Organisation and planning of anaesthesia for emergency surgery

Summary

Patients presenting for emergency surgery represent a category at high risk of complications, with substantial morbidity and mortality, whose management may be extremely challenging. In this first of two articles we consider the identification and evaluation of high risk emergency patients, the provision of critical care support, the management of sepsis, common postoperative complications and in-theatre death.

Emergency surgical conditions remain a common reason for hospital admission, and a large proportion of National Health Service (NHS) resources are directed towards managing these patients. In 2010–2011 there were 588 890 emergency episodes of care recorded for general surgery alone in English NHS hospitals [1]. At present, 1% of NHS trusts have a higher rate of deaths after surgery than expected [2].

Emergency surgical cases require triage to allow appropriate timing of surgery and appropriate and efficient allocation of resources and staff expertise. In the UK, the accepted system for surgical scheduling is the National Confidential Enquiry into Perioperative Deaths (NCEPOD; Table 1) [3].

Table 1. National Confidential Enquiry into Perioperative Deaths (NCEPOD) Revised classification of intervention 2004 [3].

NCEPOD Category	Definition	Example(s)
*The classification should reflect the patient’s physiological condition and surgical diagnosis.
Immediate (1A) Life threatening (1B) Limb or organ threatening	Immediate life, limb or organ-saving intervention – resuscitation simultaneous with intervention. Ideally surgery and anaesthesia within minutes of decision to operate	Ruptured abdominal aortic aneurysm Necrotising fasciitis in a limb Laparotomy for perforated viscus*
Urgent (2)	Intervention for acute onset or clinical deterioration of potentially life-threatening conditions, for those conditions that may threaten the survival of limb or organ, for fixation of many fractures and for relief of pain or other distressing symptoms. Normally within hours of decision to operate	Fixation of fractured neck of femur Laparotomy for perforated viscus*
Expedited (3)	Patient requiring early treatment where the condition is not an immediate threat to life, limb or organ survival. Normally within days of decision to operate	Repair of tendon or nerve injuries
Elective (4)	Intervention planned or booked in advance of routine admission to hospital. Timing to suit patient, hospital and staff	Total hip replacement

The NCEPOD sub-classifies category 1 ‘immediate’ surgical cases into those that are life saving (1A), and those that are organ or limb saving (1B) [3]. It was an initially controversial recommendation by NCEPOD to reduce ‘out of hours’ operating since operating at these times frequently reflected inadequate provision of emergency operating time during ‘office hours’ [4]. Implicit within this recommendation was the adequate provision of daytime emergency lists in hospitals [5]. Recently, NCEPOD identified that 158/218 (72.5%) of NHS hospitals were able to offer emergency theatre and 183/220 (83.2%) had provision in evening and night-time [6]. Furthermore, 98.6% of UK NHS hospitals have a post-anaesthetic recovery area, but only 66.9% have it all day every day, and provision for ventilatory support and ongoing management is only available in 82.8% of these units [6].

This review will focus on adult general surgical patients requiring immediate (NCEPOD Category 1) intervention; we will not consider emergency anaesthesia in more specialist areas e.g. obstetric, cardiac or neurosurgery. However, many of these principles and recommendations can be applied to those requiring ‘urgent’ surgery.

Approximately 3 million surgical operations a year are conducted in the UK [7]. Of these, 170 000 are in ‘high risk’ patients [8, 9], with an overall hospital mortality rate of 0.8–1.0% [7]. However, there is clear variation in morbidity and mortality even outside the emergency setting [10, 11]. The ‘high risk’ group may be defined by a predicted mortality > 5% and predictably consists mainly of older patients, with comorbidities, undergoing major surgery [8, 12, 13]. Over 80% of surgical deaths occur in this subpopulation of surgical patients, representing only 12.5% of procedures [13].

