Leadership comes in many forms and is demonstrated at all levels of the organization. Senior leaders influence safety directly by setting up safety related committees and initiatives and allowing staff time to engage in fundamental safety issues, such as the redesign of systems. Leaders also influence safety indirectly by talking about safety, showing they valueitand being willing to discuss errors and safety issues in a constructive way. Safety is also strongly influenced by people in supervisory roles, such as a nurse in charge of a ward, both in the efficient management of processes and in the attitudes and values they foster in the people they manage. In turn, an individual nurse demonstrates her personal commitment to safety to trainees by attention to detail, by performing checks and by rigorous adherence to basic standards of care.

In this chapter we are primarily concerned with senior leadership, with the style, behaviours and actions of those with executive responsibilities. The transformational leadership theory of Bass and Avolio (1991) distinguishes two types of leadership, both of which are important for safety. Transactional leadership is, broadly speaking, an effective and efficient management style: set objectives, get agreement on what needs to be done, monitor performance and reward or sanction as appropriate. A theatre nurse, for instance, may set clear guidelines for behaviour and the standard of care they expect from the staff reporting to her. Transformational leadership in contrast, which Bass and Avolio suggest is characteristic of high performing teams, is more concerned with conveying a sense of purpose and vision, with empowering people in the team and treating people as unique individuals. A growing literature suggests that transformational leadership is significantly related to safety climate, staff compliance with rules and procedures, reduced accident rates and higher levels of performance, commitment and employee satisfaction (Barling, Loughlin and Kelloway 2002; Flin and Yule, 2004). Key behaviours are: articulating an attainable vision of safety; demonstrating personal commitment to safety; engaging everyone with relevant experience; and being clear and transparent when dealing with safety issues (Box 20.1).

BOX 20.1 Leadership behaviours for safety

(REPRODUCEDFROMQUALITY & SAFETY IN HEALTH CARE, R FLIN, S YULE. ‘‘LEADERSHIP FOR SAFETY: INDUSTRIAL EXPERIENCE’’. 13, NO. SUPPL_2, [45–51], 2004, WITH PERMISSION FROM BMJ PUBLISHING GROUP LTD.)

Safety and quality in high performing firms need not be seen as separate from financial and other matters, but an integral part of productivity and profitability. When Paul O’Neill, later US Treasury Secretary, was Chief Executive of Alcoa, he made safety his ‘signature issue’, starting all board meetings with that issue, always linking it to whatever was being discussed. Other people quickly learnt that paying attention to safety always engaged his attention, and the over riding importance of safety permeated the culture of the company (Berwick, 1999). As an example of someone who put safety at the very forefront of his tenure as Chief Executive, we consider Jim Conway at the Dana-Farber Cancer Institute in Boston, who led a root and branch reform after a high-profile patient death in 1994. Jim Conway’s reflections on the crucial elements of leadership for safety (Box 20.2) summarize and give life to the more academic treatment earlier in this section.

Ross Baker and colleagues identified a number of healthcare systems which, by reputation, had achieved major gains in safety and quality. We do not have sufficient information, either across systems or for the most part within these systems, to really identify the high performers from data, so these choices were essentially made on reputation. Nevertheless, the systems they chose have all undoubtedly placed safety and quality at the heart of their endeavours and there is much to learn. Baker and colleagues reviewed documents, carried out interviews with a wide range of people and, in the context of a broader understanding of the nature of system change, reflected on the journeys that these organizations had made. We cannot attempt to show the full richness of the study, just point to some of the most important observations in order to illustrate that there are different routes to high performance.

The VA system in the United States provides healthcare for Americans who have served their country in the military and, unusually in the United States, is under government control. The VA, like other systems, must respond to many external pressures and is particularly susceptible to political and government initiatives. Almost all VA physicians are salaried employees, again unusual in the United States. The service is imbued with a military ethos and staff regard caring for veterans as a particular privilege. The system is divided into 21 Veterans Integrated Service Networks (VISNs), which emphasize integrated treatment across a whole geographical region, encompassing all aspects of community and hospital based medicine; the 1990s saw a massive transformation from a primarily hospital based service to one that provided healthcare across an entire area. Many factors contribute to the high standards of care delivered but, in comparing the descriptions of different systems, the VA places a strong emphasis on structure, standardization and measurement.

