Maternal death: lessons for anaesthesia and critical care

Learning objectives.

By reading this article you should be able to:

• •Explain how maternal death is defined and measured along with global and regional trends in maternal mortality.
• •Describe the processes for maternal death reporting and review in the UK and the USA.
• •Identify the key considerations for anaesthetic practice learned from the current literature and enquiries into maternal death.

Key points.

• •Maternal mortality is decreasing worldwide, but internationally agreed targets for further reduction are likely to be missed.
• •Maternal mortality is thought to be increasing in the USA.
• •Annual surveillance and reporting of maternal death in the UK yields lessons for those involved in caring for pregnant and postpartum women.
• •Uterine displacement and perimortem Caesarean section (resuscitative hysterotomy) are crucial to successful resuscitation in pregnancy.
• •Multidisciplinary cooperation and the involvement of senior clinicians are vital to successful outcomes in high-risk pregnancies and maternal emergencies.

Few occurrences are as tragic as maternal death, which has wide and permanent effects on children, families, healthcare staff and wider society. This review examines the current status of worldwide maternal mortality with a focus on the UK and USA, and lessons learned from mortality investigations relevant to anaesthesia and critical care.

Maternal death and its measurement

Maternal death and its subcategories are defined in Table 1. Estimates of maternal death for each country and region are produced by agencies such as the WHO, UNICEF and the World Bank. There are several metrics used to describe maternal death, defined in Table 2. In 2017 it was estimated that 295,000 maternal deaths occurred worldwide.¹ This is an improvement on the estimated 451,000 maternal deaths that occurred in 2000. The concept of ‘obstetric transition’ describes how maternal mortality decreases as a country develops economically and socially. Low- and middle-income countries generally have higher levels of maternal death, consisting largely of direct causes such as haemorrhage and communicable diseases. With increasing development over time, overall deaths decrease and indirect causes of maternal death become more predominant.² A 2014 review by the WHO indicated that approximately 73% of global maternal deaths between 2003 and 2009 were as a result of direct causes. The main causes of death were haemorrhage (27.1%), hypertensive disorders (14%), sepsis (10.7%), abortion (7.9%), embolism (3.2%) and other direct causes (9.6%).³

Table 1.

Definitions of maternal death.¹

• Maternal death: the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management but not from unintentional or incidental causes. o Direct maternal deaths: deaths resulting from obstetric complications of the pregnant state (pregnancy, labour, puerperium) and from interventions, omissions, incorrect treatment, or from a chain of events resulting from any of the above. o Indirect maternal deaths: deaths resulting from previous existing disease or disease that developed during the pregnancy and not resulting from direct obstetric causes but were aggravated by the physiological effects of pregnancy.

•Late maternal deaths: direct or indirect maternal deaths occurring from 42 days to 1 yr after termination of pregnancy.

•Coincidental maternal deaths: deaths from unrelated causes which happen to occur during pregnancy or postpartum.

Table 2.

Measures of maternal mortality.¹

•Maternal mortality ratio (MMR): the number of maternal deaths during a given time period per 100,000 live births during the same time period. In 2017, the global MMR was estimated as 211 maternal deaths per 100,000 live births.∗

•Maternal mortality rate (MMRate): the number of maternal deaths divided by person-years lived by women of reproductive age in a population. This statistic reflects both the risk of maternal death per pregnancy and the level of fertility in the population.

•Proportion maternal (PM): reflects the proportion of deaths among women of reproductive age (15–49 yrs) that are attributable to maternal causes.

•Adult lifetime risk of maternal death: the probability that a 15-yr-old girl (in the year of the estimate) will eventually die from a maternal cause. In 2017 the global probability was 1 in 190.

•Pregnancy-related mortality ratio: an estimate of the number of pregnancy-related deaths for every 100,000 live births (used in US literature).

∗In the USA the term ‘maternal mortality rate’ can be used to describe maternal mortality ratio as defined above.

