Skip to main content
Download PDF

Influence of sociodemographic and obstetric factors on maternal mortality in Brazil from 2011 to 2021

BMC Women's Health volume 24, Article number: 84 (2024) Cite this article

Abstract

Background

Obstetric causes are classified as direct (complications of pregnancy, childbirth or the puerperium) or indirect (caused by pregnancy but not directly caused by it). This study aimed to analyze maternal mortality from obstetric causes in Brazil from 2011 to 2021.

Methods

This was an ecological study on mortality and live births. The outcomes were the specific risk of mortality from direct and indirect cause adjustment and death during pregnancy and the puerperium. Binary and multiple linear logistic regressions were used to assess the influence of sociodemographic factors and maternal and child health indicators on maternal mortality and time of death (pregnancy and puerperium).

Results

Regarding mortality during pregnancy and during the puerperium, increased (p = 0.003) and decreased (p = 0.004) mortality over the years, respectively; residing in the northern region was associated with lower (p < 0.05) and greater (p = 0.035) odds; and the Maternal Mortality Committee was the primary and least active source of investigation, respectively (p < 0.0001). The number of deaths from indirect causes increased with age (p < 0.001) and in the northern region (p = 0.011) and decreased in the white (< 0.05) and stable union (0.002) regions. Specifically, for mortality risk, the age group [women aged 15–19 years presented an increase in cesarean section (p < 0.001) was greater than that of women who had < 4 antenatal visits (p < 0.001)], education [women who completed high school (8 to 11 years) was greater when they had < 4 prenatal visits (p = 0.018)], and marital status [unmarried women had more than 4 antenatal visits (p < 0.001); cesarean birth (p = 0.010) and < 4 antenatal visits (p = 0.009) were predictors of marriage; and women in a stable union who had < 4 prenatal visits and live births to teenage mothers (p < 0.001) were predictors]. Women who had no education (p = 0.003), were divorced (p = 0.036), had cesarean deliveries (p < 0.012), or lived in the north or northeast (p < 0.008) had higher indirect specific mortality risk.

Conclusions

Sociodemographic factors and maternal and child health indicators were related to different patterns of obstetric mortality. Obstetric mortality varied by region, marital status, race, delivery, prenatal care, and cause of death.

Peer Review reports

Background

Maternal mortality, defined as death during pregnancy or within 42 days of childbirth, is a global public health problem linked to socioeconomic disparities, the availability of health care, and inadequate obstetric care [1]. The maternal mortality indicator is crucial for guiding maternal health policies, especially those related to the quality of obstetric care [2]. Obstetric causes, which are classified as direct or indirect, contribute significantly to maternal deaths [3]. Direct obstetric maternal mortality occurs during pregnancy, childbirth, or the postpartum period as a result of obstetric complications. Indirect causes of pregnancy can include preexisting pathologies or pathologies that develop during pregnancy and are not directly caused by obstetric factors but are exacerbated by physiological changes [4].

Reducing maternal mortality remains a global challenge, as outlined in the 2000–2015 UN agenda (MDGs) and continued in the 2030 Sustainable Development Goals (SDGs) [5]. Maternal mortality is preventable and represents a violation of fundamental human rights, reflecting health and socioeconomic inequalities. Brazil has implemented policies to reduce maternal mortality since the 1980s, with a decline from 1990 to 1999, stable mortality from 2000 to 2013, and a new decline from 2017 to 2019 [6, 7]. Despite these historical trends, Brazil has not met its MDG targets and is not on track to meet its SDG targets by 2030 [5]. Recent studies link the mismanagement of the COVID-19 crisis to the reversal of declining trends in maternal mortality [8].

