• No results found

Interventions to Reduce Maternal Mortality

Chapter 1: Background and Objectives

6. Interventions to Reduce Maternal Mortality

Skilled Attendant at Birth

A skilled attendant at birth is one of the key strategies to reduce maternal mortality today [135]. The skilled attendance strategy does not only imply that a trained health staff with midwifery skills (skilled attendant) is available at birth, but also that an enabling environment is put in place to support the skilled attendant to perform life-saving interventions [136].

Although the term skilled attendant should best be used for a provider having comprehensive midwifery skills, in the absence of studies that examine their skills, all nurses, midwives and clinicians are normally counted as skilled attendants [137, 138]. It is important to note that the following studies are using information on ‘skilled attendant’ defined as the presence of a health worker at birth assessed by asking women’s report in household surveys. Ideally information on ‘skilled attendance’ which would imply in-depth studies on actual skills of health workers (see more section on assessment of maternal death p 58) but such assessments are rare.

Much of the evidence supporting the importance of skilled attendance for the survival of mothers and their newborns is derived from historical studies in Europe and the United States, and a few middle income countries [139-141], or ecological studies indicating an inverse association between the proportion of birth with a skilled attendant and maternal mortality [142-144]. Buor and Bream used a few determinants in an ecological explanatory model for maternal mortality and found that skilled delivery (-0.592, p≤ 0.01) was a significant predictor of maternal mortality, followed by life expectancy (-0.489, p≤0.05), and a country’s gross national product (-0.543,p≤0.05)[145]. In a paper from Betrán et al. [146] three variables, the proportion of deliveries by a skilled attendant [parameter -0.016, standard error (SE) 0.004, p<0.0001], infant mortality (parameter 0.013, SE 0.004, p<0.0003) and health expenditure per capita (log parameter –0.272, SE 0.075, p<0.0004) were able to explain 90% of the variations in maternal mortality at country level. Scott and Ronsmans reviewed the evidence from ecological studies and concluded that causal inference is tentative and there is poor controlling of confounders [147].

More recent country examples from Bangladesh support evidence of the importance of skilled attendance for maternal mortality reduction, but also underline the need for a functioning health system and other interventions like the promotion of family planning [25, 148]. A recently published paper reviewing factors that might have contributed to the successful reduction of maternal mortality in Bangladesh, Bolivia, Cambodia, Gambia, Morocco and

53 Rwanda indicated that improved uptake of skilled attendant is likely to have contributed to a decline in maternal mortality, except in Bangladesh [149]. In these countries, remarkable increases in the proportion of birth attended by a skilled attendant between 20 – 50% were observed in the same period that a reduction of maternal mortality was described, but not in Bangladesh [149]. Similar to the observation in Bangladesh, an appraisal of the maternal mortality decline in Nepal could not clearly associate the decline in maternal mortality with the increase in birth attended by a skilled attendant [150].

Evidence from intervention studies on the effect of skilled attendance on the reduction of maternal mortality is lacking. An evaluation of an intervention study, “Strengthened Skilled Attendance Program” in Burkina Faso reported that a 30% decline (adjusted OR 0.7, 95% CI 0.4 – 1.1) in the area included in the skilled attendance program during the time of implementation. However, mortality reductions were also observed in the comparison districts and the differences between the intervention and the two comparison districts were not statistically significant (p=0.439 and p=0.278, no OR available) [90].

At the individual level, the association between birth attended by a skilled attendant and maternal mortality is even more complex. Research examining the effect of a community-based midwifery program in Indonesia revealed that maternal mortality was higher in women delivering with a skilled attendant than without in all wealth groups except the wealthiest [151]. Similarly, the early neonatal death rates were higher in Indonesia for deliveries attended in public facilities than at home [89]. In other studies examining individual-level effects, the expected reduction of maternal mortality when delivering with skilled health providers was not observed in Pakistan nor in Senegal [84, 152, 153]; however, in other settings, lower maternal mortality was found [93, 95, 153].

Higher institutional MM-ratios have been described in many settings and might be due to self-selection and care-seeking in case of complications. Moreover, access to birth attended by a skilled attendant does not automatically imply access to skilled emergency obstetric care and life saving obstetric surgery due to deficient quality [147].

