Magnitude, trends and causes of maternal mortality among reproductive aged women in Kersa health and demographic surveillance system, eastern Ethiopia

Improving women’s health and reducing maternal mortality has been a global public health priority for the United Nations international development agenda [1, 2]. Globally, the Maternal Mortality Ratio (MMR) declined from 385 in 1990 to 216 in 2015. However, during the same period, the MMR reduction in sub-Saharan Africa remained stalled and most countries in the region registered sluggish progress in reducing maternal mortality [3‐5]. According to World Health Organization (WHO) estimates, even though the magnitude of MMR in Ethiopia remains high, the level has shown a steady decline from 1250 in 1990 to 353 in 2015 [3]. In the Sustainable Development Goal (SDG) period, the goal is to reduce the global MMR to less than 70 per 100,000 live births by 2030 with no country having MMR over 140 per 100,000 live births [6].

In Ethiopia, the Federal Ministry of Health (FMOH) has applied multi-pronged approaches to reducing maternal morbidity and mortality. These approaches have included improving access to and strengthening facility-based maternal health services [7]. Ethiopia’s Health Sector Development Plan target was to reduce MMR to 267 per 100,000 live births by the year 2015 but the country was unable to meet this target [8]. It was well recognized that the huge burden of maternal mortality in Ethiopia remains the single most serious challenge to the health sector [9]. In the country, the efforts to end preventable maternal mortality is at the top of the health sector’s agenda in line with the SDGs as the issue was targeted in the 2015 health sector transformation plan [10].

The highest number of maternal deaths has been reported in countries where women are least likely to deliver their babies with the assistance of skilled practitioners, such as a nurse/midwife, doctor and other health workers [11]. In many countries, it is those women who are living in rural areas, at the lowest wealth quintile and with less education who are most susceptible to maternal mortality [4, 11]. High maternal mortality levels are also an indication of deep-seated gender inequalities that hinder women’s ability to make decisions about household resources, which in turn could limit their ability to obtain social support and to access maternal health services [12].

Despite efforts at curbing maternal morbidity and mortality, developing countries are burdened with high maternal morbidity as well as mortality and are still facing the challenge of addressing the problem with very limited personnel and material resources [13]. More than 50% of all maternal deaths were from just six countries: Ethiopia, India, Nigeria, Pakistan, Afghanistan, and the Democratic Republic of Congo [14, 15]. The modelled estimate by WHO and the World Bank for Ethiopia showed a MMR of 353 per 100,000 live births in 2015 [3].

In Ethiopia, most studies on maternal mortality have been conducted in a facility setting [16‐18]. Given that a small number of mothers deliver at health institutions in Ethiopia, maternal mortality measurement from facility-based data might not reflect the actual image on the ground [7, 19]. In addition, there is a paucity of information with regards to the local level magnitude and causes of maternal mortality at the community setting in Ethiopia. Furthermore, the differential results in the magnitude of maternal mortality across different estimates for Ethiopia calls for more rigorous and locally generated evidence [3, 7, 20]. This study therefore aimed to investigate the magnitude of, trends in, and causes of maternal mortality among reproductive aged women using surveillance data in a community setting in Eastern Ethiopia.

Using surveillance data, this study was intended to assess maternal mortality and identify the causes of death among reproductive aged women over a seven-year period in Kersa HDSS, Eastern Ethiopia. The study showed a cumulative average MMR of 324 per 100,000 live births with a decreasing trend over the study period. The main causes of maternal death were postpartum haemorrhage (46.5%) and hypertensive disorders of pregnancy (16.3%). Most of the mothers (56%) died at home, and the majority (62.8%) of the mothers died after giving birth.

For every 1000 live births in the study area during the seven year period, about three women died during pregnancy, childbirth or within 42 days of childbirth. This finding is almost similar to the national average (353 per 100,000 live births) but below the sub-Saharan average (546 per 100,000 live births) [3]. Even though the MMR is below the sub-Saharan Africa average, it is higher than the world average, as is the case with other sub-Saharan African countries such as Sudan, Ghana, and Rwanda (Fig. 6) [3]. The overall cumulative average of maternal mortality in the present study is considerably high and is deemed to require government action. However, the observed MMR across the seven-year period appears to decline. The result of the present study is almost in agreement with the finding that was reported from a review of studies conducted in Ethiopia, where the level of maternal mortality was found to decline slightly [17]. The decreasing level of maternal mortality in this study might be related to the current government efforts to improve maternal health by implementing community-based programs that involve community mobilization. Moreover, it might also be explained by the fact that the population under surveillance (through the HDSS) most likely have better awareness and use of maternal health care, which could potentially lead to a lowered MMR. Furthermore, it was found that residing in a HDSS site has a positive influence on maternal health [26].