Identifying a typical profile associated with maternal mortality risk in Brazil, controlling for variables such as age and education, reveals inequalities. However, Brazil's continental size and diverse realities affect women's health unevenly. It is crucial to address regional inequalities, taking into account sociodemographic specificities, historical context, and cultural nuances [6]. The study of specific factors associated with obstetric causes is essential, as it highlights the importance of maternal and child health indicators. These indicators reflect the quality of obstetric care in Brazil. Addressing this issue comprehensively, considering individual characteristics, the regional context and the Unified Health System (SUS), can help develop more effective strategies to promote maternal health and reduce maternal mortality in the country.

Given this, the alternative hypothesis for the empirical analysis in this study is that sociodemographic factors and maternal-child health indicators predict maternal mortality from obstetric causes, both directly and indirectly, in Brazil. This study aimed to analyze maternal mortality from obstetric causes in Brazil from 2011 to 2021. To achieve this goal, 1) we examined the influence of sociodemographic factors in predicting obstetric mortality during pregnancy, childbirth, and the postpartum period in Brazil from 2011 to 2021; 2) we verified the influence of sociodemographic factors in predicting obstetric mortality (direct and indirect); and 3) we estimated whether sociodemographic variables and maternal and child health indicators predict the specific risk of maternal mortality from obstetric causes in Brazil from 2011 to 2021.

Methods

Design, area, population, and study period

An observational ecological study with descriptive and inferential analysis [9] was conducted in Brazil according to macroregion (North, Northeast, Midwest, Southeast and South) (Fig. 1) according to the current demarcation of the Political-Administrative Division (DPA), which covers an area of 8,510,345.540 square kilometers and is composed of 5,570 municipalities in 27 federal units (UF), 26 of which are states and one is a federal district [10].

Fig. 1
figure 1

States and geographic regions of Brazil

Full size image

The population and study period were composed of women reproductive-aged (WRA) between 10 and 49 years who died maternally due to obstetric causes (direct and indirect) between 2011 and 2021 according to the municipality of residence, federal unit and geographic region. The survey from 2011 was due to the beginning of the consolidation and standardization of the new forms of the death certificate (DO) and the Declaration of Live Birth (DN), with data available for free access on the platform of the Department of Informatics of the Unified Health System of Brazil (DATASUS), and ended in 2021, since it was the maximum period with data availability [11]. It is worth noting that there has been progress in the coverage and quality of data on the causes of WRA deaths [11].

Data collection and sources

The data were obtained from the DO and DN databases of the systems managed by the DATASUS open access platform [11] of the Mortality Information System (MIS) and Live Birth Information System (LBIS) using the Tabwin data transfer application. The microdatasus package was also used by the RStudio team [12] (RStudio: Integrated Development Environment for R) to acquire the dbf database and preprocess and tabulate the microdata files in dbf format. Both systems are important tools for the development of more effective public health and social security policies.

The spatial data for the thematic presentation were obtained from the website of the Brazilian Institute of Geography and Statistics (IBGE) from the Digital Municipal Grids of the Political-Administrative Division of Brazil, according to the current political-administrative structure. The cartographic bases provided are appropriate for the original scale of 1:250,000 used in the project and use the Geodetic Reference System (SIRGAS 2000) in conjunction with the Geographic Coordinate System and UTF-8 text encoding. The data were processed with QGIS Desktop version 3.26.

Eligibility criteria, operational procedures, and variables

The deaths of WRA according to the place of residence of the deceased in the period from 2011 to 2021 were included, according to the type of obstetric cause, direct (coded in the ICD-10 as O00.0 to O08.9, O11 to O23.9, O24.4, O26.0 to O92.7, D39.2, E23.0) and indirect (coded in the ICD-10 as O10.0 to O10.9; O24.0 to O24.3; O24.9, O25, O98.0 to O99.8, A34, B20 to B24). Deaths of women older than 49 years of age and deaths in which the underlying cause did not include ICD-10 obstetric cause values were excluded. Figure 2 shows the eligibility criteria for the study population for testing the study hypotheses.

Fig. 2
figure 2

Data eligibility flow diagram

Full size image

Once the data had been tabulated in the RStudio environment, the exploratory data surveillance analysis stages began to construct dependent and independent variables. A classification for understanding the variables studied is presented in Table 1.