Elderly surgical patients continue to be a group of growing concern, with an increasingly ageing population and a tendency for this age group to present as emergency cases. In a recent NCEPOD report, 83.4% of surgical admissions were as emergencies for those over 80 years old [14]. Previous studies have suggested that the interval from onset of symptoms to admission, and conditions that permit only palliative surgery (e.g. intestinal bypass procedures) predict mortality, but that increasing age in the elderly (over 70 vs over 80) is not associated with mortality, morbidity or length of hospital stay [15].

Of equal concern is the almost certain under-reporting of critical incidents that is highly likely to reveal significant ‘concealed’ morbidity; the National Patient Safety Agency (NPSA) Reporting and Learning System received only 155,258 reports of incidents relating to surgical specialties, and only identified 1172 cases of ‘severe harm’ or death during 2009 [16].

Emergency work accounted for almost 16% of the surgical workload in NHS hospitals in 2001/2012 [4], and may be even higher than this in some specialties e.g. neurosurgery [17]. In 2001/2002 there were 20 130 reported surgical deaths, with most occurring within 30 days of surgery [4]. This is now estimated at 25 000 deaths per year [8].

For matched patients the risk attributed to emergency surgery is greater than a comparable elective procedure [7]. The first report of the UK Emergency Laparotomy Network (ELN) [18] demonstrated a non-risk adjusted mortality of 14.9%, rising to 24.4% in those aged 80 or over [19].

Evolution of the management of emergency surgical patients

Organisational and medical interventions for improving morbidity and mortality around elective major surgery have been identified [7, 20] but there are fundamental challenges to delivering many elements of these in the emergency surgical population including time, availability and applicability of investigations (e.g. exercise stress testing) and the benefit of some interventions e.g. goal-directed therapy for which benefits in this population have not been consistently demonstrated [21, 22].

In the absence of robust evidence, empirical calls have been made for improved risk stratification and care bundles, already established in managing conditions such as sepsis [23] or mechanical ventilation [24, 25] for emergency surgical patients [26].

The Improving Surgical Outcomes Group (ISOG) recommendations require that more patients are identified and admitted to critical care facilities in a planned fashion, as outcomes following admission ‘in a crisis’ following deterioration are demonstrably worse, see below [7]. The recent ‘The Higher Risk General Surgical Patient’ report identifies five steps in improving patient care [8]:

• 1
   Rapid identification.
• 2
   Adequate resuscitation.
• 3
   Investigation to identify the underlying problem.
• 4
   Rapid definitive treatment of that problem.
• 5
   Appropriate critical care provision to prevent further complications.

The National Institute for Health and Clinical Excellence (NICE) and other sources recommend ‘track and trigger’ systems for adult patients in hospitals i.e. the observation and weighted scoring of physiological values to initiate appropriate responses e.g. Modified Early Warning Score (MEWS) (Table 2) [8, 27, 28]. Such systems may detect complications early [29], and reduce mortality [30], although the impact on outcomes of ‘track and trigger’ systems remains controversial [31]. A recent Royal College of Surgeons recommendation suggests that cases with a predicted mortality >10% should be managed by a consultant anaesthetist and surgeon and that postoperative admission to a critical care facility should be planned; currently this occurs in only approximately one third of cases [8].

Table 2. An example of a Modified Early Warning Score (MEWS) as applied to a surgical population [28].

Score	3	2	1	0	1	2	3
Scores for each variable are recorded at the time that observations are taken. If the total is 4 or more then the ward doctor is informed.
Respiratory rate; min^−1		≤ 8		9–14	15–20	21–29	> 29
Heart rate; min^−1		≤ 40	41–50	51–100	101–110	111–129	> 129
Systolic BP; mmHg	≤ 70	71–80	81–100	101–199		≥ 200
Urine output; ml.kg^−1.h^−1	Nil	< 0.5
Temperature; °C		≤ 35	35.1–36	36.1–38	38.1–38.5	≥ 38.6
Neurological status				Alert	Reacting to voice	Reacting to pain	Unresponsive

The prediction of outcomes for an individual patient is challenging and almost all scores provide a population-based estimate attempting to match key features e.g. age and comorbidities. Perhaps as expected, 30-day outcomes for patients prospectively identified as high risk by the supervising anaesthetist vary dramatically with surgical urgency. High-risk elective cases have 1.5% mortality at 30 days, whereas non-elective cases have a mortality of 18.1% [6].

