Background
Improving maternal health has received recognition at the global level as evidenced by its inclusion in the Millennium Development Goals (MDGs) with goal to reduce maternal mortality by three fourth by the end of 2015 [
1,
2]. The maternal health issue has also continued to be one of top post-2015 sustainable development goal (SDG) agendas and a target for maternal mortality reduction was set [
3]. Though, reduction in maternal mortality has traditionally been used as a critical measure of maternal health, it represents only a “tip of the iceberg” of the burden the maternal morbidity and resulting short and long-term sufferings. For every woman who dies of pregnancy-related causes, 20 or 30 others experience acute or chronic morbidity [
2]. Therefore relying solely on maternal mortality to assess maternal health overlooks pregnancy continuum from normal to death. In that continuum, pregnancy may be uncomplicated or complicated. Complications range in severity from minor morbidity to potentially life-threatening conditions (PLTC) and life-threatening complications (LTC) [
4‐
7].
In life-threatening pregnancy related complication the woman has one of the two SMO: she may die (maternal deaths) or narrowly escape death (maternal near-miss) cases [
5]. Maternal death is defined as death of women during the time of pregnancy, labor and delivery or within the first 42 days after delivery/termination of pregnancy, irrespective of the duration and the site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes. On the other hand, maternal near-miss case is defined as “a woman who nearly died, but survived a complication that occurred during pregnancy, childbirth, or within 42 days of termination of pregnancy”. Practically, women are considered near miss when they survive organ dysfunction [
5,
6].
Women who survive life-threatening conditions (MNM cases) have many aspects in common with those who die of such complications. The assumption is that, a woman suffering from a near miss event is exactly like the one who died, except for the outcome [
7‐
9]. Similar to the case of MD most cases of SMM are preventable. This SMM cases can also be considered a near miss for maternal mortality because without identification and treatment, in some cases, these conditions would lead to maternal death. Thus, the identifications of SMM is crucial for preventing complications that led to mortality and for highlighting ways and opportunities to avoid similar cases in the future.
This has stimulated an interest in investigating cases of SMM in the last few decades and led to the development of the near-miss approach for maternal health [
5,
6]. This approach has been found to be useful for the identification of health systems failures and a relevant source of information for policy makers in the selection of maternal health care priorities. Among other positive characteristics, investigating near miss events allows calculation of comparative indices and quantification of intensive care requirements. It also gives the opportunity to interview the woman, as she survived, providing valuable information on the risk factors and substandard care she received. Moreover, investigating the care received may be less threatening to providers because the woman survived [
5,
9,
11].
In recent years the near-miss approach for maternal health is being widely used and found to be a valuable tool in, among others, understanding local patterns of maternal mortality and morbidity, strengths and weaknesses in the referral system, and the use of clinical and other health-care interventions [
5,
6,
11]. Despite its wide application, there have been challenges with its use mainly due to absence of universal criteria for identification of cases. Criteria for MNM identification were developed by experts from WHO working group for maternal health so as to standardize detection of MNM cases. This technique of MNM identifications proposed by WHO is a two-step process. First; maternal cases with potentially life-threatening conditions, which may or may not be near-miss cases (e.g. specific complications such as severe preeclampsia and/or critical interventions such as blood transfusion) are identified and then identification of near-miss cases based on organ system dysfunction and organ-dysfunction proxies including clinical, laboratory and management criteria [
5,
12]. This criterion is validated in different setups and its applicability depends on the local context and availability of resources [
12,
13].
In Ethiopia also, MNM review is considered one of the many strategies to tackle the high maternal mortality [
6,
14]. The national Maternal Death Surveillance and Response (MDSR) system recommends all MDs and 50% of MNM cases be reviewed in hospital facilities [
6]. However, there is limited experience with the use of near-miss approach in Ethiopia and there are few studies available. Specifically, MMR in the study facility, Jimma University Teaching Hospital (JUTH), is very high, 888.5 per 100,000 live births [
15] in a review 10 years back. The reason for this high MMR is not studied and there is no study on MNM in JUTH.
This study therefore has determined the incidence and determinants of SMM and SMO in the study facility. In addition, the quality of maternal health care provided at JUTH was assessed using WHO near-miss approach and calculated indicators recommended by WHO. It is the first prospective cross-sectional study conducted to assess cases of MNM and MD in the obstetric practice of southwest part of Ethiopia.
Discussion
This prospective cross-sectional study of woman with various pregnancy related complications using the near-miss approach to maternal health is first of its kind in the study institution. Our study has found that the burden of maternal ill-health in population the hospital serves is high. In addition the finding of high proportion of complication already present on arrival to JUTH indicates the significance of delay at the different level of care. Our study reaffirmed this theory. Unlike many other studies on near-miss, our study has compared women with extremely severe outcome to woman with less severe outcome and identified significant contributors to their severity. However, comparison of our study findings with that of other similar institutions in the country may not be enough because of the paucity of studies done using this approach of maternal health assessment.