Table 1 Classification of variables
Full size table

The specific mortality risk (SMR) was calculated for Brazil, region, and federative unit and adjusted for WRA characteristics according to age group (10 to 14 years, 15 to 19 years, 20 to 29 years, 30 to 39 years and 40 to 49 years), race/ethnicity (white, black, yellow, brown and indigenous), schooling (no years of study, 1 to 3 years, 4 to 7 years, 8 to 11 years and 12 years and over) and marital status (single, married, widowed, legally separated and cohabiting). Unlike studies that use adjusted mortality ratios [13,14,15] to investigate mortality, the adjusted SMR is a measure used to assess the likelihood of a woman dying from obstetric factors compared to a similar reference population, taking into account the variables studied [2, 3]. The other dependent variables in the study were categorical and dichotomous and included the type of obstetric death (direct or indirect), death during pregnancy (yes or no), or death during the puerperium (yes, within 42 days of delivery or no). It was not possible to obtain this variable in the delivery period because the corresponding values mostly indicated zero data.

The independent variables are the sociodemographic characteristics mentioned above for the SRM calculations, grouped into categories, and the source of death investigated [Maternal Mortality Committee, Home Visit (Family), Medical Record, Medical-Legal Institute, and Death Verification Service]. The other factors are continuous variables calculated from the LBIS information: the proportion of cesarean births out of the total number of births (vaginal or cesarean) by year and place of residence of the mother multiplied by 100; the proportion of live births to mothers who had 3 or fewer antenatal visits divided by the live births to mothers who had 4 or more antenatal visits multiplied by 100; and the number of live births to adolescent mothers (age groups 10 to 19 years) per 1,000 women in these age groups. Each of these variables provides specific and relevant information for the analysis of maternal mortality risk in the context of the study. The proportion of live births with mothers who had an inadequate number of visits (3 or fewer) compared with those who had a more complete follow-up (4 or more) is a sensitive measure for assessing access to and quality of prenatal care, and the number of live births to adolescent mothers is a measure of the prevalence of adolescent pregnancy, which allows us to assess how this specific population may contribute to the risk of maternal mortality from direct obstetric causes.

Statistical analysis

Descriptive statistics were expressed using measures of central tendency, with the mean for parametric data and median for nonparametric data, and measures of dispersion [standard deviation (parametric) or minimum, maximum, and interquartile range (nonparametric)]. The absolute values of maternal deaths from obstetric causes (direct and indirect) and live births by year of birth and by mother's residence (region and UF) were used to construct direct and indirect SRMs adjusted for WRA characteristics of age group, race/ethnicity, marital status, and education. The data were subsequently tabulated in SPSS software version 27.0 (IBM SPSS Statistics for Windows, Version 27.0). The descriptive and inferential statistics were obtained from Armonk, NY, IBM Corp.

Binary logistic regression was used to 1) assess the influence of sociodemographic factors in predicting obstetric mortality (directly and indirectly) and 2) examine the influence of sociodemographic variables on the time of death (during pregnancy and postpartum). We checked the necessary assumptions to ensure the validity and correct interpretation of the results [16]. First, collinearity between the independent variables was diagnosed using the variance inflation factor (VIF), with the criterion that it should have a value less than 10 and a tolerance of less than 0.2. The "backward: conditional" method in SPSS was used to select the variables for the model. This iterative method adds and removes variables from the model, evaluating the quality of the fit at each step. The decision to add or remove variables was based on statistical criteria, such as a p value < 0.05, which indicates acceptance of the hypothesis that the model with the predictor variables provides a significantly different fit than the model without the predictor variables. The quality of the model fit was assessed using log likelihood. We also calculated Nagelkerke's R2, a measure of fit that ranges from 0 to 1 and indicates the proportion of total variation in the dependent variable that is explained by the model. Nagelkerke's R2 is an adjusted version of Cox and Snell's R2, calculated as the ratio of the log-likelihood of the full model to the log-likelihood of the null model. The reference classification for categorical variables was based on the first category for each disposition according to the coding of the data values in the MIS and LBIS databases. Finally, we calculated the confidence intervals for the odds ratios with a significance level α set at 0.05.