Changing the paradigm: increasing peri-operative planned admission to critical care

There is a recurrent message that emergency patients have high rates of morbidity and mortality [6, 7]. Furthermore, amongst such a high-risk category of patients their outcomes are demonstrably better if critical care admission is planned rather than unplanned [7] as suboptimal care of severely ill patients before admission to intensive care is common [32].

Central to achieving this in the UK is the correct identification of predicted mortality and morbidity (see below) through accurate assessment and scoring systems, integrated into ‘track and trigger’ alongside critical care outreach services, and adequate provision of critical care facilities, stratified into different levels, which is not synonymous with ‘intensive care’.

The UK Comprehensive Critical Care report defined levels of critical care in the UK (Table 3) [33]; these are recognised by the Intensive Care Society [34].

Table 3. Levels of critical care in the UK [33].

Level of care	Definition – UK Comprehensive Critical Care report [33]
Level 0	Patients whose needs can be met through normal ward care in an acute hospital
Level 1	Patients at risk of their condition deteriorating, or those recently relocated from higher levels of care, whose needs can be met on an acute ward with additional advice and support from the critical care team
Level 2	Patients requiring more detailed observation or intervention including support for a single failing organ system or postoperative care and those ‘stepping down’ from higher levels of care
Level 3	Patients requiring advanced respiratory support alone or basic respiratory support together with support of at least two organ systems. This level includes all complex patients requiring support for multi-organ failure

However, the UK continues to lack capacity for emergency surgical admissions to critical care facilities with a consistent and paltry critical care bed provision below 2% of acute hospital beds as compared with 20% in the US [6]. Local models and adjustments, such as overnight ventilation facilities, reflect ingenious solutions to resource challenges and at the time of writing it is difficult to envisage the required level of funding for critical care provision to match other developed nations.

Everything in good time: avoidance of delays

It is a principle of care that all possible delays are avoided so that definitive management (i.e. the surgical procedure) can be provided [35]. Within the steps above, perhaps ‘adequate resuscitation’ and ‘investigation to identify the underlying problem’ present the greatest potential for delay.

Individual patients should be resuscitated to end-points defined by their presentation, comorbidities and likely recovery following emergency surgery. Often appropriate endpoints are not the same as those that pertain in good health; e.g. a perfusion pressure that is ‘adequate’ rather than the same as in the pre-morbid state may be appropriate [35]. Short-term goals include restoration of organ and tissue microperfusion and global indices of tissue oxygen delivery. The specific aims for resuscitation will vary with presenting features and likely diagnosis. For example, in acute haemorrhagic shock the consensus has shifted from ‘fluid resuscitation and review response to guide management,’ to the provision of early blood product transfusion, warming and definitive anatomical control of bleeding via surgery or interventional radiology, with a focus on early arrest of haemorrhage. Transient hypotension and reduced organ perfusion may be tolerated if definitive control is imminent and some animal and human studies have shown that aggressive fluid resuscitation and restoration of ‘normal’ blood pressure are associated with worse outcomes including hypothermia, coagulopathy of trauma and death [36, 37].

Appropriate investigations should also be emphasised. Close liaison between the senior anaesthetist and surgeon is the best way to address both resuscitation and investigation around definitive management. It is essential to be mindful of the hazards of transferring patients for imaging or procedures [38, 39], and the additional risks of remote site anaesthesia e.g. CT scanner [40], alongside the diagnostic performance and ability of investigations to influence management. A significant evolution in imaging and interventional radiology has meant some blurring of boundaries. Many conditions are now undergoing evaluation for diagnosis and treatment at the same time; e.g. angiographic control of pelvic bleeding [41], and the emergency placement of endovascular stents, in selected cases, following abdominal aortic aneurysm rupture [42].