Uptake of birth attended by a skilled attendant varies a lot in relation to the factors described above, such as age, education, distance, and household economy [88, 126, 127, 151, 154]. As most factors might, to a large extent, work through increased uptake of skilled care, skilled attendant might be on the causal pathway between factors such as education or wealth and mortality reduction.

54 Emergency Obstetric Care

The term emergency obstetric care is used to describe a set of interventions needed to save a mother’s life in case of complications. Often Basic Emergency Obstetric Care (BEmONC) and Comprehensive Emergency Obstetric Care (CEmONC) are distinguished (see section 1.5 assessment of quality of maternal care)[155, 156].

A review done by Paxton et al. [157] provides evidence that emergency obstetric care is effective in reducing maternal mortality. The evidence is mainly derived from ecological studies in which relatively strong inverse correlations between MM-ratio and Caesarean section were observed [112]. Country analyses from Malaysia and Sri Lanka point to the importance of emergency obstetric care [148]. Also, country experiences from Nepal and Bangladesh suggest that access to emergency obstetric care might be an important driver of mortality reductions [150]. Further, the six assessments done by the Unmet Obstetric Need (UON) network in West Africa and Tanzania indicate that countries with high maternal mortality have very low levels of surgical intervention [158].

Intervention studies also support the impact of emergency obstetric care. An evaluation of the reduction in maternal mortality in Matlab, Bangladesh over 30 years supports the importance of facilitating access to emergency care. When comparing mortality in the intervention area and comparison area, a skilled attendance strategy improved uptake of emergency obstetric care, and where emergency obstetric services were close by, a lower MM-ratio was observed [25, 159].

In an uncontrolled study using before-and-after analysis, it was reported that improved access to emergency obstetric care (better transport and communication + financial support + improved quality of care at the hospital level) reduced institutional maternal mortality in rural Mali by 50% within two years after the start of implementation [160].

However, emergency obstetric care interventions, primarily Caesarean section, may also bear health risks. An investigation into risks of Caesarean section across 24 countries in 273 health facilities with 286,565 deliveries in Asia, Africa and Latin America indicated a threefold higher risk of dying if the mother had a Caesarean section without indication compared to a spontaneous birth (adjusted OR 3.1, 95% CI 0.8 – 13.2). The risk of death and severe maternal morbidity comparing birth by antepartum Caesarean section without indication and spontaneous birth was highest in Africa (adjusted OR 71.3, 95% CI 32.16 – 158.6) followed by Asia (adjusted OR 2.1, 95% 1.0 – 4.4) and Latin America (adjusted OR 1.9, 95% CI 0.8 – 4.9).

55 These data provide some evidence that Caesarean sections bear health risks, which cannot be explained by the reasons why a Caesarean section is done and that these risks seem to be much greater in Africa than Asia and Latin America[161].

The determinants of access to Caesarean section or other major surgical interventions have been the subject of several studies, which have highlighted major differences between and within countries, with regard to urban and rural settings, wealth and education [158, 162, 163]. Caesarean section may also need to be considered on the causal pathway explaining how place of residence, wealth and other individual factors influence maternal mortality.

Antenatal Care

Antenatal care is an intervention package including a wide range of preventive and curative interventions. The interventions that are included in an antenatal care programme differ between countries. Some interventions are context-specific like the intermitted preventive treatment of malaria in pregnancy (ITPp).

Review articles published in the 1990s questioned the long standing claim that ANC is effective in reducing maternal morbidity and mortality [164, 165]. Investigations into the effectiveness of several single interventions to improve maternal, newborn and child health commonly included in the package concluded that effective interventions might be: iron and folate supplementation; blood pressure screening for high risk groups; dipsticks in urine for detection of infection and hypertensive disorders; and detection, prevention and treatment of sexually transmitted infections, in particular, syphilis. However, the effectiveness of other interventions such as symphysis fundus measurement and palpation of the abdomen to detect malpresentation depend heavily on the skills of the provider. The risk approach, aiming at distinguishing high-risk and low-risk births, has been generally questioned, as the predictive value of most criteria is low [166, 167].

In recent years, a more focused and time reduced approach (only four visits) has been recommended. The recommendation is based a on randomized trial undertaken in the 1990s across different settings, which proposed that a more focused approach with fewer visits had a similar effect on maternal and perinatal outcomes compared to an approach with more visits [168, 169].