Multiple linear regression was used to analyze whether the proportion of cesarean deliveries, the proportion of live births to mothers with 3 or fewer prenatal visits, the number of live births to teen mothers, and the region of WRA residence were able to predict SRM for each adjusted sociodemographic category (age group, race/ethnicity, education, and marital status).

To construct the SRM models for each sociodemographic group, the necessary assumptions for multiple linear regression were met [17]. First, the significance of each model was assessed by evaluating the calculated F-statistic and performing a hypothesis test. The null hypothesis that all regression coefficients (β) are statistically equal to zero (indicating the absence of a regression model) was rejected, indicating the presence of a significant regression model (p < 0.05). We then performed Durbin–Watson analysis to test for autocorrelation in the residuals of the model. The results were within the acceptable range of 1.5 to 2.5, indicating independence of the residuals. Collinearity was diagnosed based on tolerance (values greater than 0.2) and VIF (values less than 10), ensuring that the independent variables were not highly correlated. The absence of outliers was confirmed by analyzing the standardized residuals, with reference values between -3 and +3. Finally, we evaluated the interdependence between the residuals and Cook's distance for each observation. The reference value for the interdependence between the residuals was between 1.5 and 2.5, and for Cook's distance, it was less than 1.

Results

The number of live births in Brazil from 2011 to 2021 was 31,702,562, while the number of maternal deaths from obstetric causes studied during the same period was 15,022, 8,072 of which were attributed to direct causes and 6,950 to indirect causes. In terms of data completeness, for the race/ethnicity variable, approximately 3.42% of the data for direct obstetric causes were not correctly completed, while for the indirect causes dataset, this figure is slightly lower at approximately 3.06%. For marital status, 8.13% for direct causes and 7.53% for indirect causes were not filled in correctly or were ignored. The education variable has the highest percentage of uncorrected data, with 15.35% for direct causes and 14.59% for indirect causes.

Sociodemographic predictors of pregnancy, delivery and postpartum mortality

First, the prediction of total obstetric mortality during pregnancy, childbirth, and the postpartum period was analyzed. As shown in Table 2, a binary logistic regression model was used to test whether sociodemographic factors were predictors of obstetric death during pregnancy, and a significant model was selected [χ2 (25) = 165.943; p < 0.001; R2 Nagelkerke = 0.044]. The odds of obstetric mortality during pregnancy increased over the years (OR: 1.029; 95% CI: 1.009–1.048; p = 0.003). Residence in the northern region was associated with lower odds of mortality during pregnancy than was residence in the central-west (OR: 1.431; 95% CI: 1.085–1.888; p = 0.011), south (OR: 1.414; 95% CI: 1.067–1.874; p = 0.016), or northeast (OR: 1.368, 95% CI: 1.082–1.730; p = 0.009) regions. The Maternal Mortality Committee was the primary source of investigation for obstetric deaths during pregnancy compared to others: Institute of Forensic Medicine (OR: 0.231, 95% CI: 0.158–0.339, p < 0.0001); Death Verification Service (OR: 0.309, 95% CI: 0.203–0.470, p < 0.0001); Home Visit (Family) (OR: 0.587, 95% CI: 0.466–0.739, p < 0.0001); and Medical Record (OR: 0.717, 95% CI: 0.615–0.837, p < 0.0001).

