Accidental awareness under general anaesthesia (AAGA) is a term referring to a state of unintentional inadequate general anaesthesia, resulting in the patient remembering all or part of a procedure. It is almost exclusively associated with techniques involving the use of neuromuscular blocking drugs, when the patient is unable to move or otherwise attract attention to her plight. Awareness is not an all-or-nothing phenomenon and may range from unpleasant ‘dreams’ to the extreme situation where the patient is fully conscious and experiencing pain but unable to move.
Problems and special considerations
Before the early 1970s it was common practice to avoid volatile agents altogether and use a 50:50 mixture of nitrous oxide and oxygen; not surprisingly this led to maternal awareness in 12-26% of cases. The addition of a low concentration of a volatile agent (e.g. isoflurane 1%) reduces this to less than 1%, and current evidence suggests a ~0.15% risk of AAGA for caesarean section under general anaesthesia (compared with ~0.02% for the general population).
The recent 5th National Audit Project (NAP5) has furthered our understanding of AAGA and its associations. It has long been known that AAGA is particularly associated with anaesthesia for caesarean section, and AAGA is thought to have a higher incidence in the obstetric population. In NAP5 this was supported by the finding that obstetric general anaesthetics accounted for an estimated 0.8% of the dataset denominator, but 10% of the cases of AAGA. Numerous factors may account for this, including the perceived need to protect the fetus from the sedative effects of the anaesthetic agent, coupled with an exaggerated fear of the adverse effects of volatile agents upon uterine contractility. Fortunately, pregnancy reduces anaesthetic requirements by as much as 40%, or complaints of awareness would probably be more common.
Most obstetric general anaesthetics are for category-1 caesarean sections, which by their nature involve many risk factors that were identified by NAP5 such as a rapid-sequence induction, use of thiopental and neuromuscular blockade, increased risk of a difficult airway and/or obesity, urgency and out-of-hours, non-consultant-led care. Other factors suggested include the omission of sedative or analgesic premedication, and the physiological changes of pregnancy including tachycardia and increased cardiac output.
The need for adequate fetal oxygenation during caesarean section has led to the use of higher inspired concentrations of oxygen (and therefore less nitrous oxide) than for other surgical procedures, and this also has an impact on the depth of anaesthesia. There is little evidence to support the use of more than 30% oxygen for delivery of the unstressed fetus, but when fetal distress is present, 50% or even 100% oxygen has been advocated. The loss of the anaesthetic contribution of nitrous oxide and the second-gas effect mean that higher concentrations of volatile agent should be used throughout these cases, and the initial concentration should be higher still (overpressure) to drive up the alveolar concentration quickly. With the decreasing use of nitrous oxide generally, and its replacement with air as the ‘carrier gas’ in anaesthetic gas mixtures, the influence of inspired oxygen concentration on the risk of awareness should become less of an issue.
The contribution of opioid drugs to the anaesthetic should not be ignored, and part of the explanation for the high incidence of awareness during obstetric anaesthesia lies in the common practice of withholding these drugs until after the baby is delivered. This also has adverse consequences for cardiovascular stability during tracheal intubation, and there is an increasing tendency to use a modest dose of short-acting drugs such as fentanyl and alfentanil to obviate both of these problems.
Nearly half of those who experience AAGA have been reported to experience long-term psychological morbidity, typified by ‘waking dreams’, difficulty in sleeping, depression, and fear of hospitals and doctors. Full-blown post-traumatic stress disorder may also occur.
In addition to awareness, inadequate anaesthesia may result in release of catecholamines, which further decrease placental perfusion and promote fetal hypoxia.
Management options
It has been recommended by some authorities (including the authors of NAP5) that patients undergoing general anaesthesia for a caesarean section should be warned about the risk of awareness as part of the consent process, but this is not widely practised. It is advisable to warn the woman that she may wake up with a ‘tube in her throat’ so that her memory of this is less likely to be confused with intraoperative awareness.
Some incidents of awareness may be clearly traced back to a technical problem with the anaesthetic apparatus, vaporiser faults being the most common. When checking the anaesthetic machine, correct seating of the chosen vaporiser on its mount and adequate filling should be ensured. The anaesthetist should be familiar with the breathing system and ventilator, and should understand how air or oxygen can be entrained into the system and how the inspired concentration of volatile agent can be lower than that set on the vaporiser (e.g. a circle system). A volatile agent monitor is considered mandatory.