In the UK there has been a gradual decrease in the number of maternal deaths. Between 2016 and 2018 there were 217 deaths, compared with 295 between 2003 and 2005. Direct deaths decreased by 39% and indirect deaths by 22% over this period. Some 42% of maternal deaths are now from direct causes, of which thrombosis and thromboembolism are the most common. Fifty-eight percent of deaths are from indirect causes with cardiac disease being not just the most common indirect cause but the single most common cause of death overall (Fig. 1).⁴

Fig 1.

Causes of maternal death in the UK for the 2016–18 triennium. Green bars show direct causes of death, red bars show indirect causes of death. Adapted from the 2020 CEMD report.⁴

Presently, the USA does not have a uniform system for collecting maternal morbidity and mortality data at a national level. The current system relies on reporting from individual states. The heterogeneous quality of data makes it difficult to compare maternal mortality rates internationally. Despite having a system that can provide some of the most advanced healthcare worldwide, maternal mortality rates have risen in the USA over the period of 2000–2017 (Figs. 2 and 3).¹

Fig 2.

Causes of pregnancy-related death in the USA for 2014–7 (CDC PMSS).¹³

Fig 3.

Trend in the estimated maternal mortality ratio (MMR) for the UK and USA between 2000 and 2017. Data obtained from the WHO report ‘Trends in maternal mortality’.¹

Improving global outcomes

Global variation in maternal death rates is stark. In 2017 the lifetime risk of maternal death for women in Sub-Saharan Africa was one in 37, compared with one in 7800 for women in Australia and New Zealand.¹ The United Nations Sustainable Development Goal 3.1 aims to reduce the average global maternal mortality rate to <70 per 100,000 live births by 2030. At the current slow trajectory of improvement this goal will be missed, resulting in one million additional maternal deaths by then.¹ The WHO's 2015 strategy for ending preventable maternal mortality is based on strengthening the position of women in society and involves empowering women and girls to make decisions about their health and education and encouraging greater gender equality within societies.⁵ Lower levels of education amongst women have been associated with increased risk of adverse outcomes in pregnancy and maternal death, especially in the developing world.⁶ High quality reproductive, maternal and newborn healthcare that is universally available, accessible and acceptable to women is essential. It is recognised that in order to achieve a sustained reduction in maternal deaths, individual countries need to have an effective political, social and financial infrastructure that will support high-performing health systems.⁵ Lack of accurate death registration and misclassification of causes of death make it difficult for countries and organisations to fully comprehend the extent of maternal mortality, especially in low- and middle-income countries. The WHO recommends international standardisation of death certification and recording pregnancy status on death certificates.⁵

Confidential enquiry into maternal deaths in the UK

The UK's National Surveillance and Confidential Enquiry into Maternal Deaths (CEMD) began in 1952. Since 2012, the enquiry has been conducted by a collaboration called ‘Mothers and babies: reducing risk through audits and confidential enquiries across the UK (MBRRACE-UK). Before 2014 reports were published every 3 yrs, but since then they are produced annually to enable faster changes in clinical practice to be made in response to the findings.⁷ Each annual report is still based on data from the previous 3 yrs. For example, the 2020 report uses data from maternal deaths occurring between 2016 and 2018. This allows analysis of more cases of rare causes of death whilst maintaining patients’ anonymity. Specific causes of maternal death are the subject of chapters every 3 yrs, with anaesthesia being covered in the 2014, 2017 and 2020 reports. The CEMD reports contain anonymised case vignettes to highlight key issues in the care received by women who died. An expedited review into maternal deaths during the global pandemic caused by SARS-CoV-2, has recently been published.⁸

It is not possible to demonstrate a statistical effect of the CEMD reports on maternal death rates because maternal death is a rare occurrence. The clinical impact is likely to be highly significant because maternal death is only the ‘tip of the iceberg’ of adverse events in pregnancy. Therefore, any lessons implemented from the enquiry may reduce the morbidity experienced by women in the future.⁹

