different forms, develop from isolated examples to widespread norm? Collaborative approaches, particularly in the context of diminishing resources, may be key to identifying, designing, establishing consensus for and implementing the more innovative SDO changes necessary to improve effectiveness, quality and safety.

This discussion points to a challenging research agenda. Currie et al. 262 show how distributed leadership is diluted by bureaucracy, power differentials and a centralised performance management regime focusing on a small number of individuals. Denis et al. 64 note that distributed leadership may have negative outcomes, and research has not explored the discord, conflict and rivalries that might arise. Gronn254 recognises the continuing importance of hierarchical leadership alongside distributed forms. Currie and Lockett259 also argue that enthusiasm for distributed leadership may dilute the role of hierarchy. We therefore need to develop a better understanding of management in the plural. In which health-care contexts are such approaches appropriate? Which models make the biggest difference, to those involved, to organisational outcomes and to patients?

Contribution of authors

Publications

See Appendix 5 for project outputs and Appendix 6 for related publications.

Disclaimer

This report contains transcripts of interviews conducted in the course of the research and contains language that may offend some readers.

Research protocol Version 1: 09.02.2009

Cranfield University

School of Management and Cranfield Health

NHS National Institute for Health Research Service Delivery and Organization Research Programme

241 Management Practice in Healthcare Organizations: proposal #0001926

How do they manage?: research protocol

A study of the realities of middle and front line management work in healthcare

Three and a half years, starting January 2009

Research team:

Professor David A. Buchanan

Dr Catherine Bailey

Dr David Denyer

Dr Clare Kelliher

Ms Cíara Moore

Dr Janice Osbourne

Dr Emma Parry

Dr Colin Pilbeam

Dr Janet Price

Professor Kim Turnbull-James

Dr Charles Wainwright

Scientific summary

Managers constitute 3 per cent of the NHS workforce. That figure underestimates the impact of management practice on clinical outcomes, quality of patient care, and organizational performance. The NHS has concentrated on senior leadership, and less is known about the experience and attitudes of middle and front line managers in acute care, who are the focus of this project. Exploring the realities of management work, their role in change, and links between practice and performance, this study has four aims.

The first is to contribute to the practice and theory of healthcare management in order to improve patient care and organizational performance. The second is to provide evidence-based guidance for management development, strengthening the impact of management practices on hospital performance, streamlining the implementation of changes following adverse events in the interests of patient safety.

These first two aims will be achieved with new perspectives, approaches, frameworks, diagnostics, methods, tools, and processes based on new evidence. We will identify the organizational features that support managers in contributing to clinical and corporate performance, building an ‘enabling environment’.

Our third aim is to engage stakeholders as co-researchers through our collaborative research design. In addition to respondent validation, this approach will develop ‘high impact’ channels for communicating the implications of findings. Our final aim is to develop the theory of managing, synthesizing current models, theories of distributed leadership, and processual-contextual perspectives on change.

Our collaborative research design involves six acute trusts over six stages: (1) set up (research assistant, background information, literature review, ethical approval), (2) management focus groups, (3) management survey concentrating on the themes of realities, changes, and contributions, with 60 per cent of items common for all sites for comparison, and 40 per cent based on local trust issues and priorities, (4) management briefings to check findings, explore implications, consider diffusion mechanisms, and identify cases for the next stage, (5) case studies of change following adverse incidents, and (6) publication and knowledge transfer. We will also track changes in the management role in one PCT, linked to one of the acute sites, exploring through interviews with a small number of key informants (middle and senior managers) the implications for inter-organizational relationships, and the impact on acute management roles. Research methods thus include document analysis, focus groups, self-report survey questionnaires, interviews, and case studies of ‘extreme change’. Analysis methods include context profiling, content analysis, statistical analysis, visual mapping, event sequence analysis, and ideas capture from briefing groups.

Outcomes can be measured in terms of service impact. For patients and service users, this concerns management practices that will improve quality of care and clinical outcomes, and rapid changes following ‘extreme’ events leading to improved patient safety. For middle and front line managers, this means a better understanding of how the role is evolving, new competency requirements, methods for influencing clinical and organizational outcomes, and techniques for managing ‘extreme’ change. For senior managers, we will provide guidance on management development and support needs, and advice on developing an ‘enabling context’ for the management impact on clinical, organizational outcomes, and change. For policy makers, this research will deliver a model of management work, explaining the demands and pressures, the new competencies required, the contributions to change and performance outcomes, and the implications of extending clinical engagement in management.

Lay summary

How do hospital managers handle the pressures and demands of a constantly changing health service? What effect do managers have on the quality of patient care and the outcomes of treatment? We know little about the work experience and attitudes of hospital managers, but when things go wrong, this is the group which usually takes the blame. Patient safety is a national priority, and we particularly want to find out how changes to working practices are managed after serious incidents. This can be a problem, as the advice of enquiries, in health and elsewhere, can often sit on the shelf. This study will build on what we already know about the realities of middle and front line management work and organizational change. We will collect the information we need using focus groups, a survey, and interviews which will enable us to develop case studies of serious incidents and the changes to which they lead. We will also ask the middle and front line managers involved in this study to help us with the design of the survey, and with choosing examples of serious incidents. As participating managers are among the main users of the findings from this project, we will ask them to check our understanding of the data, and to help us to develop novel ways to communicate the implications, so that this work does have an impact on management practice. The study will do this through advice on management support and development based on a better understanding of how managers work and of the factors that are shaping their roles. We will look at how managers can contribute more effectively to the quality and outcomes of patient care as well as to overall hospital performance. And we will develop guidelines for effectively implementing changes to healthcare working practices.

Details of research proposal

Relevance to SDO call for proposals

Our project focuses on the ‘realities of management’ theme (iii); ‘work life, roles and behaviours’, addressing priorities identified by Christian and Anderson (2007, p. 19) who concluded that, ‘Management issues were seen as a fundamental determinant of organizational performance: in particular the importance of different management practices; the competency of managers to fulfil their roles; the ability to link in with front-line staff; and involving key figures in proposed changes’. We will explore related themes, such as clinical–managerial relationships, decision-making, and knowledge utilization (Rousseau, Manning and Denyer, 2008; David Denyer, is a member of the American evidence-based management collaborative established by Denise Rousseau). But a better understanding of managing realities, changes, and contributions are where this research will have the most significant impact on organizational performance, and quality and outcomes of patient care.

A second intent of this call for proposals is to promote exchange between academic and practitioner communities. Our project engages participating managers throughout the research process, from developing this proposal, through advising on the collection and interpretation of data, to developing implications for practice, and disseminating findings.

Background; NHS context and relevant literature

Of the 1.3 million employees in the NHS in England, there are approximately 36,500 managers, less than 3 per cent of the total (The Information Centre, 2007). That probably underestimates the number of staff who as part of their role perform management functions. And that percentage understates the significance of management contributions to performance. The desire to engage medical staff in management and leadership dates from the 1980s, and has achieved new urgency in current proposals, such as the ‘medical leadership competency framework’ approved by the Academy of Medical Royal Colleges (NHS Institute, 2008; Hamiltonet al., 2008). In the context of the theoretical underpinning explained shortly, it is interesting to note that John Clark (Clarket al., 2008, p. 33), director of the Enhancing Engagement in Medical Leadership project observes that, ‘Enhanced clinical engagement should work towards a model of diffused leadership, where influence is exercised across a complex set of relationships, systems and cultures. It is a set of behaviours that should apply to all rather than a few’.

Although the Next Stage Review promises ‘no new targets’, the change agenda is sustained. Lord Darzi focuses management attention on accelerating the pace of change with regard to quality of care (linked to funding), patient choice, personalized budgets and care plans, and integrated care, complemented by clinical and board leadership programmes (Department of Health, 2008). The Operating Framework for 2008/09, noting a shift in emphasis away from finance and waiting times, declares an ‘ambitious new chapter’ in the transformation of the NHS, focusing on other issues including patient safety, access, better health and reduced inequalities, improving the patient experience and staff satisfaction, and enhanced emergency preparedness; not a recipe for stability (Department of Health, 2007). These aspirations will be achieved by empowering local management and staff to deliver with less central direction.

The Operating Framework also makes clear (p. 32) that the status of Foundation Trust is no longer an aspiration, but an expectation for all. The governance arrangements of Foundation Trusts, particularly with service line reporting, mean that trusts, and their clinical services, run like businesses. Plans and decisions are now commonly couched in commercial discourse; business units, customers, competitors, marketing (‘promotion of services’), cost allocations, profitability, portfolio analysis, mergers and acquisitions, business development (e.g., Shepherd, 2008). This reflects values different from those that have inspired a publicly funded healthcare system for the past 60 years. Our anecdotal evidence indicates that managers generally welcome these developments, but that many clinical staff remain sceptical. There is evidence to suggest that these changes are creating new tensions (Sambrook, 2005). It is in this dynamic context of the ongoing – accelerating – transformation of healthcare that this study is positioned.

Plan of investigation

Data collection and management procedures

This section explains the project data collection and handling arrangements, explaining how ethical issues arising from this study will be addressed.

