Inadequacies in the reporting and assessment process and lack of autopsy examinations and reports considerably impaired the review process. Nevertheless, more information is available than in the previous report and supports the conclusion made then that there is a considerable problem with the level of training of doctors practising obstetric anaesthesia. Most of the deaths (18 of 27) appear to have been directly due to anaesthesia, the majority of which (13 of 18) were due to general anaesthesia. Human error or incompetency was the main element of the deaths, rather than lack of equipment or facilities. Most deaths occurred in level 1 or 2 hospitals where medical practitioners may be placed in the position of providing anaesthesia for obstetrics without appropriate training.
Key Recommendations
There were 28 deaths reviewed in this category. The term "anaesthetic deaths" was used to denote a death occurring unexpectedly either during or shortly following anaesthesia, which was either directly attributable to anaesthesia or for which no cause could be found. Autopsy examination in one case subsequently revealed an undiagnosed intracardiac tumour as the cause of death, leaving a total of 27 cases.
Generally, the standard of reporting was very poor. A frequent complaint by assessors was of inadequacy of absence of anaesthetic records. Completion of the maternal death notification form was by a doctor in 16, a nurse in 9 and unknown in 3 cases. In the majority of cases insufficient details were presented of the events leading up to the cause of death to enable a cause to be ascribed with certainty. In these cases, cause of death was ascribed on the basis of probabilities. In 10 cases it was impossible to establish a cause of death.
The majority of deaths occurred in level 1 and 2 hospitals (Table 7.1).
Anaesthesia was the direct cause or significantly contributed to death in 19 cases and in one other case anaesthetic care was substandard. The majority of these deaths were associated with general anaesthesia. The commonest single factor responsible for death was difficult or failed intubation (Table 7.2).
Difficulty with tracheal intubation can occur unexpectedly and create problems even for experienced practitioners. However, there are features found on airway examination that may predict difficulty. Insufficient details were provided in the reports to determine whether preoperative airway examination was performed. Morbid obesity was cited as contributory factors in two of the deaths. In one case where intubation difficulty was encountered in a level 1 hospital, the patient had a delayed recovery from anaesthesia and a grand mal convulsion. Nevertheless, the patient was extubated and transferred to a general ward prior to transfer to a level 3 hospital where she arrived in extremis and died following a cardiorespiratory arrest subsequent to further difficulty with reintubation. Although the history suggests significant hypoxic brain damage prior to transfer, the death might have been prevented if the patient had received continued respiratory care in the level I hospital before transfer.
Inadequate preoperative patient evaluation contributed to 2 deaths. A patient died unexpectedly postoperatively in the general ward. An autopsy was not performed but subsequent interview of relatives revealed a history of a "heart condition incompatible with general anaesthesia". (sic)
Following induction of general anaesthesia and intubation, a huge gas leak was discovered in the patient circuit of the ventilator. A second circuit was called for but was found to be in a similar condition. The anaesthetist failed to provide adequate ventilation and the patient died following hypoxic cardiac arrest. This case can only reflect the incompetence of the anaesthetist, not only in his failure to perform a preoperative equipment check, but also his failure to maintain ventilation either with a manual anaesthetic circuit or directly by exhaled air to the endotracheal tube.
There were six deaths associated with spinal anaesthesia due to either high motor block resulting in respiratory paralysis, or catastrophic hypotension and one unknown. Few details were available for evaluation but in one case, respiratory arrest and subsequent cardiac arrest occurred following the start of caesarean section. Instead of continuing to deliver the baby, the surgeon stopped to assist the anaesthetist with resuscitation attempts, with subsequent death of both mother and child. In another case of high motor block an inappropriately high dose of local anaesthetic (17.5mg of hyperbaric bupivacaine 0.5%) was used. This led rapidly to respiratory paralysis and cardiac arrest. This patient also died undelivered. Following maternal cardiovascular collapse, the only hope for the infant is expeditious delivery, and attempts at maternal resuscitation will be fruitless whilst the gravid uterus impedes venous return to the heart.