Maternal death in the USA

As previously mentioned, the USA does not have a single comprehensive reporting process for maternal death. Information on maternal mortality is collated by the National Center for Health Statistics (NCHS) which collects data on deaths within pregnancy and 42 days postpartum based on death certificates and the International Classification of Diseases (ICD) code listed for the cause of death.¹⁰ In 2019, the NCHS reported a maternal mortality rate of 20.1 per 100,000 live births, compared with 17.4 in 2018.¹¹ Whether an apparent increase in deaths is genuine or reflects improving data capture is uncertain as variability in reporting and data collection persists, despite the best efforts of the Centers for Disease Control and Prevention (CDC), which receives and publishes maternal mortality data.¹² The Pregnancy Mortality Surveillance System (PMSS) is a separate national surveillance system covering all 50 states and also separately New York City and the District of Columbia for a total of 52 locations. The PMSS uses different sources of data and includes late deaths within 1 yr of the end of pregnancy (13–14% of all deaths) to generate a pregnancy-related mortality ratio.¹³ The pregnancy-related mortality ratio has increased steadily from 14.5 per 100,000 live births in 2000 to 17.3 in 2017.¹³

In the USA, in contrast to mortality data, there are reliable morbidity data that reveal the increasing prevalence of obesity and chronic medical conditions such as heart disease, hypertension and diabetes in obstetric patients. There is an association between comorbidity and increasing risk of maternal morbidity and mortality, as demonstrated by the Bateman Comorbidity Index for use in obstetric patients.¹⁴ This may reflect a growing tendency to delay childbearing until later in life.¹⁵ In response to the increase in maternal deaths and marked racial disparities, the CDC has initiated several projects aimed at facilitating the creation and improving the function of existing maternal mortality review committees (MMRCs).¹⁶ Several states are now collaborating to produce recommendations to prevent future maternal morbidity and mortality based on the amalgamated findings of their individual MMRCs. Through data sharing amongst these organisations, it is hoped that a better understanding of maternal mortality is gained and high impact recommendations can be disseminated nationwide.¹⁷

Racial disparities in maternal death in the USA and UK

A disproportionate number of deaths occur amongst women from ethnic minority backgrounds in both the UK and US. Rates are particularly high amongst US-born Black women compared with US-born White women.¹⁰ In the USA in 2014, the ratio of pregnancy-related mortality was found to be between three to four times higher for Black women as compared with White women.¹⁸ The difference in both morbidity and mortality is independent of age and socioeconomic status. Such disparities highlight issues of both access to medical care and variation in the quality of care available to non-White women in the USA.¹⁰ Black women access antenatal care later in pregnancy and attend fewer antenatal appointments than White women. The hospitals used by Black women often have higher rates of maternal morbidity and mortality and are less frequently used by White women.¹⁹

In the UK between 2016 and 2018, the risk of maternal death was more than four-fold higher for women from Black ethnic minority backgrounds compared with White women. Women from Asian and mixed ethnic backgrounds were also at higher risk of death.⁴ Although further research is needed to fully understand the causes of these disparities, it is likely to be multifactorial in both countries. In a recent review, Howell suggested a conceptual model for understanding the poorer outcomes in pregnancy within some ethnic groups in the US.¹⁹ This consists of:

• (i)patient-related factors (biological and genetic determinants of health and disease, socioeconomic circumstances, personal beliefs and behaviour);
• (ii)community and neighbourhood-related factors (housing, social networks);
• (iii)provider-related factors (implicit bias, communication issues); and
• (iv)system-related factors (presence of structural racism [defined below], social and political policy, ease of access to healthcare and its quality).

The 2020 UK Confidential Enquiry report highlighted the ‘constellation of biases’ faced by many women who died.⁴ These included a mixture of physical, mental and social problems, with ‘siloed’ systems of care that prevented women from receiving all the help they needed. In addition, women from ethnic minority backgrounds encounter models of care that may not take account of cultural differences. Structural racism refers to the unfair disadvantage suffered by racial and ethnic groups that is perpetuated by current rules, policies and laws.²⁰ The potential effect on healthcare outcomes has recently captured wider interest of both public and healthcare professionals on both sides of the Atlantic. In response to these concerns, there are now advocacy groups to raise awareness and bring about change, such as ‘Five X More’ in the UK.

