Anaesthetists' violations of safety guidelines

We read Beatty and Beatty's paper on attitudes to violations with interest (Anaesthesia 2004; 59: 528–40). They correctly state that monitoring violations is very difficult as reporting them might lay the individual open to accusations of negligence even if (as is usually the case) there is no intention to harm the patient. As part of our study into anaesthetic knowledge and expertise [1] we observed anaesthetists at work and have transcribed field notes from over 130 h of observation, mostly in the operating theatre. Our study was not designed primarily to capture adverse incidents and as these are uncommon, it is no surprise that we witnessed none. We did, however, note a considerable number of what we have termed ‘aberrations’ in the process of anaesthetic care. There were 103 in all, including 26 instances of ‘lack of smoothness’ during anaesthesia (coughing, patient movement, low but not dangerous blood pressure, etc.) and 19 instances of ‘procedural difficulty’ (difficulty inserting cannulae, laryngeal mask airways, etc.). The main focus of our observations was the work of the anaesthetist from the time the patient arrived in the anaesthetic room so we cannot comment on pre-operative visiting or pre-anaesthetic equipment checks (two of Beatty and Beatty's scenarios). Our analysis of responses to abnormal monitor readings and alarms (including some silencing of alarms and why this might occur) has already been published [2] and we have made some suggestions for training in this area [3]. However, with reference to what Beatty and Beatty term violations, we observed 21 instances of what might be termed ‘variances from usual process’ (even the choice of name has to be made carefully to avoid normative overtones). We were quite strict in our interpretation – for instance, if the anaesthetist left theatre even for a few seconds (for instance to fetch an antibiotic from across the corridor) this was noted. One of the commonest variances was the disconnection and subsequent re-use of a partially used bag of intravenous fluid. This is in direct contradiction to a Medical Devices Agency Hazard Notice [4]. However, we did not observe anaesthetists pressurising the reattached infusions (the other predisposing factor that led to the death by air embolism which triggered the hazard notice). We know that many of the anaesthetists we studied were aware of the notice. Why then did they persist in a behaviour that was ‘officially’ proscribed? In this regard, we were particularly interested in Beatty and Beatty's comment that professional norms were the most powerful influence on intention to violate. This is in keeping with our own work on professional expertise – we suggest that defining what is critical depends on being within a specific ‘community of practice’[5]. For instance, there may be subtle differences between ‘accepted’ and ‘acceptable’ practice depending on who is making the judgement. Those within the community of anaesthetic knowledge are likely to have a different view from those outside. Thus, in this case, anaesthetists – who understand the practice context of the error – feel that they are able to re-use fluid safely, without pressurising the bag, and this is accepted within this group of experts. Given that a Hazard Notice now exists for this practice, however, this may make it less acceptable in the health care community at large. One fear in discussing ‘violations’– or the minor aberrations we observed – must be that such behaviour may be misconstrued by those who may not fully understand the context. Finally, we would add that, although such minor aberrations may be accepted by anaesthetists as ‘just part of the furniture’, a recent study by Boëlle and colleagues [6] suggested that many adverse events were preceded by events which, though not critical in themselves, were deemed ‘undesirable’. If there is indeed a link between the two, perhaps more attention should be paid to the many instances where things do not go as smoothly as they might but will apparently remain in control. This is worthy of further study. We greatly welcome Dr Smith and colleagues' letter in response to our paper. It is a useful and interesting contribution to the debate. They raise three points, on which we would like to comment. Firstly, what is the real incidence of violations in anaesthesia? At best, the Theory of Planned Behaviour used in our paper gives an estimate of the intention to violate. There are any numbers of ways that intention may be thwarted in practice, some of which will be associated with the safety procedures in place in well-run departments of anaesthesia. Thus, the incidence level may be very low in practice as the examination of their field notes suggests. However, our study was prompted by observations by the Medical Defence Union of closed cases. There needs to be further field research using a violations focused approach if we are to find out the nature and rate of the true incidence. Secondly, we agree with them that the term ‘violations’ may have off-putting professional overtones. It is a rather aggressive term but it is one that the psychology community uses and which has established scientific definitions, which is why we used it in the paper. We agree that it may not be the most appropriate term in some circumstances. It is a definition that would include some of the variations they note. If a new terminology is to be used within anaesthesia, it should be one that does not encourage a culture in which violations are acceptable. Lastly, we find intriguing their suggestion that minor aberrations, though not critical in themselves, may be indicators of developing paths to critical incidents. It is an idea that would fit well into the sort of cascade effect in accident development demonstrated by Reason [1] and others. P. Beatty and S. BeattyUniversity of Manchester Manchester M13 9PT, UK