Intended for healthcare professionals


Improving patients' safety by gathering information

BMJ 2001; 323 doi: (Published 11 August 2001) Cite this as: BMJ 2001;323:298

Anonymous reporting has an important role

  1. Bill Runciman (wrunciman{at}, president,
  2. Alan Merry, clinical associate professor,
  3. Alexander McCall Smith (mccallsmith{at}, professor of medical law
  1. Australian Patient Safety Foundation, GPO Box 400, Adelaide, South Australia 5001, Australia
  2. Department of Anaesthesia, Green Lane Hospital, Auckland, New Zealand
  3. Faculty of Law, University of Edinburgh, Edinburgh EH8 9YL

    If the current rate of iatrogenic harm in health care is to be reduced there is widespread agreement that we need to identify how and why adverse events occur, and, in particular, how system defects may contribute to their occurrence. This view underlies reports such as An Organisation with a Memory1 in the United Kingdom and similar reports in other countries2 and has led to political commitment to national monitoring systems. As Britain's new National Patient Safety Agency, currently being established, starts to ponder the issue, what are the elements of a successful reporting and monitoring system?

    In spite of this recent recognition of the need for monitoring, disagreement remains about the attributes of the ideal reporting system. There are at least two distinct objectives. One is to identify practitioners or units whose performance is substandard and processes, infrastructure, or equipment that are manifestly inadequate or dangerous and to deal with these particular problems at a local level. To do this systematically will require the collection of data and numerators and denominators. Units and individuals will need to be identified. Information such as batch numbers, manufacturers' names, and models will be needed. The requirements of natural justice dictate that this process be objective and properly validated. With careful attention to safeguards, selected information of this sort could and should be made publicly available. However, this approach identifies the problem in only a relatively small proportion of the many cases where things go wrong in health care.

    There is also a wide range of events for which the frank reporting of all the relevant details may damage the professional prospects and working relationships of those involved. Reluctance to report in these circumstances persists in spite of moves towards greater openness in the workplace. A doctor is unlikely to report if he or she knows that this information, associated with his or her name, will be retained somewhere in a file or databank. Assurances of confidentiality may not be enough; those who know how diluted the principle of medical confidentiality has become might be forgiven for questioning whether highly incriminating information passed on to hospital authorities will go no further. If reporting is anonymous, however, such a doctor will have nothing to lose and might be more motivated to report the problem to prevent its recurrence.

    A fundamental principle is that while rare problems are not foreseeable and may never manifest themselves again in exactly the same way, the contributing factors behind them often are foreseeable and can be systematically identified and addressed. This is the second objective of incident reporting. The most important goal here is to gather the necessary information about where and why things are likely to go wrong rather than to identify the people involved. This recognises that most avoidable problems in complex organisations relate more to faults in the system than to faults in the individual.3 There is a strong case in these circumstances for anonymous reporting, in which individuals are neither required to identify themselves nor allowed to identify others. This opens the way for opinions about human performance to be expressed without fear of legal or professional consequences.

    A frequent objection to this approach is that those involved may have misinterpreted facts or failed to identify important contributing factors and that, without independent follow up, this information would be lost. Australian experience with many thousands of anonymous reports suggests that this is only occasionally a problem and is more than compensated for by a rich mass of “human factors” information that would not otherwise be recorded.4 There is also some middle ground: many who file anonymous reports are quite happy also to own up to them at quality assurance meetings with peers, allowing both discussion of possible alternative interpretations and independent validation of the facts.

    A widely held misconception is that allowing anonymity confers a special privilege on doctors. A system of anonymous incident reporting does not replace any existing legal or disciplinary processes that may follow harm to a patient. These processes are an important part of responding to the needs of those who have been injured, but they have a minor role in improving patient safety overall. Anonymous reporting should be seen as adding a safety component to existing legal and complaints procedures. It gives doctors nothing that they do not already have; it takes nothing away from the rights that patients currently enjoy; and it provides an additional, powerful, and currently unavailable tool for making health care safer.

    No system will work adequately unless those from whom the reports are needed are fully engaged and their legitimate concerns addressed. We believe that, as a minimum, a trial of reporting systems which include an anonymous option should be undertaken within the National Patient Safety Agency of the NHS.


    • AM has financial interests in Safer Sleep Limited, which promotes a system for reducing errors in anaesthesia.


    1. 1.
    2. 2.
    3. 3.
    4. 4.