Identifying the ‘high-risk’ emergency patient

There is ample evidence that clinicians have traditionally failed to recognise the severity of illness of emergency surgical patients. Efforts to improve this have largely moved towards physiology-based assessment and consideration of ‘organ systems’ and their compound failures, expressed by scoring systems, and this apparently simple process has been reviewed elsewhere [43].

There are many such scoring systems in use, some with the aim of identifying population risk, such as the ASA physical status score [44], a familiar scoring system with no adjustment for age, sex, obesity or pregnancy.

However, the ASA grading system is poorly predictive of individual patient outcome with regard to complications [45], takes no account of surgical procedure or physiological variables, and remains a relatively subjective assessment. Patients classed as ASA grade 5 are defined as those not expected to survive 24 h (with or without surgery). They present a significant challenge in terms of peri-operative diagnosis and appropriate management (e.g. surgery vs palliative care), coupled with major comorbidity and failing organ systems at the point of presentation. Yet up to 40% survive to discharge [46]. The accuracy and reliability of ASA grade 5 as an outcome predictor in emergency cases is thus questionable and alternative scoring systems may be preferable.

The Physiological and Operative Severity Score for the enumeration of Mortality and Morbidity

The Physiological and Operative Severity Score for the enumeration of Mortality and Morbidity (POSSUM) was originally devised as a retrospective tool for surgical audit and not a prospective outcome predictor [47]. The score includes physiological variables alongside operative and postoperative variables for the prediction of 30-day morbidity and mortality rates, e.g. a 70-year-old patient undergoing elective complex major surgery may have a population mortality of 1.7% and morbidity 38.5%; in the event of bowel perforation and emergency surgery these increase to 13.3 and 89.9%, respectively [48].

In recognition of the fact that POSSUM tends to over-estimate adverse events in the lower risk population, the Portsmouth-POSSUM (P-POSSUM) system was developed and validated in a UK general surgical population [49]. The elderly typically have increased comorbidities and the adjusted E-POSSUM score shows good performance in major colorectal surgery, although not specifically in the emergency setting [50].

The Acute Physiology and Chronic Health Evaluation II score

The Acute Physiology and Chronic Health Evaluation II (APACHE II) score includes variables from chronic health and acute physiological change due to illness severity, incorporating biochemical analysis [51]. It is designed for critical care use and records the worst values in the first 24 h of admission. Risk predictions for surgical patients remain better with APACHE II than with its later inception, APACHE III [52]. It has also been robustly calibrated and validated against a UK critical care population to calculate standardised mortality rates in that setting [53]. However, when compared with POSSUM, APACHE II is less able to predict mortality or morbidity in patients admitted to a high dependency unit (critical care) following general surgery [54].

Intensive Care National Audit and Research Centre data

Not all patients undergoing emergency surgery will require critical care. Currently, 14 000 patients per year are admitted to intensive care following emergency general surgery [8]. The Intensive Care National Audit and Research Centre (ICNARC) has validated a risk prediction model for adult critical care units in the UK that performs with better discrimination and overall fit than other risk prediction models, even following their recalibration [55].

Biochemical markers of risk

A number of biochemical parameters have been considered outside composite scores and appear to perform well as risk predictors e.g. serum lactate, base excess, albumin, troponin, high sensitivity C-reactive protein, procalcitonin. Most are best regarded as research tools with the possible exception of lactate (see below). It is probable that biomarkers will play a role in stratification but this will probably involve a panel of appropriate markers, dynamic profiling and a requirement for near-patient and real-time testing. The possibility of combining such profiles with genotyping is promising, with enormous potential, but this is currently not clinically applicable.