Table 2 Relationships between sociodemographic factors, source of investigation and obstetric deaths during pregnancy
Full size table

The prediction of obstetric mortality during childbirth was not feasible because 91.5% of the cases were omitted and there were no records of deaths during childbirth in Brazil. Subsequently, the significant model was selected [χ2 (25) = 184.501; p < 0.001, R2 Nagelkerke = 0.049] to identify sociodemographic variables predicting mortality in the puerperium up to 42 days after birth (Table 3). The odds of obstetric mortality at the puerperium decreased over the years (OR = 0.973, 95% CI = 0.955–0.991, p = 0.004). Residence in the northern region was more strongly associated with puerperium mortality than was residence in the mid-western region (OR = 0.741, 95% CI = 0.561–0.979, p = 0.035). The Maternal Mortality Committee was the least active source of investigation in the puerperium compared to the Forensic Medical Institute (OR: 4.681, 95% CI: 3.145–6.970, p < 0. 0001), death verification service (OR: 3.192, 95% CI: 2.082–4.892, p < 0.0001), home visit (family) (OR: 1.867, 95% CI: 1.482–2.352, p < 0.0001), and medical records (OR: 1.485, 95% CI: 1.274–1.732, p < 0.0001).

Table 3 Relationships between sociodemographic factors and sources of investigation and obstetric deaths in the puerperium
Full size table

Sociodemographic predictors of mortality from direct vs. indirect obstetric causes

Subsequently, sociodemographic factors predicting death from obstetric causes, examining the differences between direct and indirect causes, were identified by binary logistic regression (Table 4), with a significant model selected [χ2 (25) = 287.285; p < 0.001; R2 Nagelkerke = 0.068]. The odds of death from indirect obstetric causes increased with age (OR: 1.126; 95% CI: 1.106–1.145; p < 0.001). The northern region had greater odds of death from direct obstetric causes than did the southeastern region (OR: 1.336; 95% CI: 1.068–1.670; p = 0.011). Regarding race/ethnicity, compared with white ethnicity, indigenous ethnicity (OR: 0.502; 95% CI: 0.264–0.955; p = 0.036), black ethnicity (OR: 0.764; 95% CI: 0.636–0.918, p = 0.004), and brown ethnicity (OR: 0.823; 95% CI: 0.723–0.937; p = 0.003) were more strongly associated with mortality from direct causes. Regarding marital status, being in a stable union was associated with a lower odds of mortality from indirect obstetric causes than being single (OR = 0.764; 95% CI = 0.642–0.904; p = 0.002).

Table 4 Relationship of sociodemographic factors with mortality from obstetric causes: Differences between direct and indirect causes
Full size table

Multiple linear regression analysis was subsequently performed to identify sociodemographic variables and maternal and child health indicators that directly or indirectly predicted the SMR. The results of the multiple linear regression analyses for each adjusted SMR are presented in Table 5.

Table 5 Impact of maternal-child indicators on specific risk of mortality (SMR) from direct and indirect obstetric causes
Full size table

Sociodemographic factors and maternal and child health indicators as predictors of direct SMR

For the direct SMR by age group, the model was significant [F(2,52) = 31.667; p < 0.001; R = 0.741; R2 = 0.549]. For women aged 15–19 years, an increase in the proportion of cesarean deliveries reduced the risk of death (β = -0.420; t = -3.710; p < 0.001). In addition, mortality was greater among women who had 3 or fewer antenatal visits (beta = 0.416; t = 3.673; p < 0.001).

For the direct SMR by education, the model [F(2,52) = 11.958; p < 0.001; R = 0.561; R2 = 0.315] indicated that women who completed high school (8 to 11 years of education) had higher mortality when they had fewer than 4 prenatal visits (β = 0.452; t = 2.433; p = 0.018).

For the direct SMR by marital status, the model [F(3,51) = 21.853; p < 0.001; R = 0.750; R2 = 0.562] indicated that unmarried women had higher mortality when they had fewer than 4 antenatal visits (β = 0.756; t = 4.782; p < 0.001). In addition, the proportion of cesarean births (β = -0.384; t = -2.660; p = 0.010) and the proportion of live births among women who had 3 or fewer antenatal visits (β = 0.408; t = 2.717; p = 0.009) are predictors of direct SMR in married WRAs [F(3,51) = 18.415; p < 0.001; R = 0.721; R2 = 0.520]. For stable union, the model [F(3,51) = 15.017; p < 0.001; R = 0.685; R2 = 0.469] indicated that the proportion of live births to women who had 3 or fewer prenatal visits (β = 0.872; t = 3.891; p < 0.001) and the number of live births to teenage mothers (β = -0.407; t = -3.737; p < 0.001) were predictors.

