Christopher Bunch, Oxford Radcliffe Hospitals NHS Trust, The John Radcliffe, Headley Way, Headington, Oxford OX3 9DU, UK. E-mail: email@example.com
The concept of clinical governance was introduced to the UK National Health Service (NHS) by the new Labour government shortly after its election in 1997. In its first white paper on the NHS (Department of Health, 1997), the government signalled its intention to dispense with the bureaucracy and competitive environment (the ‘internal market’) that it felt characterized the previous government's reforms (Department of Health, 1989) and to focus instead on the quality of the service by insisting on collaboration rather than competition, by modernizing the NHS and its infrastructure and by defining key accountabilities and responsibilities to include the quality of clinical care as well as financial performance. Clinical governance is set to be a central feature of this new direction for the NHS.
This emphasis on quality rather than finance has been broadly welcomed by professional groups within the NHS as a refreshing change from the direction of the previous administration, although it is now becoming clear that financial control will remain tight and, although funding will be somewhat more generous over the next few years, that success will be measured in terms of performance indicators such as the numbers of patients treated, waiting times and simple outcome measures such as 30-d mortality after surgery.
The New NHS • Modern and Dependable defined clinical governance only in its glossary (Box 1) (Department of Health, 1997) and gave little indication of how it might work in practice. Eight months were to pass from its publication until detailed proposals appeared in A First Class Service, a consultation document published to coincide with the 50th anniversary of the NHS (Department of Health, 1998). Interestingly, this document uses different rhetoric to define clinical governance (Box 2) – one that is founded more in the softer language of continuous quality improvement than the risk-averse tone of the earlier definition. In practice, however, these are not mutually exclusive, and clinical governance has roots in both risk management and quality assurance (Vincent, 1997).
The origins of clinical governance
Although a novel term in itself, clinical governance introduces no fundamentally new concepts, and many of its components are well understood if not as widely practised as perhaps they should be. At one level, it may be viewed as the extension to clinical activity of well-established trends in corporate governance, which in turn reflect moves to define more transparently the responsibilities and accountabilities of organizations and those charged with running them.
A need for higher standards of governance in private companies was highlighted by various financial and operational disasters during the 1970s and 1980s which called into question the competence of certain companies to manage their affairs. Although financial problems may cause irritation, inconvenience and financial hardship to a company's stakeholders, an undoubtedly greater concern is the catastrophic effects that operational failures can have on the safety and well-being of its clients and the public at large.
The principles of corporate governance were reflected in the Cadbury and Greenbury reports which specified standards for the boards of companies in the conduct of their affairs (Committee on Corporate Governance, 1997). These principles have now been widely adopted and are, to a large extent, reflected in the constitution and composition of the boards of NHS hospital trusts.
Defining standards of conduct for the management of organizations and the accountabilities and responsibilities of its board and directors does not in itself guarantee improvements in quality of management or reductions in corporate risk. It is difficult (and in large organizations inappropriate) for individual directors to have direct personal control over every aspect of their organization's activity. What is needed is to decouple control by defining the operational standards, policies and procedures to be followed throughout the organization and to establish ways of measuring activity that can reassure directors and the board that the organization is working as intended, that any risks are properly identified and that action is being taken to reduce them.
This concept is known as controls assurance and is particularly appropriate for very large organizations and those in which there may be a high degree of decentralization. The NHS fits the bill in both respects: it is undeniably very large and it decentralized significantly by establishing self-governing hospital trusts during the early 1990s; however, in contrast to its predecessors, the present government is clearly exerting tighter central control. At present, it seems unlikely that the government will dismantle hospital trusts. Indeed, it intends to extend the principle of self-governance to primary care trusts in due course, although at the same time it is taking new powers to intervene directly in trust affairs if it sees fit (Hansard (Lords), 1999).