The VA operates with a clear structure of national regulations, standards and practice guidelines. Care delivery is organized into five broad services, which are the same in each region: primary care, speciality and acute care, mental health, spinal cord injury and geriatrics. Standardization pervades every area of work, which brings a number of benefits. For instance, infusion pumps are standardized across the entire network, so training is simpler, mistakes are fewerand staff can easily move between unit sand regions. Summarized here in a few lines, this sounds simple enough; to achieve it was no doubt a monumental effort.

From the 1990s onwards, the VA has set clear goals and performance targets which track the care delivered across the entire network, enabling comparison of units and regions within the VA and over time. Each network monitors a basic set of indicators, which now run into the hundreds. These are derived from their own information, but also by data collected from record review by the VA External Peer Review Programme. The external review ensures that the data collected is rigorous, independent and hard to argue with. As one of the staff commented when discussing trying to persuade a department that improvement was needed, ‘the data speaks volumes’ (Baker et al., 2008).

The VA has invested heavily in electronic medical records and information technology of all kinds and keeping records in electronic form is now mandatory in almost all circumstances. The electronic medical record is critical to several other features; without it the measurement, the monitoring of workload and clinical work and the accountability for performance would simply not be feasible.

While the system is constantly monitored, it nevertheless also needs to evolve and change. Within the VA, the targets and goals are constantly stretched, though, ideally, not to the point of destabilizing the current system. Considerable thought has been given to how change is actually achieved, in contrast to many systems where improvement initiatives are runinspare time ortreated as secondary activities. In VA Boston, improvement work is clear and organized, and much thought is given to the proper resources and clinical leadership:

The big change in approach that allowed improvements in the performance measures was assigning teams and supporting and encouraging them… We are very thoughtful about who we select for physician leaders and participants and we are very clear about endpoints and expectations. We are very clear about who we want, why we want them and what we want them to do (Quality manager).
(BAKER ET AL., 2008)

While this selection of a few key features has emphasized internal structures, standardization and accountability, it would be quite wrong to see the VA simply in those terms. The VA also has a long history of external involvement with IHI improvement collaboratives and with engaging in new learning; their patient safety programme for instance, was one of the first to take learning from incidents seriously. We have not even touched on culture, leadership or the timescale of change. Nevertheless, the VA approach to improvement is undoubtedly grounded in anoverall structure of performance measurement and accountability and, in a sense, driven from the top while accepting that the improvements themselves must be generated on the ground. This contrasts quite strongly with our next example, where the style of leadership and change is very different.

Jonkoping County is a regional health economy southwest of Stockholm in Sweden. Sweden’s healthcare system is publicly funded, and at the local level controlled by 21 elected county councils. Jonkoping has 3 hospitals, 34 care centres and a staff of about 10 000, looking after the health needs of about 340 000 people. It is therefore aw hole health economy, but on are latively small scale. Jonkoping regularly tops the league of Swedish County Councils when ranked on the six goals of quality, efficiency, safety, patient centredness, equity and effectiveness. Their approach to high performance is distinctive in a number of ways, and we will select some of the most important.

Jonkoping Healthcare System has been led for almost 20 years by its Chief Executive, Sven Olof Karlsson, supported by Mats Boestig, medical director and Goran Henriks, who leads on learning and innovation. These three people and their enduring commitment have been hugely important in providing both the leadership and the stability necessary for serious, purposeful healthcare improvement. This has been achieved in spite of the fact that county councils are elected for four-year terms, and each new regime reviews the leadership of their healthcare. Jonkoping’s healthcare leadership are fully aware of the necessity for structures and the usual organizational processes and their achievements rest on a bedrock of strong financial discipline. However, they also believe that there are limits to this approach and that healthcare requires some new ways of thinking and behaving, if we are to achieve safe high quality care.

Esther is a fictional, but ever present, 88-year-old Swedish woman, who has a number of chronic conditions, but nevertheless manages to live independently with a good quality of life. From 1998 onwards, clinicians and managers in Jonkoping in various settings have mapped Esther’s journey through all the parts of the healthcare system, trying to see it through her eyes and structure it according to her needs. Seeing the process through the patient’s eyes has been the foundation of a host of changes, including new admissions processes, more integrated communication, team based telephone conversations and a strong emphasis on patient involvement and self care.