Anaesthesia as a cause of maternal death in the UK

Anaesthesia is classed as a direct cause of maternal death. Between 2000 and 2017, there were 25 anaesthetic deaths, representing only a small percentage of total maternal deaths during this period.²¹ Between 2016 and 2018, only one anaesthetic-related death was reported. This is in contrast to the first triennial report in 1952–4 where 49 deaths were directly attributed to anaesthesia and a further 20 deaths where anaesthesia was a contributing cause. Anaesthetists play a broad role in the care of pregnant women beyond simply providing anaesthesia and analgesia, allowing lessons derived from many sections of the enquiry to be applied to anaesthetic practice. There were frequently deaths where input from anaesthetists, or lack of it, was felt by enquiry assessors to have contributed to the poor outcome.²²

Lessons for airway management and induction of anaesthesia

Although failures in airway management continue to feature in recent reports in the UK, failed intubation itself is not a major cause of maternal death despite much anxiety regarding this topic amongst anaesthetists. The joint 2015 Obstetric Anaesthetists' Association (OAA) and Difficult Airway Society guidance for management of difficult or failed tracheal intubation in obstetrics aims to standardise approaches to the obstetric airway against a background of reduced provision of general anaesthesia in obstetrics and therefore reduced practical experience, especially for trainee anaesthetists.²³ Fixation error, also known as ‘anchoring’ or ‘tunnel vision’, has been repeatedly identified in airway crises. It refers to situations where clinicians focus on one possible explanation for a problem without consideration of other possible causes.²⁴ An example from the 2014 report was of an anaesthetist repeatedly focussed on replacing a tracheal tube because of the loss of end-tidal CO₂ measurement without consideration of other potential diagnoses such as bronchospasm. It is recommended that all staff should be made aware of the potential for fixation error and the need for ‘standing back’ and reviewing acute situations with other team members, for example through simulation training. Based on a case where an inappropriately-sized tracheal tube may have complicated intubation during a crisis situation, the 2017 report recommended that the largest tracheal tube used in obstetric patients should be a size 7.0, with smaller tubes being available.²⁵ An individualised risk assessment for pulmonary aspiration, that takes into account risk factors such as obesity or acute illness, should be formulated for each woman. The benefit of antacid medication in reducing the risk of a low gastric pH before anaesthetic induction is highlighted.²⁵

Lessons for neuraxial anaesthesia

Serious complications of neuraxial techniques are rare.²⁶ The 2014 CEMD report stated that two women died after accidental dural puncture, one with a subdural haematoma and one a cerebral vein thrombosis.⁹ Both cases involved inadequate follow-up and delayed investigation. Persistent headache should not be assumed to always be caused by CSF leakage. In response to these deaths, evidence-based guidance on treatment of obstetric post-dural puncture headache has been developed by the OAA.^27,28 Although there have been no recent reports of deaths caused by high or total spinal blocks, a 2017 study of cardiac arrest in pregnancy revealed that a quarter of all arrests were related to anaesthesia, and more than half of anaesthesia-related cardiac arrests were caused by total spinal anaesthesia and the cardiovascular complications of this.²⁹ Since then, international guidelines on the management of hypotension for Caesarean section under spinal anaesthesia have been published.³⁰

Lessons for monitoring

In several cases deterioration after surgery was missed because of poor monitoring and lack of equipment in obstetric settings. Pregnant and postpartum women require the same standard of monitoring (and documentation) as patients receiving anaesthesia in settings other than the labour ward.⁹

Lessons for maternal resuscitation

All those involved in resuscitation of pregnant and postpartum women, including anaesthetists, should be aware of the adaptations required in pregnancy, which have been recently reviewed in this journal.³¹ In particular lateral uterine displacement to offset aortocaval compression in supine women is essential to maximise the efficacy of chest compressions and has repeatedly been highlighted in reviews of maternal deaths where it has been absent. The full left lateral position should be used in refractory severe hypotension.²⁵ During cardiac arrest, a woman in the supine position should undergo manual displacement of the uterus.³¹

