Intended for healthcare professionals


Risk, safety, and the dark side of quality

BMJ 1997; 314 doi: (Published 21 June 1997) Cite this as: BMJ 1997;314:1775

Improving quality in health care should include removing the causes of harm

  1. Charles Vincent, Senior lecturera
  1. a Clinical Risk Unit, Department of Psychology, University College London, London WC1E 6BT

    Clinical risk management was initially considered a means of controlling medical negligence litigation. Gradually, however, the need systematically to examine the underlying clinical problems became apparent, together with the need to care for injured patients rather than simply treating them as potential litigants.1 Though driven by anxiety about litigation, risk management has the potential to act as a gateway into a more important problem, which current quality initiatives have not adequately addressed: injury to patients.

    Can care that is actually harmful be encompassed in traditional frameworks of quality, such as Maxwell's dimensions of effectiveness, efficiency, appropriateness, acceptability, access, and equity?2 Certainly harmful treatment will be ineffective, inappropriate, and unacceptable but these terms imply an absence of quality rather than actual danger or harm. Maxwell's dimensions are important, but, in the positive way they have been interpreted, have perhaps directed attention away from quality's darker side.

    Iatrogenic effects of drugs and other treatments have been recorded in many studies, but only recently has the scale of injury to patients become apparent. The Harvard study found that patients were unintentionally harmed by treatment in almost 4% of admissions in New York state. For 70% of patients the resulting disability was slight or temporary, but in 7% it was permanent and 14% of patients died partly as a result of their treatment.3 Serious harm therefore came to about 1% of patients admitted to hospital. Similar findings were reported from Colorado and Utah in 1992 (personal communication, T Brennan). A recent Australian study revealed that 16.6% of admissions resulted in an adverse event, of which half were considered preventable.4 Extrapolating from these figures to a hospital with 50 000 admissions a year suggests 2000-8000 adverse events a year leading either to injury or to a longer hospital stay.

    The cost of litigation in Britain has been estimated at £100m to £150m a year,1 now probably nearer £200m. This is about 0.5% of the NHS budget and could be viewed simply as an insurance problem. The calculations look different, however, when the real costs of adverse events are considered, as these vastly outweigh the costs of litigation. An operation with an adverse outcome may lead to further operations, a longer stay in hospital, additional outpatient appointments, and so on. In Australia adverse events were estimated to account for 8% of all hospital bed days.4 Disability payments and other benefits probably far outweigh the costs of individual hospitals. Adverse events also involve a huge personal cost to the people involved, both patients and staff.1

    Recent research in medicine and elsewhere suggests actions that should be taken to improve safety. Firstly, the dictum of first doing no harm needs to be taken seriously.5 6 No treatment is risk free, but safety should at least be recognised as the first dimension of quality. An airline that functioned in the manner of the hospitals in the Harvard study could be generally effective, efficient, and so on, save for the 1% of journeys in which passengers were seriously injured.

    Potential for error must be acknowledged

    Secondly, the potential for error in medicine needs to be openly acknowledged. Skilled and routine tasks can be carried out reliably, but the probability of error and biased decision making increases greatly in novel situations as the limits of the brain's information processing capacity are reached. At times of stress and uncertainty the probability of mistakes inevitably increases, no matter how conscientious the clinician.7 Nevertheless, in medicine fault free performance is expected always. Error prevention involves exhortation, training, or disciplinary measures directed at the individual.6 But people do not intend to make errors and it is often pointless to chastise them for doing so.

    Thirdly, analyses of accidents in medicine and elsewhere have led to a broader understanding of accident causation, with less focus on individuals and more on organisational factors that provide the conditions in which errors occur.8 In medicine the root causes of adverse events may lie in factors such as communication and supervision, workload, educational and training deficiencies, the use of locums, and so on. The same organisational problems, such as poor communication within a team, may lead to a wide variety of adverse clinical events. After serious incidents the first question should be, “What does this tell us about our system?” and only then, “What does this tell us about this individual?”9

    Fourthly, targets for audit, clinical effectiveness programmes, and other quality initiatives should be explicitly selected from areas where patients are at high risk. The American and Australian studies show that high rates of preventable adverse events (20-30%), and of resulting permanent disability, occurred during treatment for infectious diseases, injuries, and poisonings and from the toxic effects of drugs.4

    Fifthly, the fact that risk management is generally centred on single adverse events may bring a new richness to the analysis of the quality of health care. Practitioners of audit know that “counting is not enough,”10 in that descriptive data do not necessarily reveal the underlying problems, and some argue that successful audits are those that specifically analyse underlying causes.11 Critical incident and organisational analyses of individual cases have illustrated the complexity of the chain of events leading to an adverse outcome.12 13 14 Finally, clinical risk management should aim at more than avoiding litigation and must be integrated with clinical audit and other quality assurance activities.

    A healthcare safety programme

    What might a safety programme in health care look like? In industry such programmes target the tasks, teams, and conditions of work rather than the staff.8 Safety needs to be tackled both at the level of the particular clinical process, as in clinical audit, and at interpersonal and organisational levels. Audits need to be supplemented by broader analyses of organisational and system features. When tasks can be clearly specified greater standardisation, clear guidelines, and less reliance on human memory and vigilance are essential. Team and communication failures have been strongly implicated in many accidents and remedial measures can be cheap and straightforward. Systems have also been developed in industry to monitor the conditions of work and the associated organisational factors and decisions that may predispose to risk and unsafe practice.

    Systemic change, at both clinical and organisational levels, involves a commitment to safety and quality at all levels of an organisation, a fact long recognised by proponents of total quality management. The examination of individual incidents in a risk management programme is a powerful way of examining the factors implicated when things go wrong, but thoroughgoing change will require a range of quality and safety techniques embedded in a comprehensive strategy. If the true costs of adverse events are recognised, rather than the comparatively trivial costs of litigation, then resources may be made available to implement the comprehensive strategies needed.

    The responsibility for adverse events rests not always with the unfortunate junior nurse or doctor caught in the crossfire but also with those taking higher level organisational decisions affecting the conditions of work; these might be senior clinicians, hospital management, the purchasing authorities, and even the secretary of state for health. From this perspective litigation can be seen not simply as a threat but as a way of revealing unsafe conditions of practice and putting pressure on those with the authority to implement change. Interesting times would indeed be ahead if lawyers sued those they perceived as truly responsible for adverse events, rather than the staff with immediate responsibility for the care of the injured patient.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.