One of the most striking features of Jonkoping is the willingness, indeed hunger, to learn from others and to embed new and useful ideas and techniques within their system. Although Jonkoping now inspires others around the world, they continue to explore and maintain a continuous dialogue with external systems and individuals. For instance, a long relationship with the Institute for Healthcare Improvement has brought a range of improvement methodologies to Jonkoping, with an underlying emphasis on engagement of all staff. Paul Batalden has been a frequent visitor, collaborating on strategies for working on clinical microsystems.

Qulturum, led by Goran Henriks, is the learning and innovation hub of Jonkoping. Here staff come together for seminars, open discussion and conversation and here the new ideas are sifted, piloted and evaluated. Staff at Qulturum are usually clinical leaders and champions of quality improvement, though not necessarily senior people. They combine a sense of practicality with an interest in culture, theory and ideas. It is noteworthy that they use very few external consultants (as opposed to collaborators), always seeking to learn the new thinking themselves rather than buy it in.

The VA system engages staff in multiple improvement projects and creates time and resource to achieve particular goals. Jonkoping also does this, but their approach is more open and more fluid. Ross Baker quotes Karllson as saying:

We involve employees in lots of quality improvement projects and help them learn how to make changes and let them define how to create results using learning and innovation… results across the small parts of the system create results for the system and lots of winners. Big, high risk projects and changing structures in a traditional way… creates losers.
(BAKER ET AL., 2008)

Rather than working within a framework of targets and goals, the Jonkoping approach has been to encourage staff to improve quality wherever they deem it important, guided by Esther and their own assessments of local needs. Rather than control the improvement process, they take a more organic approach, in which many small changes combine and build on each other to create wider gains. It is the complex adaptive system in action.

While the improvements may initially stem from the staff themselves, they do need to be integrated into the broader strategy for the delivery of healthcare. This development came at a later stage, emerging from the more local activities. From 2001 onwards, Karllson and others initiated meetings of all senior leaders for five days over a year; he mirrored the open discussions he had fostered on the clinical level at the executive level. This has led to a model for system wide improvement, the Jonkoping Diamond, which explicitly links learning and innovation to both system improvement and to financial stability and discipline.

Both the VA and Jonkoping have clearly approached the improvement of safety and quality with great integrity, strength of purpose and seriousness. Both have leaders, at all levels, committed to improvement, who have fostered a culture of aspiration and continuous improvement. Both systems have invested heavily in educating and training their staff in safety and quality improvement

Both the VA and Jonkoping have realized that if substantial progress is to be made staff must be given the time and resources to do it. This sounds blindingly obvious, and it is; in practice however, improving safety and quality is often of marginal concern atthe higher levels of an organization. In the British National Health Service for instance, many posts cover such matters as patient safety, quality, risk and governance with varying job titles and responsibilities. Remarkably however, few of these people are concerned with improvement; they may be as individuals, of course, but their responsibilities seldom allow them to work on improvement. Their days are dominated by meeting standards, regulations and dealing with incidents and the occasional disaster. This is a slight caricature, but only slight. Improvement work is fitted in around other responsibilities and is extremely vulnerable to other priorities and events.

While there are similarities of approach between the VA and Jonkoping, we can also see that the strategies adopted by these two systems, whether consciously or not, have prioritized different features of the quality and safety journey. Both the VA and Jonkoping would identify performance measurement and organizational learningas critical to safety and quality improvement; however, measurement is more to the fore in the VA, and learning in Jonkop-ing. Leadership has been critical in both systems, but the styles of leadership have been different; in the VA there has been a strong element of monitoring and performance management, while in Jonkoping there has been more emphasis on staff empowerment and disseminated decision making. If we reviewed other systems studied by Ross Baker, Paul Bate and other researchers, we would see more variants on these and other themes. How should we interpret these different styles? Should we see themas necessary adaptations or do the approaches vary because we just don’t yet fully understand how to make change on such a massive scale.

Bate, Mendel and Robert (2008) have provided us with a valuable perspective on the issue of variety in strategic approach, drawing on the ‘universal but variable’ thesis common in the social sciences, which essentially states that:

There are only a limited number of basic human problems to which all people at all times and in all places must find solutions, but the number of possible solutions to them is unlimited.
(BATE, MENDEL AND ROBERT, 2008)

Bate and colleagues suggest that the healthcare systems they studied face six core challenges–the basic human problems of the formulation above. These are:

• Structural – structuring, planning and co-ordinating quality efforts;
  
• Political – engaging the relevant parties, negotiating conflict and relationships;
  
• Cultural – the challenge of giving quality a shared collective meaning, value and significance within an organization;
  
• Educational – creating and nurturing the learning process;
  
• Emotional – inspiring and energizing people;
  
• Physical and technological – designing physical systems and technological infrastructures that support improvement and the delivery of safe, high-quality care.