Hypotension was the cause of death in two cases. In one case a referral centre refused to admit a patient in obstructed labour leaving the doctor at the level 1 hospital no choice but to perform Caesarean section as a single operator anaesthetist. Severe hypotension occurred following delivery. Resuscitation eventually restored blood pressure but irreversible brain damage ensued with subsequent death. In a second case hypotension was complicated by bradycardia following the use of phenylephrine. Inadequate doses of atropine were used and death followed further bradycardia and cardiac arrest.
This hypertensive patient underwent caesarean section for prolonged labour under general anaesthesia. She had an episode of severe hypertension following the use of ergometrine and failed to regain consciousness. Subsequent autopsy revealed a cerebral haemorrhage. However, death was caused by connection of a high-pressure oxygen supply directly to the endotracheal tube during ambulance transfer to a referral centre.
Inadequate documentation and lack of autopsy reports did not permit a definite cause to be established in 9 cases. Four of the cases are unusual in that sudden cardiac arrest occurred immediately following delivery of the placenta under otherwise uneventful general (2 cases) or spinal anaesthesia. Resuscitative efforts appeared to be well directed but with no or limited success. One patient survived long enough to establish severe disseminated intravascular coagulation, suggestive of amniotic fluid embolism which might also have been the cause in the other three cases.
Failure In this case, intra-operative desaturation was noted under spinal anaesthesia for caesarean section. The anaesthetist examined the chest and made a provisional diagnosis of either bronchopneumonia or pulmonary oedema. Haemoglobin saturation improved slightly and the anaesthetist did not intervene to assist ventilation. Had this been done a subsequent episode of hypoxaemia and cardiac arrest in the postoperative recovery room would have been avoided.
Although anaesthesia accounted for comparatively few of the total number of maternal deaths, it is an important cause of maternal mortality. The deaths are not due to disease or obstetric complications but due to substandard care and are preventable.
The term "anaesthetic death" as used in this review is inappropriate and should be abandoned. The vast majority of peri-operative deaths are not due to anaesthesia, and they are included in the relevant chapters when a probably cause of death is known. Labelling the peri-operative deaths where a probable cause cannot be found as "anaesthetic deaths" obscures the true cause of death even further and may give a false overestimation of the contribution of anaesthesia to mortality. On the other hand, substandard anaesthetic care may be contributory factor in deaths due to the other causes, notably hypertension and haemorrhage. None of the assessors were anaesthesiologists and anaesthetic contribution to deaths due to other causes may have gone unrecognised. This, together with underreporting, means that this report probably vastly underestimates the role of anaesthesia in maternal mortality. Future reports should include caesarean section as a separate chapter and both an obstetrician and an anaesthesiologist should review all deaths. Deaths associated with anaesthesia should be assessed by an anaesthesiologist and reported either as directly due to anaesthetic or as cases where anaesthesia contributed to death due primarily to another cause.
The majority of deaths directly attributable to anaesthesia were associated with complications of general anaesthesia. This is in keeping with data from both the UK and USA, which suggest that regional anaesthesia (either spinal or epidural is the method of choice for caesarean delivery unless specifically contraindicated). In this report, general anaesthesia was indicated in only one of the deaths due to general anaesthesia. Facilities for the provision of regional anaesthesia, including staff trained in its use, should be available in every hospital where caesarean delivery is performed.
Over 80% of the deaths occurred in either level 1 or level 2 hospitals, suggesting a generally low level of competence in those called upon to provide obstetric anaesthesia in these cases. This is born out by details from the reports, for example use of inappropriately high doses of local anaesthetic for spinal anaesthesia, lack of knowledge of the difference between spinal and epidural anaesthesia, inadequate patient assessment, no equipment check, use of particulate antacids. Both hypotension and high motor block are recognised complications of regional anaesthesia, are readily treatable and should not result in death, if detected early. The fact that these deaths occurred, together with several women who died undelivered, also calls into question the resuscitation skills of the staff involved. Differences in physiology and pharmacology during pregnancy make special demands upon the anaesthetist, which are not met by undergraduate training.
All those called upon to provide anaesthesia for caesarean section should have experience in obstetric anaesthesia or have received some postgraduate training in its practice. In order to determine how many practitioners require such training, and the form it should take, it is suggested that a National Manpower Survey of obstetric anaesthesia practice be conducted. This should identify all hospitals where operative delivery is performed, the number, grades, qualification and experience of those called upon to provide anaesthesia.