Serum lactate and the emergency surgical patient

The management of the shocked patient is considered in our accompanying article. However, hyperlactaemia and metabolic acidosis have both been used as an initial biochemical guide to resuscitation in the critically ill patient. Early experimental animal studies, alongside human observations, had correlated arterial blood lactate content inversely with survival [56]. Smaller studies suggested the benefit of serial lactate measurement during fluid resuscitation in shock. Repeated lactate determinations represent more reliable prognostic index than an initial value alone [57]. With increasing focus on resuscitation for surgical patients [35], and readily available near-patient testing, lactate measurement has become widespread.

The pathophysiological significance of elevated lactate has been reviewed previously [58, 59]. Typically, lactate serves as a highly sensitive marker for adverse outcomes, while its specificity for cell ischaemia per se is less good.

There is an extensive literature considering lactate in the septic population. The Surviving Sepsis Campaign recommends lactate measurement and use as a trigger for additional vasoactive support [60], and this was re-enforced in recent UK guidance [8]. However, a near-patient lactate greater than 4 mmol.l⁻¹ had with a poor sensitivity (49%) and moderate specificity (74%) for predicting 28-day mortality in emergency department patients with sepsis [61].

Serum lactate may also function as a therapeutic target [62]. This study demonstrated that therapy to reduce lactate by 20% every 2 h when elevated > 3.0 mmol.l⁻¹ significantly reduced mortality. The control group in this study had similar lactate levels throughout resuscitation, suggesting that the early resuscitation trigger that lactate provides is perhaps more important than its serial monitoring. Alternative sources have suggested that lactate clearance is an independent predictor of outcome when added to the Surviving Sepsis Campaign resuscitation bundle [63].

Sepsis – improving survival in emergency surgical patients

Severe sepsis is characterised by organ dysfunction as a consequence of infection and the systemic inflammatory response syndrome [64, 65].

Although defined by these characteristics, sepsis syndromes can present clinically differing pictures according to micro-organism, toxin production, host immune response and timeframe of infection. The consideration of predisposition, insult/infection, response and organ dysfunction (PIRO) staging has been suggested as a path for future investigation and risk stratification in this heterogeneous patient group [65, 66]. The exact prevalence of sepsis in the emergency surgical population will vary by institution, but European data demonstrated that surgical patients were the most common category admitted to intensive care with sepsis and 47% suffered pneumonia [67].

Anaesthetists play a central team role in the management of patients with severe sepsis, not least those that require emergency surgery or organ support. The anaesthetic management of patients with severe sepsis has been recently reviewed [68]. Four steps in managing sepsis were identified by a recent surgical report [8], and these largely echo the recommendations and bundles of care of the Surviving Sepsis Campaign [23, 60]:

• 1
   Measure serum lactate.
• 2
   Take blood cultures (preferably before antibiotics).
• 3
   Administer broad-spectrum antimicrobials within 1 h.
• 4
   Treat hypotension, hypovolaemia or elevated lactate with appropriate intravenous fluids.

A perhaps neglected area of impact by anaesthetists is the administration of antimicrobial agents and hence a good working knowledge of appropriate pharmacology and local flora and resistance patterns is essential. Timely sampling of infected blood and tissues and administration of adequate intravenous antimicrobial therapy should be performed as early as possible; some workers demonstrated an excess mortality of 8% per hour delay in antibiotic administration after onset of hypotension [69].

The treatment of sepsis may employ different regimens to more typical prophylaxis seen in the elective setting.

Source control measures include drainage of infected fluids (combined surgical and/or percutaneous approaches), debridement of infected tissues, removal of infected prostheses or foreign bodies and correction of anatomical abnormalities leading to recurrently infected sites [68, 70]. It remains a grade 1C recommendation of the Surviving Sepsis Campaign, indicating that its desirable effects clearly outweigh its undesirable effects, but that the quality of evidence to support it is low [60]. Anaesthetists should have a working knowledge of the sepsis resuscitation and management bundles [23]. However, the evidence base in sepsis management is evolving rapidly including:

• 1
   The role of hydroxyethyl starch colloid solutions is controversial with some workers demonstrating harm and renal impairment [71, 72]. The Crystalloid vs Hydroxyethyl Starch Trial (CHEST) results should be available in late 2012 (J. Myburgh, personal communication, 22/08/2012) [73].
• 2
   Tight glycaemic control has been undermined by a series of negative studies, [71, 74, 75] despite the original positive paper in a surgical ICU population [76].
• 3
   The role for steroids remains controversial after the Corticosteroid Therapy of Septic Shock (CORTICUS) trial demonstrated reduced duration of catecholamine infusions but higher rates of secondary infections [77].
• 4
   The lack of efficacy of activated protein C led to its controversial withdrawal in 2011 [78, 79].

Postoperative complications

Emergency surgical patients have a high incidence of postoperative complications [17]. Commonly encountered major complications include respiratory failure, hypotension and shock, myocardial ischaemia, delirium, acute kidney injury (AKI) and oliguria, pain and nausea/vomiting or ileus. Direct surgical complications e.g. intra-abdominal hypertension or anastomotic failure may exert compounding effects. Furthermore, complications related to ongoing care or condition e.g. venous thromboembolism, catheter-related bloodstream and other nosocomial infections do occur [80]. The development of sepsis syndromes following emergency surgery carries significant morbidity and mortality. Most of these problems are multifactorial and relate to pre-morbid state, surgical procedure and initial and ongoing resuscitative measures.

Acute kidney injury is common following emergency surgery, often being established at the time of admission to hospital. It is associated with higher in-hospital mortality after major surgery with an odds ratio for death of 1.88, higher than chronic renal disease, respiratory or cardiac failure and/or non-metastatic cancer [81]. Although AKI is more prevalent in medical patients who are critically ill, surgical patients admitted postoperatively to intensive care units have a high incidence of AKI; 34% of acute renal failure may be related to major surgery [82]. Other common contributing factors to developing AKI include shock and/or sepsis, hypovolaemia, trauma, nephrotoxic medications and radioiodine contrast [83]. These often feature prominently in emergency surgical patients’ natural history. The major complications of AKI remain acidosis, hyperkalaemia, sepsis, oedema and respiratory failure [84].

Respiratory failure is common after emergency anaesthesia, often requiring a period of mechanical ventilation in the intensive care unit. Common causes include pneumonia, atelectasis and mucous plugging, pulmonary oedema, pain, diaphragmatic splinting, residual effects of sedation, analgesia or neuromuscular blockade and pre-existing poor lung function as a consequence of acute or chronic respiratory disease.

For those undergoing prolonged ventilation, ventilator-associated pneumonia (VAP) occurs at approximately 2% per day. Ventilator-associated pneumonia prolongs ventilation time, length of ICU stay and length of overall hospital stay [85], and appears likely to increase mortality [86]. Measures to reduce VAP incidence in those requiring prolonged ventilation include implementation of the ‘ventilator care bundle’ [24, 25].

Death in theatre or the early postoperative period

Death in theatre is a rare occurrence, with the majority of deaths occurring expectedly in high-risk patients undergoing emergency surgery [87]. In England, Wales and Northern Ireland, there are approximately 2000 deaths on the table each year, representing 11% of all deaths within 30 days of surgery [88].

Deaths attributable to anaesthesia are more rare, estimated at 0.5–0.8 per 100 000 anaesthetics in the UK in 1983 [89], and 0.28 per 100 000 in Australia in 2008 [90]. Interestingly, the majority of anaesthesia-related deaths in this recent report occurred in those over 60, undergoing orthopaedic surgery.

Anaesthetic complications, although also rare, occur with a greater frequency than anaesthesia-related deaths, and may be just as harrowing for those involved. Major complications include airway disasters and significant aspiration [91], anaphylaxis [92], neurological harm [93] and unanticipated ICU admission.