Sociodemographic variables and maternal and child health indicators predicting indirect SMR

Women with no education [F(2,52) = 6.558; p = 0.003; R = 0.449; R2 = 0.201], who had more cesarean deliveries (β = 0.803; t = 2.132; p < 0.038), and who were divorced [F(4,50) = 2.784; p = 0.036; R = 0. 425; R2 = 0.182], who had more cesarean deliveries (β = 0.601; t = 2.604; p < 0.012), and who lived in the north and northeast (β = -0.644; t = -2.779; p < 0.008) had higher indirect SMRs.

Discussion

Based on the conclusions of this study, we can confirm this alternative hypothesis since significant associations were found between sociodemographic factors and maternal-infant health indicators and between maternal mortality from obstetric causes, both direct and indirect, in the Brazilian context. This highlights the need for epidemiologic surveillance of adjusted maternal mortality, either through the SMR or the adjusted morbidity rate [2, 3, 13,14,15].

In this study, an increase in obstetric mortality during pregnancy was observed over the years. A previous study highlighted that maternal mortality in Brazil is considered very high, and the country faces several challenges [18], in addition to considering the impact of COVID-19 [19, 20].

Furthermore, although analysis of the temporal trend of increasing obstetric deaths during childbirth over the years has been hampered by incomplete data, the literature points to gaps in care and the need to prevent more common complications, such as hemorrhage [21], hypertension [22], puerperal infection [23], labor mismanagement, uterine rupture, and anemia [24]. However, a decrease in postpartum deaths has been observed despite deficits in human and material resources, low coverage of postpartum and home visits, and a focus on care in the immediate postpartum period. Improvements in this condition in Brazil have been associated with strengthening the physical structure for postpartum care and screening for postpartum depression [25].

Another critical issue to address is the increase in mortality from indirect obstetric causes during the study period, a trend observed globally. To help countries get back on track in reducing preventable maternal deaths and to monitor progress toward sustainable development goals (SDGs), the World Health Organization (WHO) has set five key targets. Globally, maternal mortality decreased by more than one-third between 2000 and 2017 [23]. However, the COVID-19 pandemic has caused significant disruptions to health services, exacerbating these risks, especially for the most vulnerable families [18, 26].

Our findings also highlight regional disparities in the health of Brazilian women. The northern region had a greater probability of death from obstetric causes during the postpartum period and from direct obstetric causes. This may be related to the socioeconomic conditions and access to health services for women in this region, who face greater vulnerability and lower coverage of prenatal and institutional delivery services [25]. Residence in northern Brazil was associated with lower odds of mortality during pregnancy. Divorced women, especially in the North and Northeast regions, were more likely to die from indirect obstetric causes.

The northern region is also notable for encompassing the political-administrative region of the Brazilian Legal Amazon, which, in addition to its megabiodiversity, contains a great sociometabolic complexity that strongly influences the social and cultural aspects of this population [27]. These factors, combined with the globally observed inequalities in reproductive and child health, contribute to the impact of demographic disparities on maternal morbidity and mortality [28,29,30], which is particularly evident in the continental dimension of Brazil. In addition, the northern region was one of the regions most affected by the COVID-19 pandemic, which may have contributed to the increase in postpartum maternal mortality [31].