In both the UK and the USA, the importance of perimortem Caesarean section (also known as ‘resuscitative hysterotomy’) in women over 20 weeks' gestation has been emphasised. Emptying the uterus improves the chance of survival after cardiac arrest for both the woman and the fetus, if performed swiftly.³¹ It should be performed as soon as possible after cardiac arrest is confirmed and completed within 5 min whenever possible. Delaying perimortem Caesarean section can be lethal, although it should be attempted even if the 5 min deadline is missed, as survivors have been reported.²⁹ Factors causing delay, such as unnecessary fetal assessment before delivery, transfer to an operating theatre and objection to the procedure by relatives, have adversely affected outcomes and been highlighted by CEMD reports.²¹

When profound hypotension occurs as in anaphylaxis, cardiopulmonary resuscitation should be commenced when systolic arterial pressure is <50 mmHg or unrecordable despite a palpable pulse.²⁵ Delivering chest compressions to a beating heart is unlikely to cause harm. The 2020 CEMD report suggests that chest compressions may be of use in venous air embolism, where it may contribute to the breakup of intraventricular air.⁴

Lessons for management of haemorrhage

Failure to recognise the severity of a woman's condition during obstetric haemorrhage is a common cause of suboptimal care identified by CEMD reports. False reassurance from point-of-care haemoglobin values obtained during ongoing haemorrhage has led to inadequate fluid resuscitation and transfusion. Clinicians must be aware that a patient can be hypovolaemic but have a normal haemoglobin concentration if haemodilution from fluid resuscitation has not yet occurred, particularly when early in the course of a haemorrhage situation.⁹ Trends in vital signs and values such as pH, lactate and base deficit should help guide fluid resuscitation and blood transfusion with repeated point-of-care and laboratory tests to assess efficacy of treatment. A woman's body habitus, in addition to the actual (or estimated) volume of blood lost, should be taken into account when estimating the severity of haemorrhage. In several cases the severity of the situation was overlooked in smaller women.⁹ Obesity causes problems because of inaccurate monitoring and difficulty with i.v. access. Consideration should be given to using intraosseous access in emergencies when i.v. cannulation is difficult.⁹

Some of the deaths associated with haemorrhage occurred after tracheal extubation with suboptimal resuscitation beforehand. Haemodynamic instability, ongoing bleeding, significant anaemia, metabolic acidosis and hypothermia should all be corrected before extubation. Residual neuromuscular block and the effect of opioids on respiratory function have been implicated in fatal deterioration after surgery.²⁵

Cardiac disease

Cardiac disease remains the leading cause of maternal death in both the USA and UK. Most of the women who died were not known to have cardiac disease upon entering pregnancy. Cardiovascular risk assessment is recommended for all pregnant women and various toolkits for identifying cardiovascular disease are available.³² In recent CEMD reports, cardiac causes for a patient's symptoms were often overlooked by clinicians who attributed them instead to anxiety, dyspepsia or pneumonia. Pulmonary embolism was frequently considered to be the cause of chest pain or breathlessness even when evidence for it was lacking. A positive diagnosis should be made for significant symptoms and signs such as chest pain, tachypnoea, tachycardia and orthopnoea and women should receive the same investigations as non-pregnant women. Chest pain requiring parenteral opioids should be considered a ‘red flag’ requiring urgent investigation.³³ Anaesthetists have experience in managing critically unwell patients and should be able to alert the wider maternity team to potential cardiac concerns.

