Maternal and perinatal health
Causes of maternal death
Haemorrhage and hypertensive disorders together account for the largest proportion of maternal deaths in developing countries, according to an HRP study, believed to be the first to use the systematic review approach to analyse causes of maternal mortality.
The results of the review, which was published in The Lancet,(Khan KS et al. WHO analysis of causes of maternal death: a systematic review. Lancet, 2006, 367:1066–1074.) are based on an analysis of 160 datasets, or studies, that came from a multiplicity of sources including general and specialized databases, reference lists from studies produced by a search of these databases, personal contacts with WHO country representatives, nongovernmental organizations, journal articles and vital registry data. These sources produced 64 585 titles of reports, from which 1143 potentially usable datasets were identified and finally whittled down to the 160 that were used for the analysis.
Systematic review identifies main causes of maternal mortality and morbidity.
Table 1. Maternal conditions most frequently reported in sudies included in the WHO/HRP systematic review
|Morbidity||Number of studies (%)|
|Hypertensive disorders of pregnancy||885 (14.9)|
|Preterm delivery||489 (8.2)|
|Induced abortion||400 (6.7)|
|Haemorrhage (antepartum, intrapartum, postpartum, unspecified)||365 (6.2)|
|Placenta anomalies (pravia, abruptio, etc.)||245 (4.1)|
|Spontaneous abortion||235 (4.0)|
|Gestational diabetes||224 (3.8)|
|Ectopic pregnancy||146 (2.5)|
|Premature rupture of membranes||140 (2.4)|
|Perineal laceration||139 (2.3)|
|Uterine rupture||116 (2.0)|
|Obstructed labour||102 (1.7)|
|Depression (postpartum, during pregnancy)||96 (1.6)|
|Puerperal infection||86 (1.5)|
|Violence during pregnancy||77 (1.3)|
|Urinary tract infection||66 (1.1)|
|Other conditions||973 (16.4)|
Three important criteria for inclusion of a dataset or study were that: (i) the data had to cover the period 1997–2002; (ii) the study participants had to be pregnant women or women who had ended a pregnancy over the previous year; and (iii) the deaths reported had to have occurred during the pregnancy or within one year of termination of the pregnancy.
The systematic review found that the distribution of causes of maternal death varies by United Nations region. Haemorrhage, for example, is the leading cause of maternal mortality in Africa (Table 2), accounting for 34% of maternal deaths, and also in Asia (Table 3), where it accounts for 31% of maternal deaths. In Latin America and the Caribbean, hypertensive disorders, causing 26% of maternal deaths, top the list of causes (Table 4). In developed countries, the most important cause of maternal death is "other direct causes" (21%), which includes largely complications during interventions such as those related to caesarean section and anaesthesia, followed by hypertensive disorders and embolism (Table 5).
Table 2. Causes of maternal death in Africa
|Other indirect causes of deaths||16.7|
|Other direct causes of deaths||4.9|
Table 3. Causes of maternal death in Asia
|Other indirect causes of deaths||12.5|
|Other direct causes of deaths||1.6|
Table 4. Causes of maternal death in Latin America
|Other indirect causes of deaths||3.9|
|Other direct causes of deaths||3.8|
Table 5. Causes of maternal death in developed countries
|Other direct causes of deaths||21.3|
|Other indirect causes of deaths||14.4|
The previous and most commonly cited estimate, reported in 1991, on causes of maternal mortality also found haemorrhage to be the major killer among women in pregnancy or childbirth. The new study, however, provides insights into the differential ranking of other causes of maternal death in different regions. Deaths due to abortion, for example appear to be a frequent cause in Latin America and the Caribbean, account Africa, sepsis, accounting for 10% of maternal deaths, and HIV/AIDS, accounting for 6%, are clearly major problems.
Logical conclusions emerge from the systematic review. The authors believe that governments in developing countries should give greater emphasis to programmes aimed at preventing and treating the leading cause of maternal deaths, namely haemorrhage, both pre- and postpartum. “At the very least,” they write, “most postpartum haemorrhage deaths should be avoidable by appropriate diagnosis and management.” As for hypertensive disorders, the second most common cause of maternal death, greater use of magnesium sulfate is clearly called for. Finally, deaths resulting from presumably unsafe abortion highlight the need for wider availability of “services that can help women avoid unwanted births”.
Other findings of this systematic review of maternal mortality and morbidity published to date are summarized below.