Studies that compared mortality in women with and without having attended ANC found, in general, higher mortality in women who did not seek care. Koum reported a significant six-fold increase in maternal mortality in Cambodia among women who did not attend any ANC [114].

56 In Addis Ababa, a twofold increase in mortality for women without ANC compared to those who attended ANC was reported [86]. In a study of a slum in Delhi, an almost threefold increase in mortality among women not registered for ANC was reported [83].

These studies suggest that attendance of ANC reduces maternal mortality. However, there are two possible pathways to highlight how ANC reduces maternal mortality. First, women attending ANC might also be more likely to utilise skilled attendants at birth because they are more educated, richer, live closer to a health facility or face complications. Uptake of ANC and skilled attendance at birth might overlap, but uptake of skilled attendance might be the critical factor reducing mortality. Secondly ANC might also have an effect independent of skilled attendance, for example, by reducing mortality due to preventive interventions such as tetanus vaccination, IPTp and bednets to prevent malaria and giving important information about the importance of professional intrapartum care [170].

Family Planning

Family planning was, together with antenatal care, the first intervention widely made available in low- and middle-income countries as part of the GOBI-FFF6 strategy supported by UNICEF in the 1980s [171]. Later, the importance of family planning for maternal health was reconfirmed as family planning became one of the four pillars of the Safe Motherhood strategy [5]. Today, a three tiered strategy of skilled attendance, emergency obstetric care, and comprehensive reproductive health services including family planning is suggested [172, 173].

The pathway through which family planning influences maternal mortality is threefold: by lowering fertility and thereby lowering the MM-rate; by avoiding unwanted births thus preventing unsafe abortion and thereby lowering the MM-ratio; and by reducing births in high-risk age and parity groups and thereby also lowering the MM-ratio.

It has been estimated that family planning prevents 187 million unintended pregnancies per annum, including 60 million unplanned births and 105 million abortions. Based on this figure, it is estimated that 140,000 – 150,000 lives are saved every year [174]. Ahmed et al. estimated that contraceptive use averted 272,040 (uncertainty interval 127,937 – 407,134) maternal deaths (44% reduction) in 2008 [175], a level also proposed in recent publication by Cleland et al [117]. Stover and Ross concluded that during the past 10 years, there has been a 5 – 35%

decline in maternal mortality, which could be attributed to changes in parity distribution [176].

6Growth monitoring, oral rehydration therapy, breastfeeding, immunisation, family planning, female education, and food supplementation

57 Country analyses confirm the impact. A comprehensive analysis of the reduction of maternal mortality in Matlab, Bangladesh emphasised the importance of family planning for the reduction of maternal mortality [25]. Also, fertility reductions have been discussed as an important factor in the maternal mortality decline in Nepal [150]. Jain (2011) calculated that 22 – 39% of the decline in MM-ratio in Bangladesh, India, and Pakistan was due to the fertility decline [177]. Neither the population growth rate nor the contraceptive prevalence rate was a significant predictor for MM-ratio in the regression model used by Betrán et al. (parameter estimates not available)[146]. In the analysis of determinants of maternal mortality by Buor and Bream [145], the total fertility rate was also not significantly associated with the MM-ratio (parameter estimate 0.192). The actual impact of family planning on the MM-ratio is difficult to estimate.

Table 6: Summary of evidence of strategies to reduce maternal mortality Intervention Pathway Evidence

Birth attended

Unclear result from intervention study [90]

Individually-based studies indicating higher mortality (self-selection bias?) [84, 151-153]

Individual studies indicating a lower mortality (predominantly in countries with higher skilled attendance): [93, 95, 153]

Emergency obstetric care reduces mortality Review: [157]

Historical studies: [148]

Ecological studies: [112, 178]

Intervention studies: [25, 160, 179]

Antenatal care

Selected interventions are highly effective to reduce morbidity [166, 167]

Protective effect on mortality:

Cohort studies: [114]

Case-control studies:[77, 83, 86]

Family planning Preventing unwanted births and births in high risk groups

Model based estimation of reduction of MM-ratio by around 15% by avoiding birth in mothers <20 and>40 years of age[121-123, 125]

Theoretical effect of changes in parity distribution [176]

Country analysis [25, 148, 177]

58