Lessons for prescribing

The use of NSAIDs in a woman with pre-existing renal impairment has been highlighted. Creatinine concentration decreases in pregnancy, so an apparently normal level may indicate renal dysfunction.⁹

Women receiving anticoagulants during pregnancy need a plan to guide dosing during the antenatal, intrapartum and postpartum periods. In the UK, anaesthetists are often responsible for prescribing thromboprophylactic drugs after operative delivery. This differs from the USA where this is managed by the obstetric or surgical team. The 2015 CEMD report documented four maternal deaths from venous thromboembolism where postpartum thromboprophylaxis had been inappropriately delayed.³⁴ UK guidance suggests postnatal thromboprophylaxis with low molecular weight heparin (LMWH) should be given 6–8 h after delivery regardless of whether this coincides with drug rounds, with at least 4 h since spinal injection or removal of epidural catheter.³⁴ The 2018 guidelines from the American Society of Regional Anesthesia and Pain Medicine suggest a delay to starting prophylactic LMWH 6–12 h after delivery, with at least 4 h having elapsed since catheter removal.³⁵

Organisational considerations

Review of maternal deaths repeatedly highlights delays in assembling the correct personnel to care for sick women. Senior clinicians are often involved too late in the management of complex or high-risk patients. Having a senior clinician take a ‘helicopter view’ of a crisis situation was recommended in the 2020 CEMD report.⁴ Lack of antenatal coordination of the care of women with complex medical problems is also a recurring theme. Management should be led by a consultant obstetrician. In the USA, this would be the equivalent to the attending obstetrician who would coordinate the input of other subspecialties.

Lessons for critical care

Critical care should be undertaken as soon as it is warranted, regardless of location. It should not depend on a physical intensive care bed being available. Obstetric anaesthetists, critical care doctors and outreach staff should work collaboratively to enable this.³³ In the UK, there is a national recommendation that admission to a critical care unit should occur within 4 h of a decision for admission.³⁶ Women admitted to critical care should continue to receive input from obstetricians, obstetric anaesthetists and midwives who are best placed to guide treatment of obstetric conditions.

In the USA, there is guidance on the transfer of pregnant patients for both maternal and neonatal indications, with facilities and staff capabilities described for each level of care. Regional collaboration amongst institutions to ensure women receive the most appropriate level of care is recommended.³⁷ Guidance based on recent CEMD reports suggests transfer of critically unwell pregnant women should be avoided unless appropriate care or intervention cannot be provided locally. Whenever possible, expert opinion or review should come to the woman, not necessitate her transfer. If transfer cannot be avoided, it should be to a location that can provide all required input, including obstetric services.³³

The increased susceptibility to respiratory infection in pregnancy and the current ongoing COVID-19 pandemic highlights the importance of a low threshold for escalation of care in expectant mothers. The MBRRACE-UK rapid review of COVID-19 maternal deaths conducted during the first wave of the pandemic revealed previously reported issues with delayed escalation of care for critically unwell women, absence of multidisciplinary working and lack of senior input, albeit in a healthcare system under much strain.⁸

Conclusions

Maternal mortality remains a global problem that unevenly affects different countries and ethnic groups. Efforts to sustain the worldwide reduction in maternal death will require societies and governments to allocate sufficient resources to women's health and to value and empower women. In the USA, increasing maternal mortality rates are not attributable to a single cause and may partly reflect recent advances in reporting. The UK's CEMD highlights the range of clinical scenarios where the involvement of anaesthetists in multidisciplinary care could help reduce morbidity and mortality experienced by pregnant and postpartum women, beyond the provision of anaesthesia.

Declaration of interests

The authors declare that they have no conflicts of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Biographies

Alexander Walls FRCA is a specialist trainee in anaesthesia within the Imperial School of Anaesthesia, London. His interests are obstetric and regional anaesthesia.

Felicity Plaat FRCA is a consultant anaesthetist at Queen Charlotte's and Chelsea Hospital, London. She is immediate past president of the Obstetric Anaesthetists' Association and a council member of the Royal College of Anaesthetists. Her specialist areas are obstetric anaesthesia, medicolegal and ethical aspects of anaesthesia. She has been an anaesthetic assessor for MBRRACE-UK.

Angelica Delgado MD is an anaesthesiologist with an interest in obstetric anaesthesia. She completed her internship and residency training at New York-Presbyterian Hospital/Weill Cornell Medicine, including an internship in obstetric anaesthesiology.

Matrix codes: 1I01, 2B05, 3B00