‘Death on the table’

The intra-operative death of a patient is a mercifully rare occurrence, but one that can have considerable psychological and professional consequences for all members of the theatre team. When a death occurs on the operating table, the anaesthetist, one of whose primary professional tasks is to maintain patient health and safety during the ravages of surgery, may become inclined to hypercritical self-examination and subjected to prolonged, judgmental investigation by their peers. Moreover, as Professor Aitkenhead has noted ‘the focus of training in anaesthesia is concerned with the avoidance of disasters, rather than the management of their aftermath’[1], a statement reflected in the scarcity of anaesthetic literature concerning intra-operative death. Anaesthetists therefore are particularly predisposed to psychological distress should a death occur in theatre, and are unlikely to receive much in the way of emotional or professional support after the event. Peri-operative deaths solely due to anaesthesia or anaesthetic error are extremely uncommon (with an estimated prevalence of 0.5–0.8 : 100 000 anaesthetics [2, 3]). However, the intra-operative death of anaesthetised patients is a more common occurrence, with an estimated prevalence of 1–30 : 100 000 [4]. In the United Kingdom (excepting Scotland), approximately 20 000 deaths per annum occur within 30 days of surgery, 2000 of which occur within 24 h of surgery. About 100 (5%) of these deaths occur intra-operatively, a figure that has remained consistent throughout the last decade [5]. At first sight, this does not appear to be a particularly large number, particularly in relation to the number of general or regional anaesthetics performed annually (roughly 5 000 000). However, in an unreported pilot study of 12 hospitals nationwide that I have conducted, 232 (92%) of 251 anaesthetists stated that they had witnessed the intra-operative death of a patient (18 of the 19 who had not were senior house officers or registrars). Virtually all anaesthetists, then, are likely to experience at least one intra-operative death during their careers. Of course, the fact that anaesthetists are extremely likely to witness a death on the table does not mean that every death will trigger psychological distress in the anaesthetist, or affect his/her professional ability to continue delivering safe anaesthesia. By analogy, a recent questionnaire survey in the British Medical Journal highlighted orthopaedic surgeons' attitudes to intra-operative death [6]. The survey was performed in response to comments made by Sheriff Albert Sheenan, who, enquiring into the intra-operative deaths of two patients on the same operating list, recommended that surgeons should not operate for 24 h after the intra-operative death of an elective surgical patient. Commenting on these findings, Professor Sir Alfred Cuschieri said that ‘a death on the operating table of a patient is a harrowing experience for a surgeon. In my view, the surgeon is emotionally and mentally not in the frame of mind to continue to operate that day’. This reaction was echoed by Professor Arnold Maran, President of the Royal College of Surgeons of Edinburgh: ‘We can understand the pressure that single surgeons are under, but I think there would be a very strong feeling that, when a surgeon loses a patient, he should not continue operating that day’. Only 53% of the surgeons questioned in this survey had witnessed an intra-operative death, but 81% of those had performed further operations within 24 h without subjective detriment to their operating skill (a figure similar to the 77% of anaesthetists in my survey who felt perfectly competent to deliver another anaesthetic within 24 h). Nevertheless, 50% of the surgeons would have liked some time off to reflect on the death. I would suggest, moreover, that a blanket 24 h ban on continued practice after an intra-operative death, as suggested by the Scottish enquiry, is an entirely inappropriate response to this problem, for a number of reasons. The causes of intra-operative death are multifactorial, and often summative. In relating these causes to the magnitude of ascribable psychological stress for the anaesthetist, a spectrum of issues may be identified that might cause psychological distress to a greater or lesser degree, either by themselves or in combination with other factors. Broadly, these issues fall into two categories: anaesthetic factors (e.g. human error, equipment failure, organisational factors, interprofessional relationships) and patient factors (e.g. patient age, pre-operative condition, nature of the proposed surgery, doctor–patient relationship). For example, the same anaesthetist may experience considerably more psychological stress after the unexpected death of a healthy 10-year-old undergoing minor elective surgery, where there is a suggestion of medical error, than by the ‘expected’ death of a multiply-injured road traffic accident victim undergoing emergency surgery to repair a lacerated liver who dies despite exemplary anaesthetic management. The latter example, or similar cases, of ‘not unexpected’ intra-operative death account for the majority of cases, comprising as they do instances of high-risk emergency or urgent surgery on older patients with significant co-morbidities. Indeed, only 12% of deaths reported to NCEPOD in 1998/9 were assessed as ‘not expected’. In contrast, cases similar to the former example might be expected to have a greater psychological impact. Anaesthetists are not automatons: human error is a factor in anaesthetic practice and its consequences, more so than in other medical specialties, tend to be immediate, adverse and conspicuous. A death that occurs on the end of the needle, due to drug error or overdose, directly approximates the anaesthetist's actions to the patient's death, and may understandably elicit overwhelming feelings of guilt in the anaesthetist, ‘killing’ patients being an anathema to doctors. Even if the actions of the anaesthetist are temporally and spatially more divorced from the patient's death − for example, if death results from malignant