Although self-governing, hospital trusts are not autonomous; they remain part of the NHS and are subject to its external controls, including those exerted by government. In this way, trusts are required to implement inter alia internal systems of corporate and clinical governance. It remains to be seen, however, to what extent the present government will be content to leave trusts and primary care groups to develop and operate these systems for themselves. The establishment of two new statutory bodies (the National Institute for Clinical Excellence and the Commission for Health Improvement) is further evidence that trusts will be subject to much more external control than previously.
The language of corporate governance and controls assurance is foreign to most clinicians, many of whom may wish to deny that they work within an organization at all (be it the NHS, let alone an individual trust) but rather imagine that the organization exists to enable them to fulfil their vocation and is to be criticized if it fails to provide the resources for them to do so. Nevertheless, these concepts are important to a proper understanding of how clinical governance has come about.
The clinical influences on clinical governance stem in part from attempts in the 1980s to halt the rising costs of health care by encouraging greater involvement of clinicians (particularly doctors) in the management of their service and its associated resources:
• The introduction of general management to NHS hospitals followed the recommendations of the late Sir Roy Griffiths, who in fact believed that management should to a large extent be the responsibility of clinicians (Griffiths, 1983, 1991).
• The Resource Management initiative, launched in 1989, encouraged management decentralization in hospitals through the formation of clinical directorates (Buxton et al, 1989). The clinical director (typically a consultant) would manage the budget for the service and have overall responsibility for delivery of the service (Hopkins, 1993).
• The medical Royal Colleges were encouraged to promote systems of medical (later clinical) audit to develop a greater appreciation of the processes and outcomes of clinical care (Amess et al, 1995).
• Comparisons between (and within) hospitals demonstrated significant qualitative and quantitative variation in the performance of many aspects of clinical care that could not be explained simply on chance (Wennberg et al, 1987). The uptake of proven advances was patchy and slow, as was the cessation of interventions shown to be of dubious or no value. This has paved the way for the concepts of clinical effectiveness (ensuring that effective interventions are taken up, and ineffective interventions dropped) (NHS Executive, 1996), and evidence–based medicine, which seeks to ensure that interventions are supported by best–available evidence (Sackett et al, 1996).
Such initiatives are not peculiar to the UK; similar approaches have been evident, particularly in North America (Heyssel et al, 1984). Their success, however, has been debatable and they have their detractors. The National Audit Office in the UK has questioned whether the investment made so far in clinical audit represents good value for money (National Audit Office, 1995). This, however, has more to do with the manner of its implementation than the concept itself. Intuitively, it seems perfectly sensible that we should regularly inspect, audit and reflect upon our performance, and that everything we do should be supportable by good evidence of efficacy – if not by the gold standard of a randomized clinical trial, then at least by a demonstrable body of experience or opinion. It is unlikely therefore that the concepts themselves are at fault, but rather that we have much yet to learn about how to put them into practice effectively.
Several well-publicized clinical disasters have shaken public confidence in the NHS, particularly in doctors. The most notable of these have been catastrophic failures in breast screening and cervical screening programmes, the Bristol paediatric cardiac surgery debacle in which surgeons continued to perform operations despite repeated expressions of concern that their results were way below the expected standard (Horton, 1998; Keogh et al, 1998), and several high-profile cases of substandard care delivered by individual consultants. The way that the General Medical Council (GMC) has handled some of these issues has provoked serious questions about the medical profession's ability to regulate its own standards and affairs (Garrett & Savage, 1998).
Managerial accountability and professional autonomy
Like other professions, the medical profession has proved difficult to manage. Clinical freedom is highly valued and is treated almost as a divine right (Mirvis, 1993; Richards, 1998; Charlton, 1999). However, for as long as patients have ceased to pay directly for their care, but indirectly through insurance or taxation, those responsible for paying have striven to curb the freedom of doctors to do more or less as they please (Light et al, 1986; Light & Levine, 1988). The lack of sound evidence for much of clinical practice, wide variations in practice and repeated failures to observe the maxim primum non nocere have given them good reason to do so.
What does clinical governance mean in practice?
The essentials of clinical governance for practising clinicians can be summarized as being able to:
• demonstrate that we are doing things right;
• demonstrate that we are up to date and fit to practise;
• demonstrate continuous improvement in the quality of clinical services.