Benefits to the NHS

For patients and service users, although not involved directly in project fieldwork, this research will deliver:

• management practices and organizational features that have been demonstrated to contribute to improved quality of care and clinical outcomes;

• rapid changes to working practices following ‘extreme’ events, thus leading to improved patient safety.

For middle and front line managers, this research will deliver:

• knowledge of how middle and front line management work is evolving, and why;

• new competency requirements, and how these are acquired and can be best supported;

• new practices, tools, diagnostics, and frameworks for influencing clinical outcomes, care quality, and organizational performance;

• approaches and techniques for managing both ‘extreme’ and ‘normal’ organizational change.

For senior managers, this research will deliver new information on management development priorities and support needs, and a practical guide to the construction and maintenance of an ‘enabling context’ for maximizing the impact of management practices on clinical, care-related, organizational, and change-related outcomes.

For policy makers, this research will deliver a model of healthcare management work, explaining the demands and pressures which these roles generate, the competencies required, the contributions of management practices to change and performance outcomes, and the implications for extending clinical engagement in management and leadership roles.

This project will thus deliver fresh evidence about the realities of middle and front line management work, new perspectives on the implementation of change in atypical circumstances, and a better understanding of the effects of management practices. While evidence, perspectives, and understanding are intangible outcomes, they are nevertheless valuable to the extent that they redirect attention and energy, shape our understanding of problems and the settings in which they arise, and help to guide practical action.

The involvement of stakeholders

Our research design has the advantage of involving significant numbers of individuals with experience of and commitment to the service. Stakeholders will have multiple opportunities to contribute insights and to challenge. This project has several national, regional, and local stakeholders including policy makers, managers, clinical staff, and patients. These groups are not remote entities to be considered when the study is over. On the contrary, one role of our virtual advisory group is to help us to capture the views of those groups from the start.

Cranfield's mission is to improve management practice through research that generates ‘near to market’ actionable knowledge. Our collaborative design, advisory groups, the involvement of management participants, focus and briefing groups, and dissemination mechanisms, are intended to ensure continuing stakeholder involvement, particularly in the co-production of implications for practice, and innovative ideas for dissemination.

Dissemination plans

Researcher: ‘In what form would you like to see our findings presented?’ Chief executive: ‘Not another report.’ Our staged and collaborative research design means that outputs will develop throughout the project, and data streams will be ultimately combined into a series of publications, including academic journals and a book. Our final report will be complemented by briefs summarizing practical guidance, and we will publish in practitioner journals. We will also use Cranfield open and customized programmes, and our Public Sector Performance Roundtable. The project will feature on our School website, and WebEx will be used as an interactive dissemination tool. We will also contribute to practitioner workshops and conferences. But those are all relatively conventional outcomes. We are sensitive to the need to develop ‘high impact’ communication and dissemination media and channels for this project. To help us to develop more innovative methods for disseminating findings, propelling the research-into-practice process, we will be driven by ideas from our project advisory and management briefing groups. We will be seeking their ideas in this respect throughout the project, and not just towards the end.

Project timetable*

This project will run over 42 months, from 1 January 2009 to 30 June 2012:

Stage 1a Pre-research administration activity: January to October 2009

Stage 1b Site briefing and set-up processes: April to December 2009

Stage 2 Management focus groups: January 2010 to December 2010

Stage 3 The 60–40 survey: January to September 2011

Stage 4 Management briefings: June to December 2011

Stage 5 Managing extreme events: January 2010 to December 2011

Stage 6 Publication and dissemination: ongoing

* These timings are approximate, affected by delays generated by ethical approvals process, and pressures on service managers

Interim reports

We will submit interim reports during the first month following the completion of each stage of the project – in July 2009, January 2010, October 2010, January 2011, July 2011, and January 2012. These reports will summarize progress, key findings, theoretical developments, practical implications, problems arising and how these will be addressed, and will highlight any unusual, unanticipated, and particularly significant issues and outcomes.

References

Main conclusions

1. Wattle Park had previous experience of norovirus (the winter vomiting bug), but this incident was on a much larger scale and the hospital appears to have been unprepared to handle it.

2. As this incident was different in scale and consequences from previous outbreaks, this was a significant learning opportunity with regard to future infection management procedures.

3. The involvement of two external agencies potentially exacerbated the impact of this outbreak, hastening decisions to close wards and to withdraw service provision.

4. Initially defined as a serious outbreak of norovirus, this incident quickly morphed to become a serious interruption of emergency and elective services.

5. Changes in the aftermath of this incident focused on hygiene procedures that were not picked up in the recommendations of the external review; these procedures were within the control of middle management, whereas costly changes to ageing facilities were not.

6. Changes in the aftermath of incidents like this are shaped by the context in which the incident occurred, and in this case by the financial and local political conditions facing this hospital, and by the differing views and priorities of the two external agencies involved.

Introduction

In January 2010 there was a major outbreak of norovirus (the winter vomiting bug) at Wattle Park. Unlike previous outbreaks, the number of cases rose rapidly, and within a few weeks all but four of the hospital's 33 wards were closed to new admissions. The hospital was also closed to emergency admissions for 96 hours, and inpatient elective surgery was cancelled for 5 weeks. Once the outbreak was under control, an external review into the causes was commissioned. This review identified a number of contributory factors concerning infection control procedures, ward cleaning and the layout and condition of the hospital estate. Recommendations thus involved changes to infection control procedures, most notably ward closure, and isolation of suspected cases; the cleaning arrangements following a ward closure; increase in size and skill mix of the infection control team; and more rapid confirmation of cases through on-site testing. One year later, several changes had been made. However, in a number of areas, recommendations had not been implemented. No changes had been made to the infection control team and, because of financial constraints, only limited changes had been made to the premises, thereby limiting the opportunity for isolation of suspected cases. There had also been considerable focus on hand hygiene following the outbreak, although this had not been mentioned in the report, and the hospital had a good record in relation to MRSA and C. difficile. Senior managers also questioned the definition of this incident: was this the control of the virus, or the inability to provide emergency and other services? This case report concludes that the focus of activity following the outbreak was influenced in part by the hospital's context, and in particular the financial position, and the political context in which the external review was received. After an event like this, there is a perception of the need for urgent change. But what takes place in the aftermath needs to be seen in the context in which the incident occurred, the investigation took place and the report was received, particularly in relation to directing energy for change.

Assessing the progress in implementing the recommendations from the review 1 year after the outbreak, some of the focus for change was on activities not mentioned in the recommendations, notably on hand hygiene and clinical practice. This is puzzling. Given the serious nature and consequences of the outbreak, attention might have been expected to focus on the recommendations, particularly as the report was completed rapidly following containment of the outbreak. However, the focus of activity and the widening of the discourse concerning necessary change were shaped by contextual factors, including the financial position of the trust, how the report was disseminated and perceptions of the findings.

The report was not widely circulated to hospital staff. Several senior nursing staff said that they had not seen it. This was in part because the initial draft was not well received by the executive team and modifications were requested. The outbreak and its consequences generated strong emotional responses from staff, particularly nurses. It may have been the case that managers, not fully aware of the recommendations, focused efforts on general infection control, even though this had not been specifically highlighted by the review.

The trust's severely constrained financial situation meant that it was difficult to respond fully in the short term to recommendations involving investment. Faced with this constraint, and the pressures to be seen to be doing something following an incident of this nature, the result may have been a focus on activity overtly linked to infection control, involving little or no cost. Furthermore, decisions concerning the allocation of resources in this area were beyond the control of managers such as lead nurses, who could not authorise the creation of isolation bays, but were able to persuade colleagues to wash their hands between each patient episode.

The outbreak

At the beginning of 2010, Wattle Park experienced a serious outbreak of norovirus. This spread quickly, affecting many inpatient wards over several weeks. At the peak of the outbreak, all but four of the hospital's wards were closed to new admissions for about 1 month. This situation compromised the hospital's ability to provide services to the local community. Scheduled elective in-patient procedures were cancelled over 5 weeks, resulting in an estimated loss of income of around £6M. Emergency admissions were also closed for 96 hours and the hospital was in ‘major incident mode’ for 5 weeks. For some of this time, to reduce the likelihood of the virus being brought in from the community, access to the hospital was controlled by security guards and inpatients were not allowed visitors for a period of 3 weeks.

There had been several previous outbreaks of norovirus at this hospital, but none had been as serious as this. In the past, the procedures in place successfully controlled the spread of infection, with minor impact on service delivery. However, the experience on this occasion was different, and the Director of Infection Prevention and Control (DIPC) admitted that, in spite of the external investigation, it was not entirely clear why the virus had spread so rapidly this time.

Information sources

The information for this case study was collected in early 2011, around 12 months after the start of the outbreak. This involved semistructured interviews with eight hospital staff and access to hospital board papers, including the investigation report. Local press reports were also consulted. Interviewees reflected different roles and levels of seniority, including two board directors, two operational managers, two clinical staff and two staff with infection control responsibilities. Interviews lasted around 1 hour and were conducted in the hospital, in working time. With prior permission, interviews were audio recorded and subsequently transcribed for analysis. Board papers and press reports were subject to content analysis.