The findings also highlight the structural racism that affects the health of black and indigenous women, who have historically experienced violations of their reproductive rights and safe motherhood [32]. A higher likelihood of maternal mortality among black women has also been reported in other parts of the world, with studies in North America [33] and Latin America [34] highlighting that racial disparities have a significant impact on maternal mortality, with significantly higher rates among black women than white women. The increased likelihood of mortality from direct obstetric causes resulting from complications of pregnancy, childbirth, and the postpartum period and related to lack of access to and quality of maternal health care among black and indigenous women also highlights the need for increased efforts to address disparities in the SUS. It is important to note the recent pandemic scenario where the maternal mortality rate due to COVID-19 was almost double that of white women among black women [35].

Marital status, a predictor of mortality, also yielded significant results; in general, unmarried women had a greater risk than married or cohabiting women did. This result is consistent with other findings [36, 37]. However, marital status remains a relatively underdiscussed variable in terms of its impact on maternal morbidity and mortality. Several hypotheses may explain the increased vulnerability of unmarried women to specific obstetric mortality. One possibility is that unmarried women may have less access to and lower quality of health services, especially for antenatal care, childbirth and the postpartum period. This may be related to socioeconomic, cultural, and geographical factors that hinder proper and timely monitoring [38]. Another hypothesis is that unmarried women may have less social, financial, and emotional support from their partners, family, and community, which negatively affects their physical and mental health and their ability to cope with obstetric complications [39]. It is therefore essential to recognize marital status as a social determinant of maternal health and to develop intervention strategies that take into account the specificities and needs of unmarried women. Finally, unmarried, married, and cohabiting women with fewer than 4 antenatal care visits had an increased risk of death from direct obstetric causes.

Education, or years of schooling, has also been described in the literature as a determining and structuring factor in maternal mortality. For example, a study in Asia [40] revealed that educational inequality was an important factor in maternal mortality, with significantly greater rates of education among women with lower education levels. In our study, inadequate antenatal care was a predictor of the risk of direct obstetric mortality among women aged 15–19 years and those who completed high school. Cesarean delivery was a predictor of the risk of indirect obstetric mortality among divorced women and women with no years of education but was a protective factor among married women and women aged 15–19 years.

Several suggestions can be made to explain this association, including the possibility that women with low education levels may have less knowledge about the signs and symptoms of obstetric complications, as well as the necessary care during the prenatal, delivery, and postpartum periods. In addition, women with low levels of education may have less ability to communicate and negotiate with healthcare providers, making it difficult to establish a relationship of trust and mutual respect [41, 42].

The impact of maternal and child health indicators on obstetric mortality is also a critical issue that deserves further study. The number of antenatal visits stands out as an indicator of access to and quality of health services, as well as prevention and early detection of obstetric complications. Studies show that a low number of antenatal visits is associated with an increased risk of mortality from direct and indirect obstetric causes in Brazil and other countries [43, 44]. Prenatal care plays a critical role in the early detection of potential complications during pregnancy. Through regular visits, healthcare providers have the opportunity to monitor a pregnant woman's health, screen for risk factors, and implement preventive measures. However, when visits are infrequent or inadequate, these opportunities are missed, leaving women in a vulnerable position [45].

The influence of the cesarean delivery rate on maternal mortality from obstetric causes in Brazil is complex. The results of this study, based on multiple linear regression analysis, showed that cesarean delivery affects different groups of women of childbearing age. Among women aged 15–19 years, cesarean delivery was identified as a protective factor against SRM due to direct obstetric causes. This finding suggests that in early pregnancy, cesarean delivery may be beneficial for reducing direct obstetric complications that threaten the life of the mother [46]. This is an important finding given the frequency of adolescent pregnancies in Brazil. Among married women, cesarean delivery was also found to be a protective factor against mortality from direct obstetric causes, which may be related to easier access to health services and better childbirth preparation among married women, resulting in a lower incidence of severe complications [47]. However, the analysis also revealed a different scenario for the other groups; among divorced women and those with no formal education, cesarean delivery was associated with an increased risk of death from indirect obstetric causes. This raises important concerns about the practice of cesarean delivery among vulnerable women and its indiscriminate use [48].