Doing things right implies both that the right (i.e. appropriate, clinically effective and evidence based) things are being carried out, and that they are actually being carried out rather than just talked about. Demonstrating this will mean participating in appropriate clinical audit activities, which may be locally or nationally organized.
Demonstrating that we are up to date and fit to practise means that we can show that we are practising up-to-date medicine, that we are taking steps to keep abreast of advances in our field, that we are developing professionally, that we are healthy in mind and body (and thereby not a danger to ourselves or our patients) and that we are behaving towards our patients and colleagues in a professionally appropriate manner. In other words, that we are living up to the expectations articulated by the GMC in its published standards (General Medical Council, 1998).
The emphasis on being able to demonstrate these qualities is central to clinical governance and links to the concepts of corporate governance and controls assurance described above. Participation in these activities has previously been haphazard if not actually optional, but it is now the expectation of both government and professional leaders that these standards will now be universally adopted (Donaldson, 1998).
Continuous improvement of clinical services requires that doctors and other health professionals work together in teams ‘in ways that best serve patients’ interests'. Continuous quality improvement is well understood in industry and, increasingly so, in the public sector, and several hospitals in the UK and Europe have in recent years adopted the European Foundation for Quality Improvement model (European Foundation for Quality Improvement, 2000). There is an important leadership role for doctors in quality improvement (Berwick, 1996), although the medical profession should not assume that it will always lead in this area.
The implementation of clinical governance
Hospital trusts and primary care groups are expected to establish local systems for clinical governance following a framework described in A First Class Service (see Box 3), and supported by a number of national initiatives (British Association of Medical Managers, 1998; Department of Health, 1998).
Components of a clinical governance framework
As far as possible, the implementation of clinical governance should build on existing systems and initiatives. Nevertheless, certain features are central to the overall concept (Box 3) and will need to be put into place if not already present.
Although the chief executive now has statutory responsibility for clinical quality, trusts and primary care groups are required to identify a lead clinician who will have day-to-day responsibility for the implementation of clinical governance. To ensure the appropriate accountability, in NHS trusts this individual is typically an executive director, i.e. the medical director or nursing director.
Trust board committee
The Boards of NHS Trusts are required to have an Audit Committee to provide independent advice to the Board on the financial probity of the organization. Similarly, it is anticipated that the Boards will establish a Clinical Governance Committee to undertake an analogous function for clinical standards and quality, although this may not be a statutory requirement. The Audit Committee normally comprises non-executive directors, with the finance director in attendance or reporting to the Committee. A similar non-executive model could be appropriate for clinical governance, with the lead clinician (and probably also the chief executive as accountable officer) reporting on the implementation of clinical governance.
An alternative model which is being adopted by many trusts is to establish an executive committee, chaired by the lead clinician (British Association of Medical Managers, 1998). This may be a practical way for the lead clinician to discharge some of his or her responsibilities, but the accountability to the Trust Board will need to be demonstrable. The two approaches are not of course mutually exclusive; larger trusts may opt for both approaches, with the Trust Board Committee performing essentially a monitoring role on behalf of the Board (rather like a Parliamentary Select Committee) and the other driving the actual implementation of clinical governance.
Trusts will be expected to be able to demonstrate a range of activities designed to ensure quality improvement. These include widespread participation in clinical audit programmes, routine application of evidence-based practice and improvements in clinical record keeping and information collection generally.
Comprehensive clinical quality improvement programmes have become relatively widespread in the USA (Kassirer, 1993; Berwick, 1996; Berwick & Nolan, 1998), but have been less systematically pursued in the NHS (Berwick, 1998).
Several systematic approaches to quality improvement have been promoted within the NHS: for example, the King's Fund organizational audit, Investors in People, together with various external accreditation schemes such as Clinical Pathology Accreditation (CPA). Approaches adopted in North America include examination of appropriateness of clinical interventions (Phelps, 1993), implementation of practice guidelines (Lomas et al, 1989; Cluzeau et al, 1996; Renvoize et al, 1997; Woolf et al, 1999), outcomes research (Epstein, 1990) and efforts to improve physicians' behaviour (Greco & Eisenberg, 1993). These are not mutually exclusive, although each has its problems (Kassirer, 1993).