hyperpyrexia − there may still be a significant measure of guilt: the anaesthetist's actions did not directly cause death, but the treatment that the anaesthetist administered effected the patient's demise. Whether deaths are ‘expected’ or ‘unexpected’ may, of course, be irrelevant in some cases: any anaesthetist may be emotionally affected by any intra-operative death at any time. Our psychological constitution is shaped by a myriad of internal and external influences, and fluctuates with time and according to circumstance. Gautam, writing in the American Medical Association's recently published Handbook of Physician Health, notes a number of factors that may increase the risk of depression and anxiety amongst doctors [7]. These include biological factors (gender, age, family history, lack of sleep, poor eating habits, poor level of fitness and physical illness), psychological factors (perfectionism, sense of responsibility, need for control, self-doubt) and environmental factors (patients' demands, professional, institutional and legal scrutiny, occupational hazards, clinical governance, personal life and work-life balance). To these I would add medical seniority (consultants feeling greater responsibility than juniors), the ethos of anaesthesia (i.e. to maintain patient safety during surgery), work-related exposure to death and critical illness, the service nature of anaesthesia (with reduced involvement in treatment decisions, particularly amongst junior staff), and possibly the personality types attracted to a career in anaesthesia. The juxtaposition of these factors in routine practice may in turn have two consequences: the stressed anaesthetist may be prone to making more fatal errors, or they may be subject to greater psychological stress if an intra-operative death occurs (either by feeling markedly more distressed when a patient dies, or by feeling traumatised by a fatality that would not normally trigger significant stress). The Association of Anaesthetists of Great Britain and Ireland (AAGBI) have previously noted that as many as 30% of anaesthetists feel stressed a lot of the time, and that stress-related behaviour, including alcoholism, drug abuse and suicide, is particularly prevalent amongst anaesthetists, compared with other medical specialities [8]. In addition, Bacon, in an important paper describing the management of anaesthesia-related death, has previously commented that the consequences can be extremely stressful for the anaesthetist [9]. The psychological effects of an intra-operative death, on a background of continuing stress, may act as a precipitant cause for acute psychological or physical disintegration. However, it should be noted that, just as there is likely to be considerable heterogeneity amongst anaesthetists in terms of the amount of stress felt after an intra-operative death, there is also likely to be a substantial variation in their ability to cope with that stress. Throughout our careers, we are exposed to a great number of unpleasant and disturbing events, and are occasionally witness to the very worst of the human condition, yet we do not require counselling or debriefing after every patient encounter. All of us have developed coping strategies to deal with the stresses of our working lives (to the extent that doctors can be perceived as indifferent to the suffering of their patients). Nevertheless, certain circumstances at certain times may exceed our ability to cope with a death on the table, or certain of its consequences (such as the provision of information to the family, criticism by one's peers or the threat of legal action). The upshot of the above is that the entirely empirical 24 h period of work abstinence suggested by the Scottish enquiry is meaningless if its intention is to protect the psychological welfare of the doctor, or the safety of other patients. The performance of most anaesthetists will be unaffected by most intra-operative deaths, and the resource implications of employing secondary theatre teams and cancelling operating lists are likely to carry a significantly greater risk of patient harm than allowing theatre staff to continue surgery. Conversely, if it is suggested that the intra-operative death of a patient causes psychological dysfunction in doctors to a degree that abstinence from practice is merited, then it is quite possible that a more prolonged period of abstinence than 24 h may be required before the resumption of professional duties. How, then, should the profession cope with individuals who are adversely affected by the intra-operative death of a patient? In part, this may be considered an issue of clinical governance − can continued practice really be considered a risk-free venture if the anaesthetist, surgeon or other theatre staff are potentially psychologically distressed by the death of a patient? Seifert has reiterated that all anaesthetic departments should have guidelines on how to deal with catastrophes as part of their risk management strategy, which should include advice on the decommissioning of anaesthetic equipment (until its safety has been formally re-certified), record-keeping, discussion with the patient's relatives, contact with the coroner, disposal of the corpse and the early involvement of both risk managers and legal experts [10]. Implementation of these guidelines after intra-operative death is time-consuming, and will often necessarily involve the cancellation of the remainder of an operating list, providing a period of time that may be sufficient in itself for theatre staff to recover from the emotional effects of the death. In addition, I would suggest that both informal debriefing with other members of the operating team, and sympathetic peer review (in the form of departmental mortality and morbidity or critical incident reporting) will help to reassure the distressed anaesthetist that other colleagues would probably have conducted the anaesthetic in a similar fashion in the circumstances. The less formal approaches suggested above may be sufficient to deal with the majority of anaesthetists affected by a death. Nevertheless, there are likely be a small number of