Maternal mortality reviews are used globally to assess the quality of health-care services. With the decline in the number of maternal deaths, it has become difficult to derive meaningful conclusions that could have an impact on quality of care using maternal mortality data. The emphasis has recently shifted to severe acute maternal morbidity (SAMM), as an adjunct to maternal mortality reviews. Due to its heterogeneity, there are difficulties in recognising SAMM. The problem of identifying SAMM accurately is the main issue in investigating them. However, admission to an intensive care unit (ICU) provides an unambiguous, management-based inclusion criterion for a SAMM. ICU data are available across health-care settings prospectively and retrospectively, making them a tool that could be studied readily. However, admission to the ICU depends on many factors, such as accessibility and the availability of high-dependency units, which will reduce the need for ICU admission. Thresholds for admission vary widely and are generally higher in facilities that handle a heavier workload. In addition, not all women with SAMM receive intensive care. However, women at the severe end of the spectrum of severe morbidity will almost invariably receive intensive care. Notwithstanding these limitations, the epidemiology of intensive care admissions in pregnancy will provide valuable data about women with severe morbidity. The overall rate of obstetric ICU admission varies from 0.04% to 4.54%.
Introduction
Maternal deaths are studied almost globally to assess the quality of health care of a country. Most countries with organised health-care systems would have a national maternal mortality audit in place, while the others will at least be striving towards one . With improved health resources, maternal mortality has declined significantly over the past few decades. This relatively low incidence limits the value of maternal mortality data as a tool for deriving meaningful clinical conclusions. Even in resource-poor settings, where maternal mortality figures are relatively higher, maternal mortality data have been described as one of the worst-performing health indices . In order to improve the quality of a maternity care service, periodic reviews or audits of other parameters have now become necessary. Audits of severe maternal morbidities alongside maternal mortality reviews are now gaining ground for this purpose .
A classic example of this approach is the Confidential Enquiries into Maternal Deaths (CMED) in the UK, which has observed a clearly documented fall in maternal mortality ratios over the past 50 years . With the fall of maternal mortality rates to 7 per 100,000 in the early 1990s, it was recommended that maternal morbidity, in its severe form, should be audited . In the year 2000, the 1997–1999 triennial report of the CMED in the UK included a chapter on ‘near-miss and severe maternal morbidity’ for the first time .
There are difficulties in defining and recognising women who would fit into a definition of severe maternal morbidity. However, a woman who was admitted to an intensive care unit (ICU) could be considered to have suffered severe maternal morbidity. This chapter examines the importance and limitations of studying the epidemiology of intensive care admissions in obstetrics.
The major global causes of maternal deaths are post-partum haemorrhage, severe pre-eclampsia/eclampsia, obstructed labour, puerperal sepsis and unsafe abortion. However, the incidence of each of these may vary between countries and may be determined to a good extent by the availability of resources. Two studies from Nigeria and West Africa showed that, in addition to haemorrhage and hypertension, obstructed labour and sepsis remain significant risks to the mother . It is common knowledge that most women with these pregnancy complications would survive with intensive care available to them. A study of such patterns, causes and effects of health and disease allows informed and evidence-based policy decisions to be made for optimal utilisation of finite resources. At present, robust evidence on the incidence of severe maternal morbidities is limited. A recent Cochrane review on critical incident audit and feedback to improve perinatal and maternal mortality and morbidity revealed that there were no suitable randomised controlled trials in this regard . In order to audit the maternal morbidity outcome, certain predefined outcome measures are needed. A review of epidemiology of obstetric critical care was earlier published in the same journal and it reviewed available literature for 20 years from 1980 to 2004 . Here, we describe recent evidence on the same topic with special emphasis on trends in the epidemiology of obstetric critical care.