Nonetheless, the trend of maternal mortality has varied slightly over the years. For instance, there was an increase in MMR in 2011 but a decrease in 2013. The reasons for this are not clear but may have something to do with the relatively higher number of pregnancies and increased number of poor pregnancy outcomes in 2011 compared to 2013. There were more pregnancies in that year, which might be associated with a high agricultural production in the preceding year that enhanced an increased number of families formed among young couples in the district. Moreover, there was a high number of pregnancy failures and still-births in the same year [21]. In connection with this, in the same year compared to other years, the absolute number of maternal deaths may have increased. In addition, the study findings revealed that there had been temporally heterogeneous pattern in rates of maternal death, where MMR was highest during the winter season throughout the study period. The finding highlights that since the winter season has heavy rain falls, the road may get muddy, the rivers might become full and rural bridges could be damaged. Hence the parturient family might not be able to transport the women to a health facility for delivery care and thus most women remain at home to deliver their baby, which might have contributed to the higher rates of maternal death. Moreover, during the rainy season there is an increased rate of malaria transmission, and pregnant women could in turn develop adverse maternal disorders (anaemia and eclampsia) which subsequently contribute to high rates of maternal deaths. It has been shown that malaria-associated anaemia and eclampsia tend to increase during the rainy season among pregnant women in sub-Saharan Africa [27].

Generally, we found the leading cause of maternal death during the study period to be postpartum haemorrhage. This finding is in accordance with other research results from developing countries and sub-Saharan Africa where haemorrhage is the leading direct cause of maternal death [28, 29]. A similar finding was also reported from a community-based study in northern Ethiopia [30]. The finding of the current study, however, is in contrast with a systematic review of studies in Ethiopia which showed the major cause of maternal death to be obstructed labour (36%) followed by haemorrhage (22%) [16]. This difference may be partly explained by the fact that the review only included facility-based studies where most pregnant women came to a health facility very late, with advanced complications such as obstructed labour which leads to prolonged labour [16, 31]. Another possible explanation for this may be that, as the VA data used in the present study is prone to misclassification of maternal deaths, the examiners might have misdiagnosed obstructed labour that leads to haemorrhage. The mothers, however, might have died due to initially developing obstructed labour which is typically not the cause of death but rather due to the haemorrhage that resulted from prolonged labour.

In the current study, there were 543 pregnancy-related deaths per 100,000 live births during the seven-year period. Though the majority of the pregnancy-related deaths were due to causes related to pregnancy or childbirth, non-obstetric causes such as partner violence and transport accidents were also contributors. This highlights that partner violence, including murder and other injuries or accidents to women, contributes to the rates of death among pregnant and postpartum women to a considerable extent in Ethiopia. This underscores the need to design programs that could address issues of gender-based violence at the community level to synergize with current efforts to improve maternal health. The results of the study are similar to a study conducted in Mozambique, in which it was reported that a combination of partner violence and injury were the fourth leading cause of maternal death [32]. The finding further highlights that the social status of women in the community might be a root factor for the high rate of maternal mortality in Ethiopia.

The findings of this study demonstrated that the lifetime risk of maternal death is nearly 1 in 23. This result is lower than a study finding from a community-based survey in southern Ethiopia which was conducted using the sisterhood method, where women have a one-in-ten lifetime risk of deaths [33]. However, this finding is higher as compared to the sub-Saharan African (1 in 39) and national level (1 in 52) [34, 35]. The reason for the observed high LTR of maternal death in the present study might be related to the less precise estimation of the indicator using Mmrate.

In this study, the majority (62.8%) of the mothers died after giving birth and more than half (55.6%) died within the first day. These findings indicate that the first few hours and days after giving birth are a critical period during which mothers should receive immediate attention from health care providers at a health facility to avert catastrophic maternal deaths. The aggregation of maternal death around delivery or immediately after delivery also means that mothers should have access to health facilities to receive skilled care during this period.

In the present study, more than half (60.5%) of the deceased mothers had sought health care services for the health condition that led to their death. However, a substantial proportion (80.8%) of them sought the service from traditional healers at home. This has implications for maternal health behavioural change programs. Traditional healers, including TBAs, still play a paramount role in rendering services at the community level but it has been shown that they are not effective in improving maternal health even when trained [36, 37]. Yet the reliance of local women on TBAs emphasizes the need to understand, at the grassroots level, why uptake of skilled delivery care is low in comparison with TBA utilisation. Perhaps strategies should be designed to provide focused training to TBAs so as to make them capable of recognizing the critical time to seek health care for women in their village [38]. Therefore, there is a need to consider revisiting the strategies for training of TBAs to make them contribute towards the improvement of maternal health.

The magnitude of maternal mortality is considerably high, though it has shown a declining trend. The major causes of maternal mortality were postpartum haemorrhage and hypertensive disorders of pregnancy. Community-based initiatives should be strengthened to further reduce the prevailing maternal mortality. Targeted information education and communication should be provided to illiterate housewife women in their twenties. The health messages targeting these group of mothers should be tailored to their needs and match their level of literacy in order to bring better health outcomes. Future interventions on maternal health in this setting should also be tailored in such a way that women are educated through existing mother peer groups or Women’s Development Army networks at the village level. Moreover, strategic actions are required to promote skilled delivery care attendance and attention should be given to availing community based trained delivery assistants in rural communities.