Death in theatre is clearly a tragedy for the patient and their relations, but is also difficult for staff. Published guidelines are available to follow after peri-operative death or catastrophe in theatres [87, 94, 95]. Recommendations include that (notwithstanding the overall balance of risks) the surgical list (including emergencies) should stop. Operating lists should not resume until a different theatre, with a different anaesthetic team (with satisfactorily checked anaesthetic machine and equipment), surgical and theatre team becomes available. This is for the benefit of staff, as well as patients awaiting surgery.

Dead or dying patients may be transferred to a quieter and more dignified area for family and next of kin to be with them. This may be the recovery area, a nearby ward, or occasionally within the theatre, varying with situation and local facilities. Good control of symptoms and compassionate management of the process of death is paramount.

A questionnaire of 251 anaesthetists’ attitudes to intra-operative death demonstrated that 92% had experienced an intra-operative death, with the majority expected (60%) and non-preventable (77%), occurring mainly during emergency surgery (80%) [96]. The majority of respondents (71%) agreed that it was reasonable for medical staff not to take part in operations for 24 h after an intra-operative death; however, few (25%) thought this was practicable. A survey of cardiac surgeons found that intra-operative death adversely affects the morbidity (length of ICU stay and length of hospital stay) in patients operated on by the same surgeon in the subsequent 48 h. This effect was more pronounced following an emergency or high-risk surgical death [97].

In the UK, all such deaths must be referred to the coroner. All equipment should be left in-situ including airway devices, intravenous lines, drains, sutures, catheters etc. pending postmortem examination. This is primarily to preserve evidence and give the coroner’s pathologist the best opportunity for independent inspection and assessment of the case [98].

Anaesthetic equipment and medications administered (retained ampoules) should be checked by a senior anaesthetic colleague and recorded. Faulty equipment should be left untouched and, if possible, sealed and stored in a locked room. It can then be examined on behalf of the coroner and/or the Health and Safety Executive.

The coroner will consider cases of death where the cause may not have been natural and has a duty to investigate these. If an inquest is opened, anaesthetic staff will typically be requested to write a formal statement of events. Medical defence organisations and hospital legal services can both provide assistance [99, 100].

The death should be confirmed by external examination, supported by monitoring in-situ [101], as soon as feasible whilst in theatre and ‘care after death’ (formerly ‘last offices’) performed. Viewing of the body should be arranged if required, ideally outside of the theatre environment. Pacemakers, implantable defibrillators and radioactive implants should be identified so that they can be removed before the cremation if required.

A formal discussion, reflected in the patient notes, should be conducted with the next of kin at the earliest opportunity. This should involve the most senior surgeon and anaesthetist present during the case, plus a member of nursing staff, ideally familiar to the family of the deceased.

Training in this area of communication is minimal for surgeons and anaesthetists when compared with other specialties with higher rates of expected deaths (e.g. palliative medicine), and many feel uncomfortable in this situation. Excellent advice regarding communication before and following unexpected intra-operative death, is available [102].

A debrief of the theatre team involved is recommended [16], and should improve understanding of events. It may help if the cause of death and/or postmortem findings is discussed with both medical and non-medical staff.

Formal ‘single session psychological debriefing’ as an early intervention after psychological trauma (death in theatre represents psychological trauma for many involved) is ineffective in preventing post-traumatic stress disorder (PTSD) and other psychological sequelae [103]. Some evidence suggests it may be counterproductive [104]. As such it is not recommended by the World Health Organization following conflicts and disasters, and cannot be recommended albeit on a smaller scale following death in theatre [105]. Early psychotherapeutic intervention conducted over multiple sessions lacks sufficient evidence to be routinely recommended [106]. The Association of Anaesthetists of Great Britain and Ireland (AAGBI) recommend that trusts ensure that any affected member of staff has the opportunity to see a trained counsellor within the first 72 h of the event [87]. Each individual also has a responsibility to ensure they are fit to work; stress and ill-health may go unrecognised unless colleagues remain vigilant.

Competing interests

No external funding or competing interests declared.