Organisational context

Wattle Park is a teaching hospital offering a full range of services, with an emergency department and urgent care centre, with 700 beds, 3400 staff and a budget of £240M in 2010–11. The trust had faced financial difficulties for several years. In 2006–7, a survival strategy was designed to address a deficit of £25M, and over 250 posts were lost following a recruitment freeze. Training budgets were also reduced. Many managers described the rather crude approach that was used at that time as ‘slash and burn’, some indicating that they felt that the trust had ‘thrown the baby out with the bathwater’. Several managers said that the hospital was still suffering from the remaining scars.

Wattle Park was still facing cost pressures at the time of this study, compounded by deficits among local commissioners. Although the position became more secure in 2011, staff were ‘battle weary’. There was uncertainty over the hospital's future role with proposals to reorganise services in the area, and Wattle Park was considering merging with other nearby trusts. It was felt that the hospital would survive, and continue to provide emergency, maternity and related services, but it was unclear which other services the hospital would retain.

Much of the hospital was housed in Victorian buildings that had been altered over the years. Most of the medical wards were in the older areas and were traditional ‘Nightingale wards’. Less than 10% of the hospital's beds were in single rooms, and none of those rooms on the medical wards had en suite facilities. There was an isolation unit with nine single rooms and two bays with three and four beds each respectively. Only one single room in this unit had en suite facilities.

Incident background

The symptoms of norovirus include nausea, projectile vomiting and diarrhoea. The virus is easily spread from one person to another by the aerosols of projectile vomit and through faecal–oral contamination. The infectious dose of the virus is low and immunity tends to last for only a few weeks. People of all ages are susceptible. Not all of those who are infected, however, exhibit all of the symptoms and, without testing, accurate diagnosis is difficult. Furthermore, the symptoms of norovirus are common to some other conditions. Although this virus is not considered to be serious in healthy individuals, and recovery typically takes only 2 or 3 days, it may be more serious for those with existing conditions who may take much longer to recover. The virus is excreted for 48 hours after symptoms have stopped, but may linger for several weeks.

The outbreak began on Friday 8 January 2010 when a case of diarrhoea and vomiting (D&V) was reported to the infection control team in one of the medical wards. The team then found that six patients on this ward had symptoms. Given the conditions of these patients, the DIPC thought that some of these symptoms might derive from other non-infectious causes and decided to place this ward ‘under surveillance’. Restrictions were placed on moving patients out of this ward, but not on new admissions. The ward was reviewed over the weekend, when further cases of D&V were identified, but the ward remained ‘under surveillance’ and was not closed to new admissions.

On Sunday 10 January, two more medical wards reported patients with D&V symptoms. One of these wards was subsequently closed to new admissions while the other was placed under surveillance, as it was felt that the symptoms were unlikely to be caused by a norovirus infection. On Monday 11 January, the hospital formally declared a norovirus outbreak and informed the Health Protection Agency (HPA) and commissioners. The wards under surveillance were closed to new admissions and daily outbreak meetings were instituted to control the spread of the virus. Over the following 5 days, more wards reported patients with symptoms, and these were assessed by the infection control team. More wards were closed to new admissions until all patients in the ward had been symptom free for at least 48 hours.

On Tuesday 12 January, the hospital declared a Sustained Ongoing Internal Major Incident, and from Wednesday 13 January, for 5 weeks, all elective inpatient admissions were cancelled. From 25 to 29 January the hospital closed to emergency admissions for 96 hours to avoid admitting any more infected patients from the community. From 26 January, the ‘Bronze Control Team’ was based in the hospital boardroom, dealing with operational decision-making and communications and providing support to the infection control team. Additional specialist infection control nurses were seconded from neighbouring hospitals.

Following a meeting of hospital management with their commissioners and the HPA, a more risk-averse approach to closing wards to new admissions was adopted. It was agreed that reported cases would be assumed to be norovirus unless proven otherwise, and wards would be closed to new admissions when ward staff reported two or more patients with D&V symptoms. Previously, the policy had been to reopen wards 48 hours after the last symptoms had been observed. This was now changed to 72 hours, in line with HPA guidelines.

There was conflicting opinion about the progress of the outbreak. The DIPC said that he believed that the outbreak was coming to an end by 18 January, when the number of new cases reported was only three, although there was a high number of ongoing cases. This was in line with the hospital's previous experience of albeit more limited outbreaks.

Investigation

Following the incident, an external review was jointly commissioned by the hospital and its main commissioning body. The investigation report was delivered in a relatively short period of time and a first draft was available in February 2010.

The review was conducted by an infection control professional from a hospital in another area. This review largely involved conversations with senior managers and the infection control team. The resultant report highlighted a number of shortcomings. Overall, there had been a failure on the wards to recognise the symptoms of the virus and to report these to the infection control team at an early stage. This, it was felt, had resulted in the spread of infection both within and between wards. The review also pointed to the failure of the infection control team to close wards to new admissions quickly enough, as opposed to placing them ‘under surveillance’, a status that was, as became evident, poorly understood by other staff. These factors had increased the infection rate.

The report focused on the infection control and outbreak management aspects of the incident. However, there was controversy at board level, in part because views of the causes of the outbreak were subject to some considerable variation. It is also noteworthy that the report did not appear to have been widely circulated or the recommendations discussed with staff. Several interviewees said that they had not had sight of the final report and were thus not familiar with the recommendations. Furthermore, although there were some post-incident workshops run by the emergency planning team designed to learn lessons from how the outbreak was handled, one lead nurse indicated that, because of the intense pressures experienced by ward staff, some found it hard to go back and relive their experiences of those weeks:

Our emergency planning officer carried out some workshops where various people across the organisation came together to discuss the good things, the bad things, what we needed to do differently. The workshop I went to wasn't particularly well attended. I think by that time people had . . . it was such . . . and I can't explain how stressful a time it was, and people were just kind of like, ‘Oh no, no, I can't revisit that. I don't think people could really. I don't think because they didn't want to. I just don't think people could relive it.

The review also identified the following issues:

• Failure to close wards at the start of the outbreak contributed to the spread of infection.

• Staff were not clear what ‘under surveillance’ meant with regard to the movement of patients.

• Contractual arrangements for ward cleaning were not adequate to deal with the outbreak, not allowing for enhanced levels of cleaning or for terminal cleaning before reopening a ward.

• There had been debate over whether or not to cohort infected patients in the same ward (and many wards had empty beds, being closed to new patients). Cohorting also allows empty wards to be cleaned and reopened for emergency admissions. However, the cohorting process involves moving infected patients around the hospital, thus increasing the risk of further infection. It was decided during the outbreak, in agreement with commissioners, to cohort patients, allowing 10 wards to be cleaned and reopened. If one of the primary purposes of an acute hospital is to provide emergency health care to the local population, then anything which compromises that is of major significance. The decision to close emergency admissions was made when there were many empty beds in the hospital, because affected patients were not cohorted. The review noted that, had cohorting taken place earlier, the emergency pathway would not have been compromised.

• Infection control was not helped by the old Victorian buildings and the hospital layout. The Nightingale wards also made the hospital vulnerable to the spread of infection.

• The layout meant that it was not possible to admit new patients to single rooms, allowing them to be assessed for infection. In some cases, patients displaying symptoms were moved to another ward to give them a single room, thus increasing the risk of spreading the infection during transfer, and spreading infection to that other ward.

• Assessment for D&V symptoms in the emergency department was not routine, and patients infected in the community were bringing the infection with them when admitted to wards.

• Keeping potentially affected patients in the admissions area also meant that there was an ongoing source of infection which could be transferred to other wards.

• At the time of the outbreak, rapid testing for the virus was not available on site. The delay in getting test results meant that patient management decisions were also delayed.

The recommendations

The external review made the following eight recommendations:

1. Wards should be closed at an early stage, rather than designated ‘under surveillance’, which only prevents patients being moved out of a ward.

2. An area should be designated for suspected and confirmed cases that are admitted, to cohort them and to protect other patients. Bays should be provided for confirmed cases and single rooms for suspected cases.

3. The infection control team should assess risks and clinically review cases, to inform isolation and ward closure decisions.

4. The size and skill mix of the infection control team should be reviewed given the size of the hospital and the age and nature of its facilities.

5. The provision of on-site testing facilities for the virus should be considered, to inform decisions about outbreak management at an earlier stage.

6. Cleaning contracts should be revised to include provisions for enhanced cleaning levels (e.g. frequent cleaning of touch surfaces) and terminal cleaning in the case of outbreaks.

7. A process of inspection and sign-off should be introduced for cleaning clinical areas before they are reopened.

8. HPA guidance on hospital viral outbreaks of this nature needs to be updated.

Change following the incident

There was a difference of opinion among those interviewed concerning what constituted ‘the incident’. One definition concerned the outbreak and spread of norovirus resulting in ward closures. The incident generated significant management challenges, with an incident command team being established in the boardroom and the infection control team playing a central role. The outbreak also resulted in negative reporting in the local press, probably causing long-term damage to the trust's reputation. However, some staff, particularly senior managers, began to see this incident in the wider context of the trust's ability to fulfil its role as a district hospital and to provide services to the local population. As such, managing the outbreak, at least as it progressed, was positioned against the ability to carry on delivering elective and emergency services. This was felt particularly strongly following the 96-hour closure of the emergency pathway and prompted the decision to cohort infected patients to allow for wards to be reopened. A senior manager observed:

The strategic objective was to reduce the chance of cross infection [but] became a situation where we had to switch the strategic objective to protect an emergency pathway. At a certain point, ninety-six hours into having the A&E shut to emergency admissions, the nature of the incident changed.

