Statistics from Altmetric.com
The complexity of treatment, procedures, interventions, and workload of modern, inpatient paediatric and neonatal care provides a setting where errors may, potentially, have serious adverse consequences for our patients. For the purpose of this article, an error is defined as clinical performance which deviates from an ideal and, as a result, could (or does) lead to an accident or an iatrogenic incident.1 2 Active errors are those that immediately precede an adverse event and latent errors are factors inherent to a system (for example, heavy workload, inadequate maintenance of equipment, or the prevailing professional culture) that provide the conditions in which an accident is inevitable if given the right set of circumstances. Since all physicians involved with acute or emergency care may be expected to perform practical procedures,3 we need to understand why our patients sometimes suffer as a consequence of a procedure, what mistakes occur, and how we can improve on our performance.
The scale of complications and deficiencies in practical skills
“All doctors, however experienced and conscientious make mistakes”.4 In the acute setting, in paediatric intensive care practice, Stambouly et alprospectively assessed the extent and consequence of human error.5 In their experience, 115 complications occurred during 83 of 1035 consecutive admissions over an 18 month period. In all, 5% to 17% (95% confidence interval (CI)) of these complications were procedure related. Human error was involved in 41 (95% CI, 27% to 45%) of these instances, 21 of which were considered to be of such consequence to the patient that life was threatened or that further therapy, specific to the intensive care unit, was deemed necessary. The extent of these problems is not only limited to the intensive care unit. In the emergency room, management skills of front line paediatricians may also be suboptimal. Among 34 paediatric trainees responsible for after hours emergency care in Adelaide, Australia, Brady and Raftos6 found that, when assessed by a questionnaire and mock clinical resuscitation, the average trainee was deficient in one quarter to one third of the theoretical precepts considered important for acute care. Furthermore, on average, the trainees required two minutes to establish effective bag–valve–mask ventilation in an infant manikin. The situation is probably similar in America. In 45 paediatric residents in a university based training programme, White et al reported that although housestaff (individuals trained in the American Heart Association “Pediatric Advanced Life Support”) were generally able to reach the endpoint of four key resuscitation skills, they less frequently achieved the specific subcomponent of each skill.7 For example, in the assessment of emergency defibrillation, almost 90% of the participants discharged the defibrillator when required, but the median time for successful skill completion was dangerously long, 149 seconds. In this context, the performance of UK trainees is likely to be no different to the above experience in Australia and America.8
Teaching practical procedures and the influence of self efficacy
One response to the previous section is to suggest that greater attention to detail during training is needed. Certainly, there is much to be improved in the way we teach practical skills during on the job training9 10: how to plan ahead; how to take a long procedure, break it up into small bites, and adopt a detailed step by step approach; and how to give supportive feedback and get the trainee to have insight into the best path for improvement. However, other psychological factors may be equally important. Simon and Sullivan11 examined “confidence in performance” of paediatric emergency medicine procedures in a cohort of 117 emergency department physicians who were all required to treat children. The authors rated the physicians on a four point scale of comfort (1, comfortable; 2, moderately comfortable; 3, uncomfortable but would perform in an emergency; 4, uncomfortable and would never perform) for all procedures in which the American Academy of Pediatrics recommended competence for paediatric emergency specialists. More than one quarter of the cohort were uncomfortable (grades 3 and 4) with performing certain life saving procedures, even for tasks such as tracheostomy replacement, chest tube placement, and intraosseous line placement.
Taken together with the previous section, we can conclude that, despite appropriate knowledge and training, in an emergency, resuscitation techniques and practical procedures may fail to be skilfully applied unless the operator also has an adequately strong belief in their capability. This attribute should not be confused with self confidence which is a relatively stable general personality trait and may or may not be founded in reality. Rather, such belief or “self efficacy” may vary within an individual depending on the particular task or situation. Maibach et al have addressed the importance of this issue to training in post-resuscitation procedures.12 For example, for the trainee, it is important to recognise that self efficacy beliefs may influence performance: such behaviour may be apparent from particular practical choices one makes, or even specific tasks avoided. For the trainer or supervisor, it is important to ensure that the trainee experiences success and mastery with the practice and application of practical skills, even if it is vicarious and incomplete at first.