There is no guaranteed ‘sleep’ dose of an induction agent, and the drug must be titrated against the patient’s response, bearing in mind that it will be responsible for maintaining anaesthesia throughout the onset of neuromuscular blockade and tracheal intubation. Factors such as obesity will influence the induction dose that is used. Thiopental is probably still considered the drug of choice for induction (see Chapter 36, General anaesthesia for caesarean section), and the anaesthetist should have a minimum of 5 mg/kg available in the syringe. Despite its shorter duration of action, propofol was associated with less AAGA than thiopental in NAP5, possibly because of greater familiarity with its use and appropriate dosing. A second syringe of induction agent should be accessible in case of prolonged airway manipulation. One particular risk associated with thiopental is that it is easily mistaken for antibiotic, and vice versa. Concern has been raised that current guidelines advocating administration of prophylactic antibiotics before skin incision could lead to the preparation of two very similar-looking syringes at a time of potential stress, increasing the chance of a drug error; it is advisable that antibiotic syringes should not be present during induction.
Volatile agents with a low lipid solubility will achieve alveolar-inspired equilibrium most quickly. Isoflurane is the best of the ‘established’ agents, but desflurane and sevoflurane are commonly used as a result of their more rapid onset times. Concentrations representing at least 0.5 minimum alveolar concentration (MAC) should be used during the procedure, and this should be higher if the inspired nitrous oxide concentration is less than 60% (a total MAC of at least 1.0 should be aimed for; a higher MAC of 1.3 may be necessary if opioids have not been given). The effect of these concentrations upon uterine contractility is minimal, and even at 1.5 MAC the uterus should respond normally to oxytocic drugs. An overpressure of 1.5-2 MAC should be employed in the first 2-3 minutes if a more soluble agent is being used.
The patient should be closely watched for signs of lightening anaesthesia (tears, sweating), and the monitors should be observed frequently for evidence of sympathetic overactivity (tachycardia, hypertension). Some practitioners advocate the use of specific monitors of anaesthetic depth, but these have not yet been widely evaluated in obstetrics and none has so far been shown to be any more effective than simply watching vital signs (Table 60.1). A meticulous record should be kept, which should include vaporiser settings and end-tidal volatile concentrations, if available.
Table 60.1 Methods for monitoring depth of anaesthesia
Clinical signs - PRST score (pressure, rate, sweating, tears)
Isolated forearm technique
Lower oesophageal contractility
Skin resistance
Evoked auditory/somatosensory potentials
Electroencephalogram/cerebral function analysing monitor and derivations thereof (e.g. bispectral index, entropy)
A generous dose of a suitable opioid drug should be given directly after cord clamping, and the volatile agent left on until the skin is being sutured. It is better to wait a few minutes at the end of the operation rather than risk awareness.
All mothers undergoing general anaesthesia for caesarean section should be followed up within 24 hours of delivery and questioned about dreaming or sensation during the operation. The psychological sequelae of awareness can be minimised by a sympathetic approach. Midwives should be alert to the possibility of awareness and ensure early referral to an anaesthetist. A suggested management pathway has been created by the authors of NAP5; this includes a meeting between the patient, anaesthetist involved and senior anaesthetist to explore what occurred, analysis of the event to attribute contributing factors, and formal support at 2 weeks and by a clinical psychologist if necessary.
Some patients will mistake their memory of awake extubation for true intraoperative awareness. This risk can be minimised by careful preoperative explanation, but any markers as to the timing of such memory should be sought in order to reassure the patient if possible. Just because true awareness did not occur does not mean that the patient will not be traumatised.
Key points
• Clinical signs are still the best indicator of awareness.
• A clinically effective concentration of volatile agent, suitably monitored, should be used at all times.
• Patients should be warned about awake extubation.
• Complaints of awareness should be treated seriously and sympathetically.
Further reading
Paech MJ, Scott KL, Clavisi O, et al A prospective study of awareness and recall associated with general anaesthesia for caesarean section. Int J Obstet Anesth 2008; 17: 298-303.
Pandit JJ, Cook TM (eds). Accidental Awareness During General Anaesthesia in the United Kingdom and Ireland: Report and Findings. NAP5: Fifth National Audit Project. London: Royal College of Anaesthetists and the Association of Anaesthetists of Great Britain and Ireland, 2014. www.nationalauditprojects.org.uk/NAP5report (accessed December 2018).
Platt F, Lucas N, Bogod DG. AAGA in obstetric anaesthesia. In Pandit JJ, Cook TM (eds), Accidental Awareness During General Anaesthesia in the United Kingdom and Ireland: Report and Findings. London: RCoA and AAGBI, 2014, pp. 133-43.
Robins K, Lyons G. Intraoperative awareness during general anesthesia for cesarean delivery. Anesth Analg 2009; 109: 886-90.