Overall, there was a need to balance the consequences of controlling the virus with maintaining emergency services. The same senior manager commented:

So, our strategic objective then changed to maintaining the integrity of the emergency pathway, and the board had an open conversation about this, knowing that we might as a consequence of this, expose some patients to norovirus who otherwise wouldn't have been. But the relative risk of norovirus, which is unpleasant, it's a forty-eight- to seventy-two-hour self-limiting condition, whereas driving patients round in the back of ambulances, looking for hospitals who can take them, is a much higher risk, and will affect many more people than potential exposure to norovirus.

Elective surgery was cancelled for 5 weeks. However, the considerable loss of income which this entailed (around £6M) was felt to be less crucial to the trust, because cancelling surgery had other significant non-financial consequences. For example, recovering from these cancellations took much longer than recovery from the closure of emergency services in relation to the rescheduling of operations. A shared notion of what constitutes an incident is likely to have implications for how an organisation responds, and the types of changes that are planned to avoid further similar incidents.

A year after the outbreak, a number of changes were reported to have taken place to avoid a further similar situation. Nevertheless, it was acknowledged that norovirus would remain in the community and that measures were required to prevent the infection spreading and limit the damage that an outbreak would inflict on the hospital's ability to continue to provide services. Interviewees also reported issues that had not changed, or changes that had not been sustained.

The nature of the hospital estate was seen to be a major factor in the spread of the outbreak. This was not only because of the predominance of Nightingale wards and the lack of single and side rooms, which would have enabled infected patients to be isolated, but also because of the fabric of the building and the age and quality of the furnishings. One senior manager remarked, ‘If you wash a porous wall, it is still porous’. In addition, in some parts of the hospital, the doors were not wide enough to allow the type of bed used for elderly patients to pass through, and isolation and cohorting thus became more problematic. There was widespread agreement that there was limited scope to deal with these problems. One senior manager commented that, ‘In truth, the place should be razed to the ground and something built that's fit for the 21st, not the 19th century’.

The lack of single rooms meant that it was difficult to isolate patients who were displaying symptoms of the virus before they were sent to a ward. Single rooms would allow time for patients to be isolated, to check if they were infected or whether or not their symptoms had other causes. Many felt that this was a main reason for the rapid spread of the virus in the hospital. Nightingale wards were seen as problematic because, if there was an infected patient on the ward, this exposed around 20 other patients to the virus. However, there were some differing views about the extent to which Nightingale wards really were the problem. One lead nurse indicated that good hygiene practices could minimise cross-infection irrespective of layout. It was noted that, if patients were in four- or six-bed bays, with one nurse in each bay, then cross-infection was likely to be restricted to the other three or five patients in that bay. Nightingale wards also meant that, when restrictions were placed on patients moving out of wards where there had been an infection, a larger number of beds were affected rather than the small number in a bay. If admissions were restricted because of infection, this could result in large numbers of empty beds, unavailable for emergency admissions or elective surgery patients.

At the time that the research was conducted, progress had been made in providing side rooms so that patients admitted through the emergency department displaying symptoms of infection could be isolated and their condition monitored. A ward refurbishment programme was also in place, which included reconfiguring the Nightingale wards into bays. However, these plans resulted in an overall loss of beds, as bays require more space. It was estimated that, following refurbishment, 85 beds would have been lost across the hospital. Progress on the refurbishment was slow because, in an already financially challenged trust, resources to fund this were limited. One board member indicated that they had put some money aside each year to fund this. In addition, there had been unforeseen problems, such as the discovery of asbestos, which had to be removed first and which delayed building work. At the same time, however, an increase in the number of side rooms and single rooms would allow the trust to make progress on achieving its targets for single-sex wards.

The report highlighted ward cleaning as an area for attention with regard to the additional cleaning required during a virus outbreak and procedures for ensuring that cleaning had been carried out to an acceptable standard. Those interviewed said that good progress had been made and that they had established a new set of arrangements in collaboration with the cleaning company.

The reporting of suspected cases to the infection control team and the timing of the decision to close wards were also identified as contributing to the speed of the outbreak. Several respondents said that they felt that wards were not closed early enough and that this had contributed to the spread of the virus. Following the outbreak, staff were encouraged to send specimens for testing early on, so that suspected cases could be confirmed and the appropriate steps taken as soon as possible. This was reported to be generally improved, with staff being more vigilant. However, it was acknowledged that, for example, a patient who did not need assistance to go to the toilet could have symptoms and not report them initially, but that by the time they did report symptoms several other patients might have gone to the toilet and been exposed to the virus on touch surfaces such as door handles or taps. Although rapid reporting of suspected cases was seen to have improved, the infection control team indicated that this was an area where staff needed constant reminders to remain vigilant, and that over-reporting was preferable to under-reporting.

The report also noted that the infection control team was small for a trust of this size, and at the time of the outbreak one staff member was on long-term sick leave. At the time of this study there had been no changes to the size or composition of the team, but there was no indication that this lack of resource had impacted on the spread of the outbreak. Additional resources were brought in from neighbouring trusts when the outbreak escalated. The lack of in-house resources may have affected the executive's decision to call in the HPA at an early stage. Some comments were made by staff about the clarity of the advice given by the infection control team and how in some instances this conflicted with the advice from the HPA.

Interestingly, following the outbreak, hand hygiene attracted a lot of attention. Hand hygiene was not identified in the report as a contributory factor in the spread of the virus and the trust had a good record of controlling MRSA and C. difficile infections, indicating that hand hygiene had not historically been a problem. However, at the time of the research, hand hygiene was mentioned as a contributory factor by many interviewees, who also noted that ward layouts meant that hand washing was not made easy for staff. In many Nightingale wards, for example, there might be two basins in the middle of the ward, requiring staff to walk up and down the ward frequently to wash hands between patients. Alcohol gel was provided by each bed, but this is not effective against norovirus. An operational manager argued that there was a need to be more proactive, that portable sinks should be placed at the entrance to wards, that ward staff should take responsibility to ensure that they were replenished regularly and that hand-drying materials were available and disposed of safely. It was noted that the provision of washbasins was to be addressed in the refurbishment of wards.

It was also observed that there was some misunderstanding in this area among staff. For example, some thought that, if they wore gloves, frequent hand washing was less necessary. The infection control team had run many training and awareness programmes about hand hygiene and dress codes since the outbreak, but there was a general view that this remained a problem. Interviewees indicated that good progress had been made initially following the outbreak, but that this was difficult to sustain. It was acknowledged that staff were busy and there was a need for constant reminders to be vigilant. Trust management had made it clear that they would use disciplinary procedures against members of staff not adhering to hand hygiene practices. The director of operations said:

Fair enough, you know, there may be a situation where, for whatever reason, you've had a mental aberration and you're reminded to wash your hands and you do it. The second time, you get a letter saying, ‘You still haven't got this as a normal part of your daily business.’ And on the third occasion you are then disciplined.

There were, however, no reports of any disciplinary action being taken against members of staff, and some interviewees felt that senior management were not serious about this. One executive team member defended the lack of disciplinary action, noting that it was recognised that staff worked under very difficult circumstances and that failure to wash hands was not see as deliberate but as human error. Other senior managers indicated that a lesson learned here was that ‘you cannot drop your guard’ on hygiene and that there was a need for constant reinforcement.

Although improved hand hygiene is always likely to be beneficial in a hospital, it is interesting to observe that, 1 year on from the outbreak, this remained a major focus of attention and was deemed by many interviewees to have contributed to the spread of the virus. This is particularly interesting as it was not mentioned in the investigation report and the trust had a good record in controlling other infections. This outbreak had a traumatic impact on staff, many of whom had never experienced anything as widespread as this. Only senior managers were able to control the refurbishment of buildings and to increase the size of the infection control team, so hand hygiene may have become the focus of activity as this was in the control of all members of staff. Hand hygiene is also an issue over which responsibility is widespread – doctors, nurses, cleaners – and that crosses specialties and levels in the organisation hierarchy. Moreover, a number of those interviewed, from different areas, felt that there was a ‘blame culture’ in the hospital, and that blame for the norovirus outbreak was attributed to ward staff, which had been demoralising.

A number of general lessons were also learned:

• it is necessary to limit the number of people who move between wards during an outbreak, including non-clinical staff, such as the newspaper seller

• to prevent cross-contamination, domestic staff should work only on either infected or non-infected wards

• agency nursing staff (the use of which was common) should not work shifts in both infected and non-infected wards.