Trainee's response to medical mistakes
One consequence of medical training is being initiated into the experience of either having done something to a patient which had a deleterious consequence or else having witnessed peers do the same. Before considering how clinicians respond to their mistakes, it is important to first acknowledge that within a blame culture there may be personal or institutional reasons why professionals in training might respond in a particular manner to a mistake, especially when there are questions of culpability and responsibility. As a result of these pressures, defensive responses are commonly seen: such responses may not be justified but we do need to recognise them if we are to make changes for the better. In this regard, almost 20 years ago, Mizrahi13 observed that, when such events occurred, junior medical staff used a variety of collectively acquired psychosocial coping mechanisms for defining and defending their various mishaps. For example, denial, which resulted in one of three responses: negation of any thought of error by emphasising that the practice of medicine was an art and so not easily subject to rigorous analysis; repression of the facts; or revision of the event by redefining mistakes as non-mistakes. Alternatively, the process of discounting was sometimes used. In this form of defence, blame was externalised to circumstances beyond the control of the individual physician concerned and directed towards other staff, or the disease process, or even the patient. Last, when a mistake could not be denied or discounted, junior clinicians resorted to distancing techniques.
These responses may, however, have more important and far reaching consequences on career development and practice. In an anonymous questionnaire study, Wu et al reported that 114 of 254 (95% CI, 39% to 51%) house officers in internal medicine said that they had made a serious medical mistake.14Multivariate analysis of data from this group indicated that those who coped with their mistake by accepting responsibility were more likely to make constructive changes to their practice, but to experience more emotional distress. Whereas those who coped by escape–avoidance were more likely to report a defensive change in practice.
Mishaps and analysing the human factor
All mishaps have both a context in which they occur and a chain of events from which they appear to have arisen. By examining these ingredients of an accident and by looking at aspects of the ensuing organisational process, such as communication, stress, and supervision, it should be possible to identify an accident's key elements or potential causes.1 Such an approach can also be applied to the assessment of the human factor in iatrogenic medical events.15 For example, the nature and anatomy of an accident can be categorised according to the scheme shown in table 1. Briefly, in this framework, the occurrence of an unsafe act can be analysed according to whether it was the result of an error or whether it was the result of some deliberate deviation from a regulated code of practice. These events can then be described as being either “unintended” or “deliberate” and represent detrimental acts committed by those at the “sharp end” or front line of practice. Latent factors are the inevitable consequence of some remote decision, made some time previously, or they are the result of fallible organisational systems. These “in built” dormant faults may not necessarily be evident to those working at the sharp end of a particular system. Alternatively, there may be potential promoting conditions within the working environment.
When applying the above structure to the study of a particular procedure related incident, the main purpose—beyond identification and categorisation—is to enable a reasoned approach to understanding why a particular event has occurred and what can be learned from it. For example, this analysis may provide insight into whether the person at the sharp end of an event has committed a rule based mistake—that is, misapplication of a rule which may be perfectly alright in other circumstances. Alternatively, in the event of a violation, categorisation may help to signify personal and institutional features which have produced, promoted, and permitted such an event.16
Practical procedures are an integral part of inpatient paediatric care and physicians must know how to perform these safely and effectively. Near misses and actual mishaps, unfortunately, are inevitable. Aside from issues such as personal recrimination, peer review, patient complaint, and legal action, we must both teach and learn how to deal with iatrogenesis in a constructive manner. In this context, one successful approach has been the “critical incident technique” used, originally, in World War II as an objective method for selecting recruits with the appropriate skills to become successful pilots.17 In this method, factual accounts of incidents that were related to successful outcome or failure were collected and used to measure performance and change training. The emphasis was with the incident and not the apportioning of blame to the individual concerned. In medicine, application of this technique has been applied to the practice of anaesthesia and intensive care18: where, at best, the process results in defining the core skills needed for practice within a certain area, the development and dissemination of new, safety conscious guidelines, and the timely identification and alerting of potential system inadequacies.19 Similarly, in paediatric cardiothoracic surgery, application of a similar technique to a cluster of surgical failures led to retraining20; an exemplary description of failure stimulated professional conduct. Training and maintaining practical expertise among paediatricians should be no less stringent. There is, therefore, a need to adopt approaches which will enable us to audit and investigate our own performance,21 not least because they will also engender good professional habits, a system of self appraisal and, if necessary, insight into when to seek retraining.
The author wishes to acknowledge the helpful comments given by Nandu Thalange, Specialist Registrar in Paediatrics, and Robert Ross Russell, Clinical Director of Paediatrics, Addenbrooke's Hospital, Cambridge.