Our study showed the ration of SMO incidence and MNM incidence to be 59.2 and 50.4 per 1000 live births. The MMR was 876.9 per 100,000 live births whereas, the MI was 14.8%. In general, the rate of occurrence of SMO indicators were found to is higher than the findings of earlier study done in other part of the country [
9,
14,
16] and other developing countries [
10,
11,
13,
17‐
24]. Moreover, the MMR was higher than the national average MMR report for the year 2013 [
1]; and the findings of earlier facility based study done elsewhere in the country [
14,
15,
25‐
29], and that of other developing countries [
11,
13,
19‐
22,
30]. This high incidence ratio might be due to the fact that our study facility is the only referral hospital in south-west region of the country serving complicated cases referred from other health facilities in its catchment. Moreover, the occurrence of a high rate of intra-hospital SMO and MI, higher than the findings from otter facilities in the country [
31] and other developing countries [
19] puts the quality of obstetric care offered to women with obstetrical complications in JUTH under question.
Our study showed that direct obstetric causes were the most common underlying factors of SMM and SMO. The most common underlying morbidity of SMM were hypertensive disorders and obstetric hemorrhage similar to the findings of the study done elsewhere in the country [
16] and other developing countries [
19,
22]. However, uterine rupture, obstetric hemorrhage and hypertensive disorders were found to be the top leading underlying complications among cases of SMO and this is comparable to the findings from studies in other parts of the country [
9,
14,
25,
31] and other developing countries [
10,
13,
17‐
24] and sub-Saharan countries [
32].
Of the total of 24 MD, 21 of them died of direct obstetric causes and the rest three were indirect maternal deaths comparable with earlier studies elsewhere in the country [
14,
15,
26] other developing countries [
11,
13,
17,
22,
24,
33]. Our study showed the leading causes of MD to be hypertensive disorders followed by obstetric hemorrhage and this finding is in agreement with the findings of earlier study done in different hospitals from Ethiopia as a part of FIGO logic initiative and the review of maternal mortality trend in Ethiopia [
14,
26]. Eclampsia was associated with the highest case fatality; similar to the findings of earlier studies done in other parts of the country [
9,
14] and other developing countries [
17,
21,
32].
Unlike earlier studies in the same facility and other parts of the country [
15,
26,
28], uterine rupture is no more the leading cause of MD. This could be due to improved care for uterine rupture cases at the facility. The fact that 30 out of 43 (70%) cases of uterine rupture cases presented after laboring at home for over 12 h show the delay in seeking care. Uterine rupture as the leading causes of MNM. This is in contrast to the findings from earlier studies in this and other health facilities in the country [
15,
26,
29,
34]; in our study there was no case fatality because of complication from abortion supplementing the findings of study stating case fatality from complications of abortion in Ethiopia was declining [
27,
28]. Perhaps this could be attributed in part to the recent ‘revision of abortion law and its legislation’ in the country which ensures the provision of safe abortion services for selected group of women, avoiding fatal complications from unsafe abortion [
35].
Emergency obstetric care use by women is influenced by a complex interaction of factors leading to delay in decision-making, accessing services and receipt of proper care once a health facility is reached [
1,
5,
18,
24,
30,
36]. Similar to the case of our study; delay one, delay two and delay three were reported as a significant contributor to SMO in several earlier studies from developing countries [
10,
18,
30,
36]. In this study, the occurrence of delay one, two and three was 45.1, 57.1, and 59.1% respectively, similar to studies done elsewhere in the country [
9,
14,
29]. Delay-1 (delay in seeking health care) is seen more frequently than the findings from other developing countries [
37] however it is better than the finding of earlier studies done in Ethiopia [
9,
14]. This can be justified by a difference in socio-demographic characteristics of the study population and the year of study.
According to WHO the fact that large number women arrived to health care facility with SMO indicate occurrence of the first (delay in recognizing a condition as a complication and delay in seeking help) and second (delay in reaching a health-care facility once the decision to seek care has been made) delays in the health district [
5]. This fact is well supported by the findings of our study. In our study, close to three-fourth of women had SMO on admission or within twelve hours of arrival. This finding is also in line with studies from other institutions in Ethiopia and other developing countries [
11,
14,
16,
28,
31].
In this study, delay three (at first and/or last facility) was found to have the strongest association with SMO: with a four-fold increase in the risk. This supports the WHO hypothesis relating a high case fatality in the hospital as an indicator for the presence of delay in receiving an adequate and appropriate treatment [
14,
37]. Seeking care from a facility that is ill-equipped to give emergency obstetric care contributes to significant delay even after reaching the health facility. These factors were reported as significant contributors of delay in several studies [
25,
37,
38]. These non-functional health facilities are physically accessible and described as “physical obstacles” for pregnant women in accessing a functioning health facility in time [
36]. In line with this, studies from northern Ethiopia attributed 88% of all maternal deaths to health system failure [
29]. In our study, 77% of SMO cases, and 75% of the 24 MDs had health system related factors as a possible reason for delay three.