These common challenges are met in different ways by organizations; what works in one setting might not work for another organization. However, although Bate and colleagues emphasize that the journeys are different, they are clear that the challenges are non-negotiable and an organization neglects themat their peril. No matter how wonderful your technology, you still need to inspire people. Conversely, the most charismatic leadership will not get you far, unless backed by education, training and technical understanding. This last scenario was common in the early days of patient safety, when ‘changing the culture’ was held by some to be the royal road to patient safety.

If the factors for successful improvement are reasonably well understood, then why have healthcare systems not made more progress? Ross Baker and colleagues pose this incisive question in the introduction to their book on high performing systems. In part it is that every healthcare system that takes this journey is finding its way, although obviously inspired and guided by those taking similar journeys. Knowing the factors is, as Ross Baker puts it, a static understanding and does not show you the strategies, how to deploy the resources, the nature of the leadership and so on (Baker et al., 2008). However, to continue the journey metaphor, it also depends on where you are starting from and the resources available. One organization may be culturally ready, because they have reviewed and discussed safety and quality issues for years, but may not have developed formal organizational structures necessary for larger-scale transformation. Another may have the structures and the techniques, but a punitive and repressive culture which means that any attempts to engage staff are viewed with suspicion and even hostility. There are also a host of local circumstances which either support or impede these larger-scale activities. These range from the personalities of senior people in the organization, the huge range of other priorities that take resources from improvement, the fact that a merger has occupied the attention of senior staff for a year, a new building programme, and changes and the interest and knowledge that the wider staff have in safety and quality. There is, therefore, an awful lot to juggle and, just as there are many routes to success, there are innumerable ways of failing.

Many organizations do appear to have found a way of combining all the necessary aspects and making a sustained and long-term commitment to safety and quality improvement. There is, however, still some doubt as to how far this has translated into improvements for patients. This is partly because much of the data that might inform a response to this question has not been made public or reported; the VA for instance, clearly has large volumes of data but little of it is in the public domain. There is also, as we have seen, heated debate about the quality of data, the extent to which local improvement data reflects real improvement and whether it is necessary to ground all assessments in a formal clinical trial framework (Brown and Lilford, 2008). Measurement has also not been a strong point in even some of the finest programmes. John Ovretveit and Anthony Staines, clearly admirers of Jonkoping’s achievements, commented that despite the embedded culture of quality improvement, the extensive and effective learning and the consistent improvement work in many settings:

Jonkoping County Council still has little improvement data that would be credible to many researchers or clinicians. We noted the lack of a strong culture of measurement, we observed fewer measurement systems and activities than expected and relatively little specialised support to teams for designing, collecting, analysing and presenting data.
(OVRETVEIT AND STAINES, 2007)

Do we assume that patients are seeing the benefits or do we pessimistically conclude, as some do, that the safety and quality rhetoric far outstrips the evidence of real change? This depends on how we see the whole quality improvement process; is it like the action of a drug on a disease or is it a more organic, evolutionary process? Ovretveit and Staines point out that in the case of Jonkoping in particular, with a strong emphasis on staff engagement and change at ground level, the quality improvement process is likely to take time to produce results, but when they are seen they should be widespread. Conversely, if you plough all your resources into a small number of critical targets, such as waiting times in the emergency department, then you will see results but probably only in very specific areas (Bevan and Hood, 2006).

Ovretveit and Staines suggest that, although leaders hope for rapid change, there is probably a necessary period of building a quality infrastructure. There may even be a threshold to achieve before real improvements are seen. In the first phase then, perhaps lasting some years, it isnecessary to investina ‘quality infrastructure’, but not to expect any immediate return on investment. In the short term one hopes for engagement and enthusiasm and the sense of embarking on a journey; in this phase one is building awareness, leadership and freeing up resources to create the capacity for improvement. In the subsequent phase, one aims for some signs of improvement in specific areas; these may be incomplete, inconsistent and may be hard to distinguish in the face of other changes and developments. In the longer term, when measurement and capacity are mature, then patient experience and outcomes begin to improve in a sustainable manner; they suggest that this could takeas longas ten years, even in as successful a system as Jonkoping.

