Discussion

The NHSLA database is compiled primarily to monitor claims and not for clinical purposes. As a result it may contain inaccuracies. The nature of the incident is usually clear, but the cause of the incident, the injury and the specialty classification may occasionally be inaccurate.

The author also approached a number of other regulatory bodies and organisations that hold information on medical errors. However, due to problems with the classification of these data and concerns about patient confidentiality, very little of this information was accessible. Thus, while there is no doubt that there is a great deal of information on errors, in practice it is not readily available. Thus in spite of its limitations, the NHSLA database is the single best available source of information on paediatric errors in England.

Only a minority of complaints develop into claims handled by the NHSLA and of those the majority are settled out of court or abandoned by the claimant. Court cases can be very expensive and time consuming and can have an unpredictable outcome. Only 4% of NHSLA claims actually reach court and in this analysis it has been assumed that the payment of damages by the NHSLA to the claimant indicates that an error was made, although this may not always the case.

The commonest incidents related to a delay or failure in diagnosing infections, particularly septicaemia and meningitis. Medication errors were also common and will hopefully lessen with the development of electronic prescribing that can check drug doses and drug interactions. Although many of the cases that led to litigation were predictable, others such as delays in diagnosing anorectal malformations, misdiagnosis of epilepsy and gastrostomy related errors were less expected. Psychological and psychiatric effects on a parent or child as a result of an error also led to litigation. Technical problems with medical devices such as cold lights and potentially avoidable nursing issues such as pressure sores call for better training in the use of medical devices and more vigilance in the monitoring of pressure sores.

In Carroll and Buddenbaum's data from the Physician Insurers Association of America, the five most common events that led to claims were brain damage to infants, meningitis, ‘routine infant or child health check’, neonatal respiratory problems and appendicitis. Nearly half (43%) of the incidents occurred in an ‘office paediatrics’ setting, out of hospital.4 In Najaf-Zadeh et al's survey of paediatric malpractice in France, the commonest reasons for claims were meningitis, dehydration, malignancy, pneumonia and appendicitis. However, the French data excluded infants under 1 month of age and dealt mainly with adverse incidents in general practice.3

The causes of the errors were very much as expected, with a delay or failure in diagnosis and treatment being by far the most common. This is in keeping with Carroll and Buddenbaum's observations and Najaf-Zadeh et al's data in which an incorrect diagnosis was the commonest cause of an error in 32% and 47% of cases, respectively.3 4 The NHSLA data do not provide information on the underlying reasons for these errors, which may be due to individual mistakes and/or process errors. Inadequate nursing care was another cause of errors and emphasises the importance of team working and efficient communication between doctors and nurses. Close observation of infusions, especially those in which extravasation can lead to tissue damage, is also necessary to avoid another common cause of litigation. Failure to consider the possible complications of a disease or its treatment (eg, a ventriculoperitoneal shunt infection) was another source of errors.

Most medical errors result in no harm. However, in this study of the small number of adverse events that led to successful claims against the NHS, the commonest result of an error was death. Unnecessary pain was the second commonest result and may reflect on the adequacy of pain management. Psychiatric and psychological morbidity in the parent(s) and/or child as well as being a cause of litigation were also results of an adverse incident. Additional surgery was also a common result of an adverse event. The fourth commonest injury was brain damage.

Successful litigious complaints were commonest in paediatric medicine (64%) and second commonest in neonatology (21%). Obstetric claims which may relate to cases of cerebral palsy and developmental delay in infants and children were not part of the NHSLA paediatric database.

The physical and psychological consequences to the child and their family of these adverse incidents can be very high. The financial cost to the NHS, especially in this age of limited funding and cuts in public services, is considerable. The largest claims resulted from cases that had led to brain damage. In some of these cases there are continuing intermittent payments which are planned for the duration of the claimant's life.

Many hospitals now have a clinical risk department. However, feedback and the dissemination of lessons from errors is often limited. Regrettably, the same error may have to recur several times before effective action is taken.5 Despite the encouragement of the former Chief Medical Officer, Professor Sir Liam Donaldson, for doctors and hospitals to learn from their mistakes, this is only happening to a limited extent. Furthermore, the more open culture in which errors and service failures can be reported and discussed has still not materialised.

Publications such as Why Children Die6 from the Centre for Maternal and Child Enquiries provide useful educational information. There is also a surgical online site, CORESS,7 sponsored by the Association of Surgeons of Great Britain and Ireland, which enables the confidential reporting of surgical errors with the aim of improving patient safety and of providing an important educational resource. A similar paediatric site would be very useful.

It is regrettable that institutions tasked to deal with errors often do not classify their data in a manner that is conducive to its analysis.2 Doing so would help to produce a useful national picture of common errors in paediatrics and other medical specialties.

It is hoped that the information provided about errors in this article will inform the training of junior doctors and nurses and in so doing will help avoid the recurrence of errors and improve the safety of medical and nursing practice.