One operational manager indicated that, when it happened, the hospital did not publicise the outbreak widely until measures such as preventing visitors from coming to the hospital and closing the emergency pathway had already been taken. She indicated that the trust had become more confident and assertive in publishing information about infections and communicating to visitors the need to be vigilant, especially with regard to hand washing. This research was carried out approximately 1 year after the outbreak, at which time the hospital had a small number of cases of norovirus, having closed five wards. It was felt that this was a good test of its ability to manage the outbreak effectively, and to protect emergency and elective pathways.

Managing the outbreak

The infection control team believed that the outbreak could have been managed differently, with less impact on service delivery. Team members felt that, when the decision was made to bring in the HPA, measures were put into place which resulted in a greater degree of ward closure than was necessary. They also felt that, once the outbreak had started to develop, there was a degree of over-reporting (although it was desirable that staff notified the infection control team of suspected cases) and that, for some patients, their symptoms had other causes. Their view was that the outbreak had been contained in about 2 weeks but that, as a result of the involvement of external bodies (commissioners and the HPA), it was prolonged. Team members suggested that the HPA had been cautious and that the management of the outbreak was taken over by people more experienced in managing infection control in the community rather than in a hospital. It was argued that, had this approach not been taken, hospital services may not have been compromised.

Several interviewees observed that, when the outbreak team was set up in the boardroom, too many people became involved and noise levels were high, with lots of mini-meetings taking place. Because the boardroom was located on the main hospital corridor, it was easy for people to ‘drop by’ and ‘see how things were going’. This was felt to have been unhelpful, and with this volume of traffic through the boardroom, some from infected wards, this could have been a source of cross-contamination.

It was also felt that one of the problems that the outbreak team faced was dealing with conflicting advice from different sources. In particular, there were differences of opinion between the infection control team and the HPA. The HPA was felt to be cautious in its approach and this inhibited the ability to deliver services. However, it was acknowledged that the HPA had to be involved, for the hospital's own protection, and as a result the hospital was faced with comparing different forms of risk. By contrast, the commissioners were concerned with the political implications associated with the reduction in service provision and as a result were keen to see actions that would allow services to be resumed.

Research briefings

We also published a series of short (four-page) research briefings. These each addressed a single topic or theme and they were written for our end-user management audience. They were available in hard copy (colour printed A4 card) and also as PDF web downloads.

Buchanan DA. Releasing time to manage: what stops middle and front line managers from doing their jobs more effectively, and what can you do about it? Cranfield Healthcare Management Group Research Briefing 1, September 2010.

Buchanan DA. Names, ranks, and numbers: how many managers does the NHS have, and is that enough? Cranfield Healthcare Management Group Research Briefing 4, October 2010.

Buchanan DA. Spot the manager: how to identify middle and front line managers in the NHS. Cranfield Healthcare Management Group Research Briefing 5, November 2010.

Buchanan DA. Are healthcare management jobs becoming extreme jobs: and what are the implications of this trend? Cranfield Healthcare Management Group Research Briefing 7, February 2011.

Buchanan DA. Elsenham crossing: a wholly avoidable accident. Cranfield Healthcare Management Group Research Briefing 9, May 2011.

Buchanan DA. Reinforcing the management contribution. Cranfield Healthcare Management Group Research Briefing 10, May 2011.

Buchanan DA, Moore C. Management capacity: framework and assessment guidelines. Cranfield Healthcare Management Group Research Briefing 14, October 2011.

Buchanan DA. Grandmother’s footsteps: the institutional context of management work. Cranfield Healthcare Management Group Research Briefing 18, February 2012.

Gascoigne C. A review of research on extreme jobs: long hours, intense effort, high challenge. Cranfield Healthcare Management Group Research Briefing 19, May 2012.

Moore C. Knowing me, knowing you: interpersonal relationships between managers and hospital consultants. Cranfield Healthcare Management Group Research Briefing 11, May 2011.

Moore C. You are what you email: how do you manage your relationship with your inbox. Cranfield Healthcare Management Group Research Briefing 12, May 2011.

Moore C. Sticking plasters over big problems: are you managing your wicked problems or using sticking plasters? Cranfield Healthcare Management Group Research Briefing 13, September 2011.

Moore C. Mental toughness: applying sports and navy principles to healthcare. Cranfield Healthcare Management Group Research Briefing 16, December 2011.

Moore C. Sweat the small stuff: minor problems, rapid fixes, major gains. Cranfield Healthcare Management Group Research Briefing 17, January 2012.

Osbourne J. What motivates NHS managers?: is it just money? Cranfield Healthcare Management Group Research Briefing 3, December 2010.

Osbourne J. Challenges facing healthcare managers: what past research reveals. Cranfield Healthcare Management Group Research Briefing 6, January 2011.

Osbourne J. Managers have feelings too: emotions of healthcare managers. Cranfield Healthcare Management Group Research Briefing 8, April 2011.

Pilbeam CJ, Buchanan DA. After the crisis: the maintenance model of effective change. Cranfield Healthcare Management Group Research Briefing 2, September 2010.

Turnbull James K. What leadership practices are needed in complex healthcare organizations? Cranfield Healthcare Management Group Research Briefing 15, November 2011.

Submitted and under review

Buchanan DA, Denyer D. Researching tomorrow's crisis: methodological innovations and wider implications [published online ahead of print 13 November 2012]. Int J Manag Rev 2012. doi:10.1111/ijmr.12002. (Submitted on invitation from guest editors for special issue on management research methods and practice.)

Abstract

Solving small, annoying problems may appear to be unimportant in a political and economic climate that focuses attention on radical long-term changes to the organization, delivery, funding, and regulation of the healthcare system. However, as the initiative reported here suggests, small changes can generate major benefits for patients, staff, and hospital performance. This approach can also strengthen clinical–managerial relationships, which are key to larger scale initiatives, and which can suffer when ‘the small stuff’ is not fixed. The success of this initiative is explained with reference to a process model of change combining context, content, process, and individual dispositions. This model could also identify other settings where this approach is likely to be effective. On the basis of this experience, healthcare managers may be advised to be alert to what appear to be minor unresolved issues, and to address these in addition to undertaking larger scale, longer term projects.

Acknowledgements: The research on which this paper is based was funded by the National Institute for Health Research Service Delivery and Organisation programme (award number SDO/08/1808/238, ‘How do they manage?: a study of the realities of middle and front line management work in healthcare’).

Disclaimer: This paper is based on independent research commissioned by the National Institute for Health Research. The views expressed in this presentation are those of the authors, and not necessarily those of the NHS, the National Institute for Health Research or the Department of Health.

For want of a nail . . .

For many middle managers, they can see the problem and know how to fix it, but it's easy to let it drop. Will I pick up the coding battle today? Maybe not.

Operations manager, Clearview hospital

Very slow, very frustrating, and quite often not enough progress in an acceptable time, and the base problem remains. And after some time it becomes embedded and accepted because the middle manager concludes it's too hard to address.

Interim director, North Somerton hospital

The aim of this paper is to explore the potential benefits to patients, staff, and overall organizational performance, of addressing minor problems in an acute hospital setting – unanswered e-mails, delayed equipment orders. These issues can be overlooked in a context where major change is at a premium. However, a small-scale initiative designed to address such issues reveals the potential to generate savings, increase staff morale, improve quality of patient care, expose underlying problems, and strengthen clinical–managerial relationships. The success of this initiative set the foundation for further improvements, in this and other services. The change continuum from ‘fine tuning’ to ‘transformational’ implies the relative superiority of the latter (Stace and Dunphy, 2001; Kotter, 2008), which may be misleading.

In the current economic climate, attention focuses on radical changes to the ways in which healthcare is funded, organized, delivered and regulated (Department of Health, 2010). For the National Health Service (NHS) in England, meeting ‘the Nicholson challenge’ to cut the annual budget by £20 billion by 2014–15 involves implementing major cost improvement programmes (CIPs), which have meant cutting tens of millions of pounds from hospital budgets. National policy means that these reductions in spend must be accompanied by plans to maintain Quality, encourage Innovation, increase Productivity, and strengthen Prevention (of incidents which jeopardize patient safety); the so-called QIPP agenda. This implies ‘disruptive innovation’ (Christensenet al., 2000), not ‘tinkering’ (Abrahamson, 2004).

These policy-driven changes accompany the challenges facing all modern healthcare systems: rising patient expectations; ageing society; wider information availability; changing nature of disease; advances in treatments; changing workforce (Department of Health, 2009). The traditional model of the acute hospital is thus now in question, with the advent of polyclinics, care in the community, improved management of long-term conditions, and the integration of services in further attempts to reduce costs and streamline care (Edwards, 2010a).

The Clearview context

When the computer on the ward was bust, I said to the ward manager, we need a new one, and she said we need to ask the operations manager. When I asked the operations manager, she said the ward manager could do it. So it takes ages to just sort out the computer. It's this inefficiency which is irksome.

Consultant, medicine division, Clearview hospital

Clearview is an acute hospital with an annual budget of around £600 million, 7,000 staff, and 1,000 beds. Clearview was implementing cost improvement projects (CIPs) to reduce a projected £10 million deficit in 2011/12, and it was predicted that further savings of over £18 million would be required to reach break even, given the rising costs of pay and medical supplies. Clearview had seven divisions: cancer; women and children's and; emergency and perioperative care; diagnostics; medicine; neurosciences; surgery. The medicine division had CIPs totalling £9 million at the time of this initiative.