The idea that a period of investment is necessary before returns are seen potentially makes sense of some of the conflicting evidence. Enthusiasts are not being na€ıve in seeing real change in an organization, even if more sceptical observers believe that patients are no better off. However, the enthusiasts need to understand that there is much further to go than they anticipated. The timescales we have discussed, however, are not absolute, but more reflect the current understanding of large-scale quality improvement. Only a very few systems have probably understood the nature and scale of capacity development that is actually needed; most have relied on enthusiasm, culture change and on people doing quality improvement in their non-existent spare time. If we trained people properly, gave them time and resources, and set up proper economic and clinical evaluations, then the period between initial investment and tangible patient outcomes could be a great deal shorter.

If this book has succeeded in its aims you will, I hope, be convinced that patient safety is critically important for both patients and healthcare staff in every setting throughout the world. Hopefully too, something has been conveyed of the landscape of patient safety, the central concepts and an understanding of the nature, causes and prevention of error and harm. Perhaps you will also agree with the assertion in the preface that patient safety is a tough problem.

At the end of the first edition, I wrote that we had many good ideas and concepts, some solid evidence and many promising avenues to explore, but we were nevertheless still at the beginning of the safety journey. Although we understood the problem quite well, many interventions were haphazard in nature and their effects uncertain. Furthermore, we did not understand how to integrate the various safety interventions and components of safety into a coherent whole. As I hope the later chapters have shown, the last five years have brought considerable progress on many fronts and there is now reason for real optimism about what can be achieved. The earlier years of patient safety were mainly devoted to uncovering innumerable holes in the healthcare Swiss cheese. Now we can envisage the possibility of attaining the reliability and resilience associated with the classic Swiss watch.

Above all, for anyone who has read this book, working or associated with healthcare in whatever capacity, I would hope that you feel that patient safety is a subject worthy of your attention. Understanding and creating safety is a challenge equal to understanding the biological systems that medicine seeks to influence. While the challenge is immense, it is clear that we are making some progress in awareness, in understanding and on action to prevent harm and care for those affected. Treating patients one at a time brings obvious and immediate benefits, but working to improve the safety of healthcare as a whole may ultimately benefit many more.

Baker, G.R., Macintosh-Murray, A., Porcellato, C. et al. (2008) High Performing Healthcare Systems. Delivering Quality by Design, Longwoods, Toronto.

Barling, J., Loughlin, C. and Kelloway, E.K. (2002) Development and test of a model linking transformational leadership and occupational safety. Journal of Applied Psychology, 87, 488–496.

Bass, B.M. and Avolio, B.J. (1991) The Multifactor Leadership Questionnaire, Consulting Psychologists Press, Palo Alto CA.

Bate, P., Mendel, P. and Robert, G. (2008) Organising for Quality. The Improvement Journeys of Leading Hospitals in Europe and the United States, Radcliffe Publishing, Oxford.

Berwick, D. (1999) Taking action to improve safety. How to improve the chances of success, in Enhancing Patient Safety and Reducing Errors in Healthcare, National Patient Safety Foundation, Chicago IL, pp. 1–11.

Bevan, G. and Hood, C. (2006) Have targets improved performance in the English NHS? British Medical Journal, 332 (7538), 419–422.

Brown, C. and Lilford, R. (2008) Evaluating service delivery interventions to enhance patient safety. British Medical Journal, 338, 159–163.

Ferlie, E.B. and Shortell, S.M. (2001) Improving the quality of health care in the United Kingdom and the United States: a framework for change. Milbank Quarterly, 79 (2), 281–315.

Flin, R. and Yule,S. (2004) Leadership for safety: industrial experience. Quality and Safety in Health Care, 13(Suppl_2), ii45–ii51.

Ovretveit, J. and Staines, A. (2007) Sustained improvement? Findings from an independent case study of the Jonkoping quality program. Quality Management in Health Care, 16(1), 68–83.

Westrum, R. (1997) Social factors in safety-critical systems, in Human Factors in Safety-Critical Systems (eds R. Redmilland J. Rajan), Butterworth-Heinemann, Oxford, pp. 233–256.