Ensuring and demonstrating effective clinical care involves both knowing (or being able to find out) what is effective and what is not, putting it into practice and regularly reviewing or auditing practice to ensure that processes have been followed correctly and/or outcomes are as expected (NHS Executive, 1996).
Choosing the most appropriate investigation and treatment in a given situation has traditionally been guided by a combination of skill, acquired knowledge and experience. The rate of growth of knowledge, however, makes it nearly impossible for the average practitioner to keep abreast of the literature, let alone to spend time critically reviewing and appraising its worth. The so-called evidence-based medicine movement has grown out of an attempt to develop an educational paradigm designed to equip the practitioner with critical appraisal skills to underpin day-to-day clinical decision making (Rosenberg & Donald, 1995; Sackett & Rosenberg, 1995; Sackett et al, 1996).
Keeping up to date and abreast of the literature is aided by the development of data bases of critical reviews, for example the Cochrane Library, and by published reviews, e.g. Effective Health Care Bulletins from the Centre for Reviews and Dissemination in York and the American College of Physicians (ACP Journal Club). New skills need to be acquired to make best use of these resources (Rosenberg et al, 1998), and medical libraries will also have to be developed to support the approach generally (Broerling, 1993).
Knowing that one's practice is what one thinks it is and that it is as effective as intended requires regular review or audit. Medical audit has had a chequered and largely unsuccessful career so far (Buttery et al, 1995; National Audit Office, 1995; Hopkins, 1996), Early attempts at audit were, however, flawed in three significant ways: (1) it was exclusively medical and thus focused narrowly on just one part of the total management of the patient; (2) it was intermittent and project based and, thus, susceptible to selection bias (i.e. auditing practice that is going well rather than difficult areas); and (3) those undertaking it were poorly trained, a factor aggravated by the rather low status of the activity generally. It is now appreciated that successful audit is comprehensive, multidisciplinary and a part of everyday practice (Teasdale, 1996; Auplish, 1997; Ayres et al, 1998; Burnett & Winyard, 1998).
Clinical risk management will be an important component of clinical governance. It developed initially as a defence against litigation, especially in North America where escalating litigation costs have a longer history than in the UK. This has tended to produce a rather negative image, one associated with criticism of failure, damage limitation, etc., focusing on protection of the organization rather than the interests of the patient. It has also tended to perpetuate defensive medical practices, further driving up the cost of health care. More recently, however, it has developed a more positive, proactive role in reducing harm and improving the quality of care (Vincent, 1995). It has become customary to separate clinical from non-clinical risk management, but in practice there is much overlap in both substance and principles and many trusts will take a common approach.
Two significant changes in the past 10 years have altered the way in which the NHS approaches clinical and non-clinical risk.
Before 1990, all NHS doctors were individually responsible for the financial consequences of medical negligence and were contractually obliged to subscribe to a defence organization or insurance company to cover this. In 1988, the Government agreed to reimburse two-thirds of the subscription cost, but in 1989 – faced with rising subscription levels and the declared intention of one major defence organization to introduce differential subscriptions based on specialty – it decided instead to provide ‘Crown indemnity’ for hospital doctors, placing the financial burden for this on health authorities (Brahams, 1989). Employers are liable for the acts and omissions of their employees committed in the course of their employment, and this change in effect gave hospital doctors the same indemnity as other employees.
Similarly, before 1990, the NHS was exempt from several aspects of legislation that applied to other organizations, notably in the areas of environmental health and safety. This loss of ‘Crown immunity’ has had a significant impact on health authorities and, latterly, trusts (who have never been Crown bodies), forcing them to improve standards and manage associated risks. Many of these risks have been covered by commercial insurance policies, but insurance companies have shown an active interest in their clients' operational standards and risk management capabilities and have sought to improve these. In 1998, however, the Government became concerned about the cost of insurance premiums, and now plans to extend NHS self-insurance to cover such risks by a risk-pooling scheme similar to the mutual scheme for clinical negligence described below.