Terminology
Severe obstetric morbidity was earlier referred to as ‘near misses’, a terminology that originated in the aviation industry where it described a critical event where no loss of life or collision occurred. In obstetrics, it implies any severe morbidity due to any cause related to or aggravated by pregnancy or its management but not due to accidental or incidental causes regardless of the site or duration of the pregnancy, up to 42 days from delivery . Filippi et al. defined it as a severe life-threatening obstetric complication necessitating an urgent medical intervention in order to prevent the likely death of the mother .
Mantel et al. defined a woman with severe acute maternal morbidity (SAMM) as “a very ill pregnant or recently delivered woman who would have died had it not been but luck and good care was on her side .”
A systematic review conducted by the World Health Organization (WHO) on severe maternal morbidity and maternal near-miss cases in 2003/2004 emphasised a substantial heterogeneity in terminology and definitions used in obstetric morbidities . As a result, it established a technical working group comprising obstetricians, midwives, epidemiologists and public health-care professionals to develop a standard definition and uniform identification criteria for maternal near-miss cases . In April 2009, a paper was published by this group defining maternal near-miss morbidity as “a woman who nearly died but survived a complication that occurred during pregnancy, childbirth or within 42 days of termination of pregnancy .”
Terminology
Severe obstetric morbidity was earlier referred to as ‘near misses’, a terminology that originated in the aviation industry where it described a critical event where no loss of life or collision occurred. In obstetrics, it implies any severe morbidity due to any cause related to or aggravated by pregnancy or its management but not due to accidental or incidental causes regardless of the site or duration of the pregnancy, up to 42 days from delivery . Filippi et al. defined it as a severe life-threatening obstetric complication necessitating an urgent medical intervention in order to prevent the likely death of the mother .
Mantel et al. defined a woman with severe acute maternal morbidity (SAMM) as “a very ill pregnant or recently delivered woman who would have died had it not been but luck and good care was on her side .”
A systematic review conducted by the World Health Organization (WHO) on severe maternal morbidity and maternal near-miss cases in 2003/2004 emphasised a substantial heterogeneity in terminology and definitions used in obstetric morbidities . As a result, it established a technical working group comprising obstetricians, midwives, epidemiologists and public health-care professionals to develop a standard definition and uniform identification criteria for maternal near-miss cases . In April 2009, a paper was published by this group defining maternal near-miss morbidity as “a woman who nearly died but survived a complication that occurred during pregnancy, childbirth or within 42 days of termination of pregnancy .”
Classification
Many different classifications of severe maternal morbidity have been described in the literature. For the purpose of better description here, we discuss two major approaches: (a) by outcome identification criteria and (b) by region. Outcome identification criteria can be further analysed by: (1) disease-specific criteria (e.g., severe pre-eclampsia and severe post-partum haemorrhage), (2) management-based criteria (i.e., admission to ICU and need for a blood transfusion) or (3) organ system dysfunction-based criteria .
Identification of disease-specific criteria
An audit of SAMM by predefined specific clinical entities such as severe pre-eclampsia, eclampsia, massive haemorrhage, severe sepsis and uterine rupture has been popularised, as these clinical entities are unlikely to be under-reported. The use of such clearly defined clinical morbidities may be appropriate for units with limited facilities to store health records electronically. As it is relatively easy to define severe obstetric morbidities according to the specific clinical entities, criteria based on specific pregnancy complications report a higher percentage of near-miss cases and a wider range of estimates compared with the other criteria . A drawback of this method is that whilst some conditions are easily defined and diagnosed, conditions such as pulmonary embolism may be difficult to diagnose or define in the survivors. Most of the studies carried out using this classification have used established criteria for the degree of severity for the selected clinical conditions ( Table 1 ). The upper range of the severe maternal morbidity rates varied from 3.21% in Middle Eastern, to 4.92% in Latin American, to 5.41% in Asian to 6.03% in African countries. By contrast, studies from high-income countries in Europe reported an upper near-miss rate from a low of 0.69% . Moreover, near-miss rates of developed countries have been consistent over the past 20 years and haemorrhage is responsible for the most near misses . Even though haemorrhage is still the leading cause of maternal near misses in developing counties, sepsis also contributes significantly .