individuals who will continue to feel traumatised after the event. Their anxieties may diminish with time, although it is possible that a few may develop symptoms of persistent anxiety disorders. Such individuals present a problem: doctors are notoriously reluctant to access professional help when coping with psychological stress. It is crucial therefore that colleagues recognise the signs of stress or depression in a colleague in the aftermath of a death, and take steps to help that individual, possibly involving the doctor's general practitioner, the departmental Director of Anaesthesia, or assistance via the AAGBI's Sick Doctor Scheme. This is not a proposition that many doctors feel comfortable with, but bearing in mind our commitment to clinical governance, our duties to colleagues and the fact that doctors respond well to treatment for anxiety and depression, it is surely one that should be seriously considered. Critical incident stress debriefing (CISD), a formal process of crisis intervention that employs cognitive group psychotherapy, has been previously advocated to limit stress occurring after exposure to emotional trauma [11]. Early forms of stress debriefing, termed ‘crisis intervention approaches’, were tailored on an individual basis. ‘Group psychological debriefing’ was developed later, and is used to reduce immediate distress and identify individuals at risk of developing chronic psychological problems (who require referral for further treatment). Three types of group psychological debriefing are used, of which critical incident stress debriefing (the ‘Mitchell model’) is the most commonly employed. CISD involves one or more aspects of a seven-part model (e.g. assessment of the individuals involved, ‘defusion of emotion through ventilation and validation of distress’), and is provided to subjects within 24–72 h of a critical event. CISD has been used for a number of subject groups, including fire service and police personnel, intensive care nurses, bank employees subjected to armed robbery and military personnel. Its efficacy at preventing long-term distress is variable [12], being very effective in some groups, whilst actually seeming to prolong psychological distress in others. When offered to policemen (a group who might reasonably be expected to experience similar levels of psychological distress to anaesthetists), CISD did not appear to achieve any differences in psychological morbidity [13]. Some authors have suggested that efficacy between groups may vary according to who does the crisis intervention, to whom and in what circumstances [14], which seems logical − soldiers, for example, who are exposed to considerable personal danger during critical incidents, might have different psychological stresses to anaesthetists, who are subject to less personal risk, but possibly more affected by personal responsibility for incidents. Currently, there is no professional guidance concerning the management of intra-operative death. There may be a number of reasons for this: such events are rare and are perceived as part of anaesthetic practice, to be dealt with on an ad hoc basis. In addition, there may be a reluctance to accept that any additional guidance is needed to augment current practice, or to accept that individuals may require further psychological support. Moreover, there may be an understandable resistance to implementing quasi-legal directives that propound to be applicable in all instances of intra-operative death, as if such events were homogeneous occurrences. Possibly of most relevance, however, is that there is very little research in this area on which to establish any rational guidance. The authorities that have considered the management of an anaesthetic catastrophe [1, 9, 15], although admirable in their rational approach, have based their advice on personal experience and reflection, rather than hard evidence. It is unknown, for example, to what degree the intra-operative death of a patient affects the anaesthetist concerned, whether their professional abilities are affected as a consequence, whether critical incident stress debriefing may be of any benefit, or whether a proscriptive approach to intra-operative death is really indicated. In the absence of formal guidance, it would seem that the best approach to take after a death occurs is a pragmatic one. This might involve calling a temporary halt to the operating list (providing other patients were not placed at immediate risk) in order that all members of the theatre staff could discuss the event informally. Staff would then be allowed to make individual decisions on whether to carry on operating; there should be in place departmental provisions to draft in additional personnel if they are required. If there was a possibility of a significant medical error having occurred, or if the death was particularly emotive for any reason (e.g. the death of a child, or an intrapartum maternal death), the departmental director should act to relieve affected individuals from their duties, as a matter of good risk management, whilst arranging for the incident to be discussed in more formal detail. In addition, the departmental director should consider both whether CISD might benefit those distressed by the incident, and whether a mentor (for example, a close colleague) should be appointed or advised to keep a friendly eye out for signs of stress, depression or other psychological sequelae, alerting the director to suggest further treatment. We are all likely to experience a death on the operating table at some point in our careers. In the majority of cases, we will be able to cope with the psychological distress that this may cause. However, we cannot predict our reaction to such distress, due to the circumstantial nature in which it arises. It is important that the profession of anaesthesia continues to strive to prevent intra-operative death from occurring, but it is equally important that there be renewed discussion about managing the aftermath of a death, particularly with regard to providing support for all staff who may be psychologically traumatised by such a potentially distressing event.