Almost all SMO on admission cases were transferred from other facilities, indicating presence of significant deficiency in the referral system. Large proportions of woman are kept unnecessarily before decision to refer is made. Even after deciding to refer these women, most are transferred without initiation of treatment. These can be due to professional factors (negligence or lack of skill) and/or possibly lack of the necessary supplies to provide the initial treatment, which are both not assessed in this study.
The fact that large number of woman reach to a facility with a complication indicates the complexity of care required by the population served by the health-care facilities in the assessment [
5]. The complexity of care and treatment provided for obstetric patient ranges from basic to intensive care and thus the level of health facilities is different too [
39]. The provision of appropriate health care for critically ill obstetric cases requires proper staffing, equipment, and management strategies and this health services obviously contribute to a better outcome among women with life-threatening conditions [
24]. But, health facilities in the developing world are chronically under-resourced [
24,
29]. Though most women presenting with organ dysfunction needs to be managed in ICU, the overall ICU admission rate of cases with SMO was low (14.8%) in this study. Moreover, only 40% of MD cases were managed in ICU because of lack of bed and/or mechanical ventilators. Additionally, lack of dialysis was the other reason for the delay at the study facility. Cases requiring dialysis were referred out to the hospital located in the capital of the country many kilometers away. Of the total of 12 cases with renal dysfunction and therefore requiring dialysis five were referred out and seven died before such arrangement is made. This supports the WHO hypothesis relating a high case fatality in the hospital as an indicator for the presence of delay in receiving an adequate and proper treatment [
5,
24].
In general, logistic related problems like: lack of blood products, scarcity of ICU beds, unavailability of dialysis, laboratory reagents, and drugs are responsible for most of SMO with delay three (in JUTH) in our study. This is in line with the finding of the recent systemic review of delay-three, which highlighted its importance in developing nations. In that review, professional (inadequate training/skills mix (86%); staff shortages (60%); low staff motivation (44%) and logistic problems like drug procurement/ logistics problems (65%); lack of equipment (51%)) were identified as the most common barriers [
24,
37]. In contrast, professional factors were less prominent in the case of delay three in JUTH. This can be explained by the fact that; the study is conducted in a teaching hospital with different level professionals and strict morning meeting discussion and feedback system.
The tendency of seeking reproductive health services, including family planning and ANC by the woman is affected by different socio-demographic factors and cultural barriers. Uneducated women are less likely to use the services. They may underestimate the value of institutional delivery or their decision power is limited in the male dominant community. As such their health seeking behavior will be limited (delay one) [
24,
40,
41]. The fact that only 10% of deliveries in Ethiopia occurred in a health facility is clear evidence for this. In addition, more than 50% of women have no any formal education reported in Ethiopia [
38] similar to the finding from our study and with other studies done elsewhere in the country [
14,
31], but lower than that of other developing countries [
10].
In our study, the tendency of booking for ANC seems to be high, but the proportion of cases with an optimal number of ANC visits was low and it is lower than the finding from other developing countries [
10,
18,
23]. This difference could be because of the difference in education status. Similar to studies in other part of the country [
9,
14,
31] and other developing countries [
10,
18] residential area, educational status and occupation, optimal number of ANC visits, state of pregnancy at the time of the first detection of complication, mode of delivery index pregnancy and duration of hospitalization were found to associated with SMO in our study. Findings, which have significant implications for interventions in our study include: an adequate number of ANC and delivery by C/S being protective of SMO, similar to the findings from earlier studies [
40].
Even if C/S is believed to increase the risk of developing SMM or MNM because of the associated increased risk of infection and hemorrhage, the rate of C/S is expected to be higher among cases of SMO for the obvious reasons [
10,
24,
42]. The rate C/S among cases of SMO was high in our study similar to the study in other hospitals elsewhere in the country [
9], and other developing countries [
13]. In our study, C/S was found to be protective of SMO as the case study done in China [
33]. This is, however, different from findings in studies from other developing countries which have found increased risk of SMO in women who underwent C/S [
10,
24]. The higher incidence of uterine rupture (24.7%) in this study is higher than the report in those studies and can possibly account for the difference. In adition, the populations studied, the setting and delays identified at different level can also explain this difference.
This study, undertaken near the end of MDG and eve of SDG, showed an alarm to the situation of maternal health care in the study facility and hence the referral facilities and the population it serves. The high prevalence of delay in receiving an appropriate care after reaching the facility needs an urgent solution. We have witnessed cases of pregnant mothers being admitted and followed for labour in the referral facility whereas, these cases were beyond the capacity of those facilities to manage such cases like: the case of transverse lie being admitted and followed in labour for more than 24 h and later referred with uterine rupture, a mother with two C/S scars referred with uterine rupture after 2 months stay at maternity waiting homes of those health facilities and a case of cord prolapse where the cord was clamped, cut and legated before transfer, among others. The fact that more than half the cases with severe complication lacking formal education and the delay in seeking care needs to be addressed by the government and the health system. We feel that the finding of this study may pass a strong message to all parties who work to end preventable maternal deaths at the study area. The preventive effect of the ANC and C/S, mentioned above, for SMO can only be used by the majority only if this system is in place and close to the community. In addition, the significant contribution of delay three to MNM and maternal death calls for action.