Health authorities (and latterly NHS hospital trusts, which have assumed many of the relevant responsibilities) were initially ill-prepared organizationally and financially for these changes. Risk management capabilities were poorly developed, and sensible guidance was not provided until 1993 (Department of Health, 1993). In 1995, a mutual risk-pooling scheme known as the Clinical Negligence Scheme for Trusts (CNST) was introduced which now covers the majority of hospital trusts in England. The Scheme is supervised by a special health authority, the NHS Litigation Authority, although its operation is subcontracted to a commercial organization. Its objectives are primarily to protect member trusts from the financial consequences of clinical negligence, to promote good standards of risk management and claims management and, thereby, to improve the quality of patient care (Hickey, 1995).
CNST only covers high-cost claims (above a threshold dependent on the level of premium paid) received since the member trust joined the scheme. Existing liabilities are covered separately. This has allowed premiums to be set at a relatively low level initially, but these are now rising significantly as liabilities are being met. High-risk trusts (e.g. those with maternity or neurosurgical departments) pay more than low-risk trusts (e.g. community), and the scheme is organized on a ‘pay-as-you go’ principle, collecting from members each year only the sum CNST expects to pay out that year nationwide.
Trusts are able to earn substantial discounts by being able to demonstrate high standards of risk management to external inspectors (Hickey, 1995). This has produced a significant incentive for trusts to improve their standards. So far, these have focused in the main on generic issues, e.g. the organization of risk management, development of incident reporting, managing complaints, patient information and consent, clinical documentation and record keeping, the management of medical records, induction of clinical staff and more specific standards for maternity services. In future, standards will be extended to cover a range of specific clinical issues. It is also likely that similar initiatives will cover non-clinical areas of risk if and when the NHS Litigation Authority takes responsibility for areas currently covered by commercial insurance.
Even before the Bristol affair came into the public eye, professional leaders in both medicine and surgery were exploring ways of detecting and dealing with poor performance. In 1997, the General Medical Council introduced new fitness to practise procedures, enabling them to intervene when significant concern has been expressed about an individual doctor's performance. These procedures are different from the GMC's disciplinary procedures and are intended to be remedial and supportive. By their nature, however, they are only ever likely to come into play when performance is very poor indeed.
Following Bristol, the GMC has taken stock, has revised its core advisory publication on medical practice (General Medical Council, 1998), has undertaken a pilot study for local medical regulation and has recently published proposals for the regular revalidation of doctors, linked to registration.
Trusts implementing clinical governance will need to develop their own local procedures for dealing with performance problems in clinical staff. These will require sensitive handling and close co-operation between the professions and the personnel department. It will be only too easy to create a climate of fear. Staff need to know that they can express their concerns freely, and without criticism, and that these concerns will be taken on board and dealt with appropriately. Early resort to disciplinary proceedings should be avoided. However good systems and processes are, mistakes will always happen, and every attempt should be made to learn from these and not to be too quick to blame.
Continuing medical education (cme)
Although clinical governance introduces no new concepts, many new skills will need to be learned and, as already discussed, clinical staff will need to strive to keep up to date. Advising on specialist medical education has been the traditional remit of the Royal Colleges and specialty associations, and these are now actively promoting schemes for continuing medical education. The approaches being taken, however, are not necessarily the right approach.
A review in 1992 sifted 50 randomized controlled trials of CME from 1445 articles on the subject (Davis et al, 1992). Simple lecture-based activities (probably the most common approach) tend in themselves to be insufficient to change practice. Successful programmes have simple goals and go beyond simply informing participants by providing or facilitating activities that reinforce learning.