The initiative reported in this paper involved the gastroenterology service, which was part of the medicine division, with an annual budget of around £4.5 million, 70 staff, 6 consultants, a 22 ‘single-rooms’ ward, a five-room endoscopy suite, and daily outpatient clinics. This was a high volume service with a team that was receptive to change, and this initiative had the support of one of the consultants who encouraged his colleagues to participate.

Problems and fixes

Five ‘small problems’ were raised, involving patient pathways, two separate issues related to coding, a scanner, and patient safety. What were these problems, why had they not previously been resolved, how were they fixed, and what were the outcomes?

Much doing to be done

If NHS managers are struggling it is because of the size of the agenda and the lack of respect for the management task involved; and the systematic lack of investment in learning – it is hard to learn together when there is so much ‘doing’ to be done.

Malby, 2011, p. 19

Table 6 summarizes the five problems, the nature of each issue, why it had not been resolved, how it was resolved, the cost of the fix, and the benefits – which fall into five categories:

• financial: income generation

• processual: safer patient allocation

• temporal: tasks performed more quickly; less waiting time

• emotional: reduced annoyance, boredom, frustration

• relational: improved inter-professional relationships

The benefits to patients concern safety, reduced hospital visits and waiting times, and clinical staff who have more time to spend with them. The benefits to staff relate to reductions in routine manual work, and considerably less frustration. Corporate benefits include income generation, improvements in quality of care and safety, and improved relationships, thus contributing to the implementation of more radical initiatives in future.

The small issues underpinning this initiative had not previously been addressed due to:

1. workload

2. budget and logistical problems

3. incorrect advice

4. lack of understanding.

It is interesting that, in only one case, the failure to have fixed the issue was due to lack of finance (for an additional staff member), and that was resolved with the creative and cost-neutral development of visual process maps. Only one fix involved minor expenditure.

Informal feedback from the gastroenterology team was positive. The opportunity to be involved in solving those issues was welcomed, although two gastroenterology managers (deputy operations and directorate support), the animateur, and the ‘who knows who knows what’, did most of the work. However, the work involved was hardly burdensome: the animateur kept a log and calculated that, in helping to resolve those issues, she had spent 40 minutes in conversations over five days. Although it is managers who often leave those small issues unresolved, they typically have a better understanding of how to fix them. Clinical and medical staff often lack, and may have little time or desire to develop, the networks and the organisational knowledge that contributed to this initiative. Administrative staff were also a valuable source of potentially untapped knowledge, particularly with regard to negotiating the bureaucracy and finding short cuts.

Drawing on findings from the wider project, middle managers in acute hospitals are confronted with long and complex change agendas, with many conflicting priorities, all of which are always urgent. This has been described as ‘multiloading’ (Bruch and Menges, 2010; Malby, 2011), and was a feature of management work at Clearview. In this context, it is perhaps not surprising that ‘big change’ attracted attention, and ‘small stuff’ was overlooked. And as is often the case in healthcare, the solutions to these small problems saved expensive staff time, but were not ‘cash releasing’, but one would generate income.

From a theoretical standpoint, how can the success of this initiative be explained? Drawing on an influential processual perspective (Pettigrewet al., 2001; Dawson, 2003), Walkeret al. (2007) argue that change outcomes are shaped by the ‘integrative influence’ of context, content, process, and individual dispositions. Context refers to forces in the external and internal environments. Content concerns the changes being implemented, whether they are fundamental or incremental, changing the organization's character or fine tuning, episodic or continuous. Process relates to the actions of change agents. Significant individual factors include dispositions, behaviours, and reactions to change. Table 7 summarizes the nature of these four sets of factors in this case. First, there were many external and internal pressures for change. Second, the change agenda was driven by the staff of the clinical service, and was not externally imposed. In addition, the changes were not radical, but involved fine-tuning that was perceived to be important. Third, the three-person team – animateur, consultant champion, ‘who knows who knows what’ – appears to have been particularly effective, in launching the initiative, and in bringing it rapidly to a successful conclusion. Finally, the frustrations, and the ambitions of those involved were key to creating and sustaining interest in and commitment to the project.

From a managerial standpoint, it would appear to be a simple matter to extend this small-scale initiative to another clinical service. However, it also seems reasonable to assume that the probability of success elsewhere may depend on the extent to which the configuration of context, content, process, and individual dispositions broadly mirror those encountered in the, perhaps ‘ideal’, setting of this particular gastroenterology service. Where conditions may not appear to be ideal, this model can also act as a guide to remedial action.

Finally, considering the processual, temporal, emotional and relational benefits, what is the significance of an initiative focusing on minor issues in a climate of ‘unprecedented challenge’? Was this of value, or was this a diversion from the main transformational agenda? This initiative can be considered to have value in at least three respects.

First, several issues were resolved quickly, to the satisfaction of staff involved, at little cost other than small amounts of time and a £90 scanner. For those faced with the daily frustrations of those minor annoyances, those quick and permanent fixes were welcome. One of the consultants involved subsequently used the ‘quick fix’ developed in one of these instances to resolve another, similar problem just as rapidly. And it was planned to extend this initiative to other divisions and services.

Second, ‘tinkering’ has a recognized place in the toolkit of change management. Abrahamson (2000; 2004) describes tinkering as ‘fiddling with the nuts and bolts’, of what already exists in order to generate inspired solutions to current problems, rather than trying to create something new from scratch. He argues that tinkering is typically inexpensive and rapid, is not as destabilizing as large-scale radical change can often be, and is more likely to succeed (or to have lower opportunity costs of failure).

Third, previous research has suggested that one predictor of the success of the next change initiative concerns an organisation's past experience (Walkeret al., 2007). A track record of successful change encourages a predisposition to support further initiatives. Commentary in this field thus recommends achieving ‘quick wins’ (for a healthcare example, see Klaberet al., 2011). Sweat the small stuff achieved several welcome and visible quick wins, with possible medium and long term benefits beyond the five day programme. As previously indicated, another healthcare-specific predictor of the success of change lies with the quality of clinical/managerial relationships. This initiative served to strengthen relationships which were positive at the start, albeit weakened by those ongoing unresolved problems.

In other words, the cluster of ‘soft’ benefits that emerged from this small-scale initiative may contribute in significant ways to the support for, and the implementation and success of, the more radical, transformational, disruptive change agenda facing healthcare. This initiative also helped to cement the mutual interprofessional trust, respect, and collaboration on which those larger-scale challenges are likely to depend.

References

Aims and backdrop

The main aim of this appendix is to provide further details concerning the data collection and analysis methods used in this project. A further aim is to explore the wider lessons from this experience for future studies of this nature. The research methods are outlined briefly in Chapter 1. In this appendix we go ‘behind the scenes’ of the fieldwork and analysis process in more detail.

The backdrop to this project is explained in Chapter 1 with regard to the economic crisis from 2008 and the change of government in 2010. Both of those events have had a continuing impact on the funding and governance of the service as a whole, and on the management agendas of acute trusts. Chapter 3 outlines the dimensions of the institutional context in which NHS providers operate. The emphasis in those accounts lies with the challenges facing middle management in acute hospitals. But those trends and developments also affected significantly the conduct of this research, for which funding was secured in late 2008, with the aim of commencing fieldwork in 2009. This appendix thus also considers how those contextual conditions influenced the collection and analysis of data.

It was the best of times, it was the worst of times, to be conducting a study of middle management roles in acute trusts. On the one hand, managers were faced with new agendas and priorities, creating ideal conditions for exploring responses to these challenges and how managerial roles and structures were being reshaped, and identifying the new capabilities and processes that might be required. On the other hand, those new agendas and priorities generated considerable pressures, increasing demands on managers' time and reducing their availability for research meetings. The original project timetable was thus abandoned from an early stage and a 6-month extension was subsequently required, taking this project up to the end of June 2012, rather than to December 2011 as planned.

This appendix is structured in terms of the main project stages (see Table 1), considering the details of data collection and analysis, and also exploring how the wider context of this study shaped those processes, and the lessons for future research that can be drawn from this experience. We will thus return first to the recruitment of the six participating sites, followed by the set-up meetings, the focus groups, the management survey and finally the case incidents.

To secure a measure of respondent validation, we also invited participants at the six trusts to a project launch event in 2009, to project update workshops in 2010 and 2011 and to an end-of-project conference in June 2012, as well as to a series of management briefing sessions from mid-2011 onwards. As those various events were part of the verification process with regard to the analysis and interpretation of interview and focus group information in particular, their contributions in this regard will be discussed under those headings. It should also be noted, however, that those events were also offered to project participants at no cost as a courtesy, to say ‘thank you’ for contributing to this study, and to maintain the relationship for future research, should that be appropriate.