Reference | Study period | Region | Morbidities | Total deliveries | Number with morbidity | Near-miss rate (%) | Maternal deaths | Death to morbidity ratio |
---|---|---|---|---|---|---|---|---|
Jabir et al 2013 | 2010 | Sub-Sahara | Hemorrhage 40% Hypertensive disorder 22% Sepsis 5% Ruptured uterus 14% | 25,472 | 129 | 0.51 | 16 | 0.12 |
Siddiqui et al., 2012 | 2010 | Asia | Haemorrhage 34% Hypertensive disorders 29% Ruptured uterus 12% | 1508 | 111 | 7.3 | 19 | 0.17 |
Lotufo et al., 2012 | 2008 | South America | Hemorrhage 40% Hypertensive disorders 33% Sepsis 2.3% | 9683 | 158 | 1.6 | 5 | 0.03 |
Lori et al., 2012 | 2008 | West Africa | Hemorrhage 42% Anaemia 21% Sepsis 21% Hypertensive disorders 11% Ruptured uterus 5% | Not documented | 120 | 28 | 0.23 | |
Shrestha et al., 2010 | 2008-2009 | Asia | Hemorrhage 42% Hypertensive disorders 28% Sepsis 19% Obstructed labour 3% | 1562 | 36 | 2.3 | 5 | 0.14 |
Almerie et al., 2010 | 2006-2007 | Middle Eastern | Hypertensive disorders (52%) haemorrhage (34%) | 28 025 | 901 | 3.21 | 15 | 0.02 |
Roost et al., 2009 | 2006-2007 | Latin America | Haemorrhage (47%) Hypertensive disorders (46%) Sepsis (3%) | 8136 | 401 | 4.92 | 15 | 0.04 |
Mustafa et al., 2009 | 2006 | Asia | Haemorrhage (51%) Anemia (21%) Dystocia (14.8%) Hypertensive disorders (8.5%) Infections (4.2%) | 868 | 47 | 5.41 | 6 | 0.13 |
Zwart et al., 2008 | 2004-2006 | Europe | Haemorrhage (63%) Uterine rupture (9%) Hypertensive disorders (9%) | 371021 | 2552 | 0.69 | 48 | 0.02 |
Donati et al., 2012 | 2004–2005 | Europe | Hemorrhage 40% Hypertensive disorders 29% | 539 382 | 1259 | 0.23 | 90 | 0.071 |
Ben Hamouda et al., 2007 | 1999-2003 | Africa | Haemorrhage (39%) Hypertensive disorders (25%) | 19736 | 119 | 0.6 | 6 | 0.05 |
Brace et al., 2004 | 2001-2002 | Europe | Haemorrhage (50%) | 51165 | 196 | 0.38 | 4 | 0.02 |
Filippi et al., 2005 | 1999-2000 | Africa | Haemorrhage (35%) Hypertensive disorders (27%) Anemia (18%) Sepsis (5%) | 47 477 | 2864 | 6.03 | 197 | 0.07 |
Keizer et al., 2006 | 1990-2001 | Europe | Hypertensive disorders 62% Haemorrhage 18% | 18 581 | 142 | 0.76 | 7 | 0.05 |
Baskett et al., 2005 | 1988-2002 | Northern America | Haemorrhage 65% Hypertensive disorders 17% | 159 896 | 313 | 0.19 | 3 | 0.009 |
Identification of management-based criteria
Admission to intensive care
The admission of a pregnant or a recently delivered mother to an ICU is an easily definable and auditable critical event. The data are likely to be easily available and reliable, prospectively or retrospectively, even in low-resource settings. Against the backdrop of the problems that beset defining severe morbidity, this is a major reason for studying intensive care admissions. It provides a definite criterion that indicates severe illness that could be used as an inclusion criterion. The overall incidence of intensive care admissions ranges from 0.04% to 4.54% .