Estimates of maternal mortality
Estimates of maternal mortality ratios (MMRs), namely, the number of maternal deaths per 100 000 live births, calculated for 141 countries included in the systematic review showed a strong association with three factors: (i) the proportion of deliveries assisted by a skilled attendant; (ii) the infant mortality rate; and (iii) national per capita expenditure on health. The estimates were based mainly on vital registration data and on survey data for the period 1997–2002. The estimated MMRs varied considerably between countries, even countries within a region or countries grouped by development status. MMRs ranged from 127 to 1289 in the least developed countries and from two to 695 in the less developed countries. Development status clearly showed an inverse relationship with MMR: generally speaking, the higher the level of development, the lower the MMR.
Source: Betrán AP, Wojdyla D, Posner SF, Gülmezoglu AM. National estimates for maternal mortality: an analysis based on the WHO systematic review of maternal mortality and morbidity. Biomed Central Public Health, 2005, 5:131 (doi:10.1186/1471-2458-5-131).
Prevalence of severe acute maternal morbidity
A situation in which a "very ill pregnant or recently delivered woman would have died had it not been for luck and good care" has entered the obstetric literature under the term “severe acute maternal morbidity (SAMM)”, more commonly known as “a near miss”. Of the studies included in the overall systematic review, 30 included reports of SAMM cases for the period 1997–2002. An HRP analysis of these reports showed prevalence rates ranging from 0.38% to 8.23% for the period of interest. Complicating the analysis is the existence of several definitions of SAMM. One common definition is based on the disease or disorder that caused the “near miss”, such as preeclampsia, haemorrhage, and so on. A second common definition focuses more on how the problem was managed, such as by hysterectomy or admission to an intensive care unit. A third, more exact, definition is based on failure of specific body organs as a result of pregnancy-related conditions. As per this last definition, the study showed that, of pregnant women who deliver in hospitals in resource-poor areas, 4%–8% will experience SAMM, versus about 1% for women delivering in more developed areas.
Source: Say L, Pattinson RC, Gülmezoglu AM. Systematic review of maternal morbidity and mortality: the prevalence of severe acute maternal morbidity (near miss). Reproductive Health, 2004, 1:3 (doi:10.1186/1742-4755-1-3).
Prevalence of uterine rupture
Uterine rupture, or tearing of the uterine wall during pregnancy or delivery, often results in the death of the baby and sometimes of the mother. In some cases, the uterus suffers irreparable damage and has to be removed. An HRP systematic review covering all the available data on this morbidity found that in most countries prevalence rates are in the 0.1%–1.0% range. The median prevalence rates of uterine rupture in community- and hospital-based studies were 0.05% and 0.31%, respectively. In women who had had a previous caesarean section, the prevalence of ruptured uterus was about 1%. The review was based on 86 groups of women participating in 83 studies. Less developed countries had a higher prevalence rate than more developed countries. Reports from four developing countries— Bangladesh, Ethiopia, Ghana and Nigeria—showed that about 75% of cases of uterine rupture occur in women with an unscarred uterus. They also found that in 1%– 13% of cases of uterine rupture the mother dies and that in 74%–92% of cases the baby dies. The authors of the systematic review suggest four approaches to reducing the prevalence of rupture of an unscarred uterus: (i) by reducing the number of unwanted pregnancies, particularly in high-parity women; (ii) by increasing access to obstetric services, including caesarean section for obstructed labour; (iii) by innovative solutions such as symphysiotomy or caesarean section with local analgesia in areas were conventional caesarean section facilities are unavailable; and (iv) by wide distribution of guidelines on the use of miso-prostol to induce labour to ensure that the drug is used in safe doses.
Source: Hofmeyr JG, Say L, Gülmezoglu AM. WHO systematic review of maternal mortality and morbidity: the prevalence of uterine rupture. BJOG, 2005,112:1221–1228.
Prevalence of stillbirth
The prevalence of stillbirth is on average three times more common in the less developed areas of the world than in the more developed areas. This differential emerged clearly from a systematic review covering 50 countries and 70 studies. The review, which involved a meta-analysis, found that in less developed settings 1.17% of births were stillbirths versus 0.5% in more developed settings. This finding suggests that development status of a country or area is a strong predictor of its stillbirth prevalence. The highest stillbirth prevalence rates, ranging from about 3% to just over 6%, are in western Africa, the review found. The finding that even in developed countries about 1% of all births are stillbirths should, the authors of the review maintain, “alert policy-makers to initiate audit procedures to identify avoidable cases and take action”.
Source: Say L, Donner A, Gülmezoglu AM, Taljaard M, Piaggio G. The prevalence of stillbirths: a systematic review. Reproductive Health, 2006, 3:1 (doi:10.1186/1742-4755-3-1).