Trusts as employers have responsibilities to ensure that their workforce is up to date and fit to practise, and will need to allow adequate time in job plans for this to happen. Much better would be for them to take an active interest in the personal development of permanent medical staff, who represent a significant long-term investment. Job plan reviews can be developed from anodyne statements of fixed commitments to a more rounded plan for personal and professional development.
National institute for clinical excellence
A National Institute for Clinical Excellence (NICE) was established as a Special Health Authority for England and Wales on 1 April 1999. Its remit is to ‘provide patients, health professionals and the public with authoritative, robust and reliable guidance on current best practice’, and also to develop and disseminate associated clinical audit methodologies (National Institute for Clinical Excellence, 2000).
NICE intends to work closely with academic centres and existing organizations and inherits many established initiatives supported by the Department of Health. Its functions are primarily to do with setting and disseminating standards for clinical care, but it is not expected to be directly involved in their implementation.
The principal areas of activity for NICE are:
• To be a single access point for clinical practice standards.
• Appraisal of health technology, including cost effectiveness of new drugs and appropriateness.
• Production of disease management guidelines, e.g. primary care management of asthma and diabetes.
• Development of audit methodologies, which it is anticipated will be built into standards and guidelines. It may have an extended remit, e.g. taking over the existing national confidential enquiries.
• To be an umbrella for all clinical standards/audit initiatives that receive Government funding, including existing publications such as the Effectiveness Bulletin and Prescriber's Journal.
It is anticipated that NICE will also focus on the quality of guidelines and that it will incorporate published information in this area.
Commission for health improvement
A second new statutory body, the Commission for Health Improvement, was established in 2000 ‘to provide independent scrutiny of local efforts to improve quality and to help address any serious problems’ (Department of Health, 1998).
The Commission will oversee and promote the development of clinical governance nationally and plans to undertake a rolling programme of reviews, visiting every hospital and primary care trust every 3–4 years. The reviews will focus on both process and outcome. For example, it might examine complaints handling and/or incident reporting, looking not just at the systems themselves but also at whether lessons have been learned and practice changed as a result. They will also examine the trust's progress in implementing NICE guidelines and national service frameworks.
National service frameworks
In 1994, the Chief Medical Officers of England and Wales published a policy framework for commissioning cancer services (Department of Health, 1995). This sought to address variations in service provision by establishing standards for the delivery of cancer services at all levels – from specialist centres to primary care – and its principles have now been widely adopted.
The government is building on this concept by developing a series of National Service Frameworks. Frameworks for coronary heart disease, mental health and paediatric intensive care are have recently been published, and a framework for care of the elderly is expected in the near future (NHS Executive, 1998).
The Health Service and its professions face a huge challenge to introduce a degree of openness about the quality of its activities and the systems to achieve it that is unparalleled in its history. Why has this challenge arisen, and is it unique to the NHS? It appears to have some roots in a growing frustration among payers – and more recently the public – with the spiralling cost of health care, its unexplained variability, its propensity to do damage as well as good and the perceived secrecy and insensitivity of the medical profession particularly in its dealing both with individual patients and those responsible for resourcing its activities.
Private industry – in both the manufacturing and service sectors – has embraced formal approaches to safety, risk, quality improvement and general openness much more readily and to a far greater extent. But this does not appear to be simply a characteristic difference between the public and private sector. The US health care industry – which is largely private – faces similar challenges from its payers and similar reactions from the medical profession. In part it has to do with perceived threats to professional autonomy, and a long-standing struggle for control. Professions have always jealously guarded their independence and their right to regulate themselves, but recent events have eroded confidence in their ability to do so.
It is easy to see, therefore, how some aspects of clinical governance may be less palatable in practice than in theory. Clinical governance should, however, be seen as an opportunity rather than a threat. For the first time in some decades, it has elevated the quality of health care in the UK to the same level of importance as its cost. Underlying this, of course, is a strong belief that good quality costs less than bad quality in the long run, but it must be remembered that changing course requires the expenditure of energy and some resource will be necessary to make things happen.
With autonomy comes responsibility: embracing the principles of clinical governance and putting them into practice will go a long way towards restoring the public's confidence.