In considering ICU admission as a surrogate marker for SAMM, certain ground realities must be considered. First and foremost, thresholds for ICU admissions may vary across settings and between facilities. Zwart et al., in a large population-based prospective study in the Netherlands, found a direct relationship between the workload and level of a hospital and the threshold for admission to ICU . The workload, available resources in the hospital and the threshold for admission to the ICU would determine the incidence of intensive care admissions in pregnancy. Therefore, there is an element of relativism in a woman being admitted for intensive care. A woman who may receive intensive care in one hospital may have been cared for in a high-dependency unit (HDU) or labour ward in a hospital that cares for a greater number of women with a higher baseline risk. Zeeman et al. estimated that an HDU would cater to the needs of 50% of women requiring higher levels of care . Second, not all women who have SAMM will be admitted to an ICU. The study by Zwart et al. and a Scottish study found that only about a third of women with a SAMM received intensive care . The reasons for this are complex, but in general, the more severe cases of SAMM would receive ICU care. The figures from resource-poor countries are likely to be worse than those above. Admission to an ICU will also depend on the accessibility to ICU facilities, a criterion that could differ widely between settings. Obviously, the accessibility of intensive care will greatly influence the numbers studied. Admission criteria to ICUs may also vary, as does what constitutes intensive care ( Table 2 ).
Reference | Study period | Region | Reasons for ICU transfer | Total deliveries | Number of ICU transfers | ICU transfer ratio (%) | Maternal deaths | Death to ICU transfer ratio |
---|---|---|---|---|---|---|---|---|
Rios et al., 2012 | 2008-2010 | South America | Haemorrhage 52% Hypertensive disorders 37% Septic 3% | 30053 | 242 | 0.81 | 6 | 0.03 |
Bibi et al., 2008 | 2006 | Asia | Hypertensive disorders 50% Sepsis 17% | 2224 | 30 | 1.35 | 10 | 0.33 |
Small et al., 2012 | 2005-2011 | Northern America | Cardiac 36% Haemorrhage 29% Hypertensive disorders 9% Septic 9% | 19575 | 86 | 0.44 | 1 | 0.01 |
Zwart et al., 2008 | 2004-2006 | Europe | Haemorrhage 63% Uterine rupture 9% Hypertensive disorders 9% | 371021 | 847 | 0.23 | 48 | 0.05 |
Wanderer et al., 2013 | 1999-2008 | Northern America | Hypertensive disorders 30% Haemorrhage 19% Sepsis 11% Ectopic 10% | 765,598 | 2927 | 0.38 | 53 | 0.01 |
Richa 2008 | 1998-2005 | Middle Eastern | Hypertensive disorders 27% Sepsis 27% Haemorrhage 20% | 6250 | 15 | 0.24 | 5 | 0.33 |
Quah et al., 2001 | 1998-1999 | Asia | Hypertensive disorders 50% Haemorrhage 24% Sepsis 3% | 31725 | 239 | 0.75 | 3 | 0.01 |
Al-Suleiman et al., 2006 | 1992-2004 | Middle Eastern | Haemorrhage 33% Hypertensive disorders 17% Septic 6 | 29432 | 64 | 0.22 | 6 | 0.09 |
Bouvier-Colle et al., 1996 | 1991-1992 | Europe | Haemorrhage 37% Hypertensive disorders 28% Septic 6% | 140323 | 435 | 0.31 | 22 | 0.05 |
Keizer et al., 2006 | 1990-2001 | Europe | Hypertensive disorders 62% Haemorrhage 18.3% | 18684 | 142 | 0.76 | 7 | 0.05 |
Baskett et al., 2005 | 1988-2002 | Northern America | Haemorrhage 65% Hypertensive disorders 17% | 159 896 | 83 | 0.05 | 3 | 0.06 |
Baskett et al., 2009 | 1982-2005 | Europe | Combined haemorrhage and hypertensive disorders 56% | 122001 | 93 | 0.08 | 4 | 0.04 |