Site recruitment

A call for proposals in the area of ‘management practice’ was published in April 2008. Recognising that we could contribute to this theme, we began to pull together a project team and to approach potential participating trusts, while developing the outline proposal. As Chapter 1 indicates, the aim was to establish a collaborative research design, and this was adopted from this stage. With a combination of corporate and personal links with the six participating trusts, each was approached in 2008 through either their chief executive or an executive board member, who were each then asked to comment – critically – on the outline proposal. This step confirmed the interest in principle of those trusts in this project, and provided reassurance that the proposal was addressing themes and issues that they considered to be significant. One issue from an early draft, concerning teamwork, was rejected as having attracted enough research and practical attention already.

The outline proposal was accepted by SDO, as was the subsequent full research proposal, and January 2009 was the formal project start date. Each trust was thus revisited early in 2009 to identify key contacts through whom the research team would work, and to establish the project logistics in terms of distributing participant information and arranging interviews and focus groups. To gather background information about each trust, and to establish the management groups that would be involved in focus group discussions and the management survey, key contacts were asked to assist in arranging a small number of set-up interviews, as described in the following section.

The original project plan allowed 6 months for gaining ethical approval and for completing governance checks through the Integrated Research Application System (IRAS). A favourable ethical approval was granted in April 2009, by Cambridgeshire 3 Research Ethics Committee. The governance checks, however, were not completed until the end of October that year; the processes applied by the Coordinated System for gaining NHS Permission – CSP – were cumbersome and time-consuming. National Institute for Health Research guidelines at that time stated that project team members did not require research passports, given the levels of access required to hospital staff and information. One of the Comprehensive Local Research Networks implicated in this project insisted that passports were required, again contributing to delay.

A process that was expected to last 6 months thus took 10 months, consuming considerably more time and effort than anticipated, particularly on the part of the principal investigator. The research governance framework has since been revised and studies such as this, involving NHS staff in their professional capacities, no longer require the approvals that were necessary for this study.

This delay was to have three consequences for data collection and analysis. First, it was not possible to start arranging focus group meetings with middle managers (many of whom have ‘hybrid’ clinical responsibilities) during the final quarter of calendar 2009, at the point when annual ‘winter pressures’ start to mount. This meant that focus group arrangements could not be activated until well into 2010.

Second, this had a predictable ‘knock-on’ effect on the rest of the project with regard to the timing of data collection in subsequent phases. Although steps were taken to limit the damage, a 6-month extension was eventually required at the end of the project. This early delay also affected data analysis. It is more straightforward to analyse information, and especially qualitative information, that has been gathered, complete, within a defined period than it is to handle information that accumulates more slowly, in an unpredictable manner, in which the end point is unclear.

Third, a degree of momentum had been lost. Having stimulated the interest of senior hospital staff through 2008 and early 2009, having arranged interviews to gather background information and having agreed the focus group logistics, core data collection would not begin until well into 2010. Some key staff had left or changed roles in that time, one trust had appointed a new chief executive and the site recruitment process thus had to be repeated, requiring further investments of time.

Set-up interviews

As explained in Chapter 1, the set-up interviews (see agenda in Appendix 2) had three objectives: first, to identify the management groups who would be involved in the project; second, to gather background information about the trust and its management agenda; and, third, to establish the logistical arrangements at each site for distributing project information. In preparation for the subsequent focus groups and management survey, interviewees were also asked about the pressures and demands facing middle managers, about their motivations and rewards and for examples of management contributions to clinical outcomes and care quality. A final question was, ‘what has to change to allow middle managers to make an even stronger contribution?’. These interviews were relatively informal, lasting up to 1 hour, with the researcher taking notes that were subsequently transcribed (no recording took place at this stage).

These interviews began with each site's lead contact, who was then asked to nominate colleagues who would be able to provide further background information. This process had several benefits. First, it stimulated interest in the project among those approached. Second, the management contributions of ‘hybrids’ were reinforced. Third, it confirmed the scope and challenge of acute trust management agendas, which, apart from specific local issues, were similar across all six sites. Consequently, this programme was extended, eventually covering 93 interviewees (see Table 3).

The transcribed interview notes were subjected to content analysis, question by question. Content analysis is a data reduction method, commonly applied to qualitative information, and is based on identifying and categorising (coding) recurring themes. For example, answers to the ‘motives and rewards’ question from 13 interviewees at Greenhill were coded in nine categories (Table 43).

Similar analyses were conducted for ‘pressures’, ‘contributions’ and ‘what has to change’. These analyses were first carried out by the lead researcher for each site, and then checked by the principal investigator. The project research fellow collated and further checked these analyses. There is clearly a degree of subjectivity involved in categorising and labelling interview comments such as these, and content analysis can be controversial for that reason. However, in this case, and as the above illustrations suggest, only minor disputes arose over category labels, and the tables in the main report allow readers to check whether or not the categories make sense to them. The number of items coded in each category was also recorded, simply to provide a rough indication of the weight of evidence being gathered; for example, the 13 interviewees at Greenhill identified a total of 42 ‘motivation’ items.

How confident are we that this analysis has captured accurately the information that interviewees provided? In addition to discussion within the project team, these analyses were presented to a group of management participants from the six participating trusts at the first project update workshop in June 2010. Participants were asked whether or not they ‘recognised’ these analyses as an accurate portrayal of the issues; no challenges were launched. This step is sometimes called ‘respondent validation’, a term suggesting degrees of rigour and finality that are rarely present. However, these findings were also presented at the second update workshop in June 2011, in 16 management-briefing sessions and at the final project conference in June 2012, where they invariably met with agreement.

Focus groups

As explained previously, the focus group programme was delayed by more than 6 months because of delays created first by the ethical approvals and governance checks and then by the onset of ‘winter pressures’. Following the comments from set-up interviews, the aim was to involve four groups in particular at each participating site: clinical directors, directorate managers, ward sisters and matrons. However, where possible, it was also decided to involve the heads of support services departments, non-executive directors, service managers, clinical leads with service improvement responsibilities and staff in leadership and management development roles. Focus group discussions followed a topic guide (see Appendix 2) with three issues concerning motives and rewards, how management roles were changing and what would have to change to strengthen management contributions. Group discussions were each facilitated by two members of the project team, with one guiding the discussion and the other taking notes that were subsequently transcribed. Focus group discussions were not recorded (transcribing group conversations can be extremely difficult).

Focus group discussion transcripts were subjected to the same content analysis procedures as set-up interviews. The researchers involved produced a first analysis (see, for example, Figure 8), which was scrutinised by the principal investigator, with all focus group analyses being collated by the project research fellow. All of the focus group transcripts were then recoded by an independent researcher (using a NVivo database, version 9; QSR International, Southport, UK), whose choice of categories proved to be almost identical (see Table 42). As with the set-up interviews, the results of these analyses were also presented at the update workshop in 2011 and at the final project conference in 2012, thus increasing confidence that this analytical approach had accurately captured the issues raised by project participants.

The logistics of arranging focus group discussions proved to be more challenging than anticipated. The intent was to use existing meetings where possible, rather than arranging separate meetings and rooms. However, this involved, for each group, identifying a meeting that an appropriate number would be attending, at which the agenda was short enough to accommodate an additional discussion, and where the duration of the meeting permitted this. This arrangement worked well in many instances. Some staff did not appear to meet regularly with peers (support department heads at Netherby; assistant general managers at South Netley) and special meetings with mixed occupational groups thus had to be arranged. Where it was possible to use existing meeting schedules, the research often had to wait for two or more meeting cycles to occur before an appropriately ‘light’ agenda would arise, and ≥ 2 months would then elapse before that meeting could take place. At some sites it proved impossible to meet with particular groups at all during the life of the project – clinical directors and consultants at Burnside, divisional directors at Greenhill – not because of lack of interest in the study, but because of lack of diary space. A similar difficulty at Wattle Park was resolved by the research team member meeting individually with five clinical directors whose respective diaries appeared never to share empty spaces. We have no systematic evidence but many anecdotes of middle and senior managers deliberately and routinely double- and treble-booking diary slots, then hoping that one or more of those meetings would be cancelled for some reason, or making a judgement on the day as to which was the most important. One focus group meeting (Clearview) was cancelled 10 minutes after it had started as participants (an operations management team) were called to help resolve a beds crisis.

These logistical problems meant that the focus group programme, which was originally scheduled to run over 6 months, actually ran over an 18-month period. As explained above, this delayed a systematic analysis of the collected focus group information, which was one reason for involving an independent researcher to recode the transcripts towards the end of the project, in 2012. Nevertheless, 40 focus groups were conducted over 18 months with over 200 participants – around double the original number of targets. The main report does not explore the contrasts across the six participating trusts with regard to the findings from either the set-up interviews or the focus groups. This is because the conclusions – the motives, pressures, contributions, and desirable changes – were similar in each hospital.

Management survey

Details of the survey design and administration are provided in Appendix 3, along with an analysis of the sample characteristics. The majority of survey items were based on findings from set-up interviews and focus groups, and were designed to test those items with a larger sample. As explained in Chapter 5, the survey also incorporated a widely adopted organisational commitment scale, and a recognised set of stress indicators. A novel ‘extreme jobs’ scale was also built into the design, based on focus group discussions and on a ‘proof of concept’ pilot, facilitated by the project's SDO management fellow at one of the six participating trusts. The original plan was to use a ‘60–40’ design in which 60% of the survey items would be common to all sites, with 40% tailored (by local management) to local issues and priorities. Only two sites, Burnside and South Netley, took the opportunity to add a small number of additional items to the core survey; management at the other participating trusts were too busy in the second half of 2011 to contemplate this level of involvement.

The survey design passed through numerous iterations, within the project team, in discussion with the project advisory group and with the project's SDO management fellow (an operations manager) and following a (hard copy) pilot at one participating trust. Drafts of the design were also circulated to the lead management contacts at each site for comment. In addition to overall item coverage and detailed wording, one of the main problems was length: initial designs were too long and this would almost certainly have reduced the response rate. The final design thus reflects a series of compromises, resulting in a short instrument (70 items) that could be completed in less than 15 minutes.

The pattern of survey responses varied widely across the five trusts that were able and willing to participate. Wattle Park was engaged in merger discussions during the second half of 2011 when the survey was administered and did not want the distraction of an internal management survey. Senior management at Greenhill decided, without prior consultation with the research team, to restrict the survey distribution to their ‘100 leaders’ (as part of an internal development programme) rather than to the trust management population as a whole. Different administration methods were considered, including hard copy mailing and group administration, which would be complex and costly. The participating trusts all expressed a preference for online administration. In some respects this was relatively straightforward. Once the survey design had stabilised, it was uploaded to the Cranfield secure server and a unique link was created for each of the five trusts that would participate. That link was to be distributed at each trust in an e-mail from a senior manager (preferably our lead contact at each site). Through this route, the research team would know which trust the responses had come from but would have no personal details of respondents, other than in the replies that they provided. This approach, however, was not straightforward.

Two problems arose. First, the lead contacts who were to distribute the e-mail were those same busy managers who were double- and treble-booking their diaries by the second half of 2011. Inevitably, most delegated this task to a secretary, or to a colleague with little knowledge of this project, but who had work priorities of their own to deal with. The research team was not always informed of these delegations. This generated delays, and some errors, in the survey distribution. Second, the issue of directing the survey to the middle management population of ‘pure plays’ and ‘hybrids’ was dependent on the e-mail distribution lists at each hospital. Those lists use other staff categories and differ from one trust to another. The mailing strategy thus had to be determined separately at each trust, often by someone who, as just indicated, was not familiar with the project or its objectives.

Despite these problems, over 600 useable responses were received, with an overall response rate of 24%. (The annual NHS Staff Satisfaction Survey has a response rate of around 50%.) Descriptive statistics only are included in the main report, for two reasons. First, the sample size is small for inferential investigation; average cell values fall sharply and conclusions become meaningless. Second, the pattern of responses is similar across all five participating trusts and supports the qualitative findings, despite differences in local conditions.

Case incidents

The incidents reported in Chapter 6 were based in each case on interviews with relatively small numbers of staff who had been directly or indirectly involved in the incidents and on the associated documentation including statements, RCAs and timelines, investigation reports and, when available, progress reports. The evidence with respect to each incident varied significantly. The case of Mrs Mayland dated from 2005 and the Burnside C. difficile case dated from 2006, so there was consequently a greater volume of information available concerning those two incidents (both of which had been nominated by the respective trust chief executives as incidents that it would be useful to study). The case of Mr Torrens occurred at the end of 2010 and events relating to that incident were still unfolding during this project. The case of Mr Mitcham, a ‘never event’, occurred in late 2011, as the fieldwork for this project drew to a close. The collection of evidence concerning the latter two incidents was facilitated by the SDO management fellow seconded to this project, but nevertheless, given the sensitivity and recency of those incidents, less information was available in those cases than in the other two. As discussed previously in this appendix, it was once again not possible to gather precisely comparable evidence relating to each case.

Establishing access to the staff involved in these incidents was not problematic. On the contrary, those who were involved mostly welcomed the opportunity to share their experiences with external researchers (anonymously, and in confidence). As a further indication of this willingness to share information, the operating theatre team involved in the Mr Mitcham case asked two members of the project team, Buchanan and Denyer, to facilitate a workshop for them on the theme of ‘managing change in the aftermath of serious incidents’. These two project team members had already conducted workshops on this theme for several other audiences, and Denyer had previous experience with this topic in other sectors including nuclear reprocessing, fire and rescue services and high-security mental health care. The arrangements were once again brokered by our SDO management fellow and a 1-day (invitation only) workshop was designed to cover three main issues: using a ‘mess mapping’ tool to explore incident causality, helping the team to establish their own change agenda and identifying and tackling the barriers to post-incident change implementation. The ‘mess map’ for the Mr Mitcham incident appears in Chapter 6 (see Figure 7), along with information from a report produced by the trust's Never Events Group, which was provided to the two researcher-presenters as background to inform the running of the workshop.

This was not an action research intervention, but rather an attempt to provide assistance and support on request. Feedback from the event indicated that it was seen as valuable, and we were invited to repeat the workshop for another theatre's team. It was understood that information collected through this process would contribute to this project, appropriately anonymised. We are aware, however, that the risk management team at that trust have since been considering other systems-based approaches to the overall management of serious incidents and never events. With funding for this project coming to an end, the opportunity to follow its progress in that regard has not been available.

These kinds of incidents typically unfold over prolonged periods. The case of Mrs Mayland is a good example of this, with events and outcomes being shaped by a combination of factors at different levels of analysis interacting over time. This suggests a narrative-based processual approach to analysis. This has been started in Chapter 6, based on an ‘ideal narrative’: an understanding of incident causality should inform a change agenda, which should in turn trigger an appropriate change process. It appears, however, that this narrative is often ‘edited’ by events and circumstances such that the change process does not occur (or is partial or delayed) and a further similar incident occurs. The aim is to explain how and why this ‘editing’ takes place, and if possible how to prevent this in the interests of more rapid and effective change. It proved difficult within the length constraint of the project report to present the evidence (typically rich and interesting) surrounding the four cases along with more comprehensive theoretical development. The weight of treatment in the report, therefore, lies with the evidence, and theory development will feature in follow-on publications.

Limitations

As indicated in Chapter 1 of the main report, this study has several limitations.

First, data collection was uneven across the six participating trusts because of the mounting pressures and challenges facing hospital management at the time of this study, and the consequent difficulties in arranging and rearranging meetings (especially from late 2010 onwards). Initial plans for a ‘common data set’ to be collected from each trust had to be abandoned. The information that was gathered, however, demonstrates a mutually reinforcing internal consistency across the different data streams, which indicates that, had it been possible to gather a more standardised set of information, the conclusions would not have been much different. Nevertheless, the lack of a common data set for each participating trust does inhibit cross-site comparisons and contrasts.

Second, the extended time frame of this project potentially jeopardises the relevance of the findings. The SDO scoping study that led to the funding of this research was conducted in 2007–8. Publication of the final project report is in 2013. The NHS at the end of that period is a different organisation from the one that presented in 2007–8, in numerous respects, and particularly from an organisational and managerial perspective. However, the pressures and challenges facing hospital managers over this period appear not to have abated, but are continuing, and may escalate. Economic recovery in the UK is proving to be slower than anticipated, the government is still attempting to run down a budget deficit, in part by cutting government expenditure, and as a major contributor to that expenditure the NHS is being required to continue to make efficiency savings. It is possible that the findings and implications of this study may become more and not less relevant in this unfolding context. Nevertheless, the timeline of this project must be a cause for concern, and this is addressed in the concluding section of this appendix.

Third, coverage of the range of themes in this project – realities, contributions, changes following serious incidents – has been achieved at the expense of depth. It was not possible to follow up many interesting lines of enquiry because of the time pressures facing both researchers and participants. This was particularly the case with regard to some of the serious incidents, which were identified as valuable learning opportunities in the context of the research objectives, but which would have taken considerably more time to document effectively. This was also the case with many accounts of management contributions to clinical outcomes, especially in collaboration with clinical and/or medical colleagues, a theme that is developed in Chapter 7 but which merits closer attention. Nevertheless, we feel that this study has generated findings with useful implications for organisation and management, as well as identifying a number of interesting and valuable further research priorities.

Fourth, the quantitative support provided by the management survey for the findings from the qualitative components of this study will not satisfy a number of readers. Qualitative research in general still attracts suspicion (an appearance of lack of rigour is inevitable) and is often discounted by those with scientific and medical backgrounds who are more readily convinced by double-blind randomised control trials. There are a number of areas where quantitative methods would strengthen this study. One topic, for example, concerns the broad range of management contributions to clinical and organisational outcomes, with past studies tending to focus on single metrics and assuming a more linear causal model. Another area concerns the costs of not implementing, rapidly and effectively, the findings from investigations into serious incidents and never events. Those costs appear to be rising, and this research has proposed more effective approaches to change implementation in such conditions.

A study of this nature will always display imperfections. Nevertheless, we hope that the conclusions from this research will inform leadership and management development and contributions, and organisational change following extreme events, in useful, innovative and productive ways.

Lessons from this project

This experience suggests the following four lessons to inform future studies.