For the indicators of live births to adolescent mothers and SRM due to direct causes among women in a stable union, we found that having live births at a young age may act as a protective factor against direct obstetric complications among women in a stable union. These findings highlight the complexity of the relationship between maternal age or marital status and maternal mortality. While early motherhood may be considered a risk factor in some circumstances, this study suggested that early motherhood may confer protection against direct obstetric complications when combined with a stable union. It is important to note that these findings should be interpreted with caution given the specific context of the study design [9], where not all situations fit this pattern and additional factors such as access to health care, quality of health services, and other social determinants of health play important roles. Thus, our results suggest that women who had children at younger ages and who maintained stable unions may have had greater opportunities to access maternal health care, education, and social support. It is important to emphasize that health policies should be developed on the basis of sound evidence and take into account the complex interactions between age, marital status and the risk of obstetric mortality. It is also essential to ensure equitable access to quality health care and maternal health education for all women, regardless of age and marital status [1].

Strengths and limitations

The strengths of this study include the use of a comprehensive and reliable database that allows detailed analysis of maternal mortality from obstetric causes over the years. Advanced statistical techniques were used to analyze the relationships between a wide range of sociodemographic factors, including age group, race/ethnicity, region of residence, method of death investigation, marital status, and education level. This approach allowed for a more comprehensive analysis of the determinants of maternal mortality. Based on the findings of this study, maternal mortality from obstetric causes is a complex and multifaceted issue that requires a comprehensive approach. It is essential to consider the specificities of both direct and indirect causes, as well as the needs of women, considering the influence of sociodemographic factors and the quality of maternal and child health care.

It is important, however, to acknowledge their limitations. An inherent challenge of ecological studies is the possibility of committing an "ecological fallacy". This means that associations observed at the population level may not be true at the individual level. Therefore, conclusions about the relationship between sociodemographic variables and maternal mortality may not be directly generalizable to individual women. In addition, ecological studies do not allow for causal inference. The observed associations between sociodemographic variables and maternal mortality do not imply causality. Other confounding factors may affect these associations [49].

Conclusion

This study revealed an increase in obstetric mortality during pregnancy over the years, with inconclusive information available for childbirth and a decreased likelihood of postpartum mortality. The northern region had lower odds of pregnancy-related mortality but higher odds of postpartum mortality than did the other regions. Divorced women, particularly those in the North and Northeast regions, had higher odds of indirect obstetric mortality. Indigenous, black, and brown women had higher odds of direct obstetric mortality than white women did. Stable union was protective against indirect obstetric mortality. Women with fewer than 4 antenatal visits had an increased risk of both direct and indirect obstetric mortality. Inadequate prenatal care predicted direct obstetric mortality among women aged 15–19 years and those with a high school education. Cesarean delivery was a predictor of indirect obstetric mortality in some groups but was a protective factor in others. The importance of studying causes of death, especially the role of the Maternal Mortality Committee, was emphasized. Sociodemographic factors and maternal and child health indicators showed different patterns of obstetric mortality.

Availability of data and materials

The data used for the study are publicly available and can be accessed from the website (opendatasus.saude.gov.br) and Microdatasus in RStudio (check the script in the Supplementary material).

Abbreviations

IBGE:

Brazilian Institute of Geography and Statistics

DO:

Death Certificate

DN:

Declaration of Live Birth

DASNT:

Department of Health Analysis and Surveillance of Non-Communicable Diseases

DATASUS:

Department of Informatics of the Unified Health System of Brazil

UF:

Federal units

LBIS:

Live Birth Information System

MIS:

Mortality Information System

SIRGAS 2000:

Reference Geodetic System

SMR:

Specific Mortality Risk

SDGs:

Sustainable Development Goals

SUS:

Unified Health System

VIF:

Variance Inflation Factor

WRA:

Women of reproductive-age

WHO:

World Health Organization

References

  1. World Health Organization. Strategies toward ending preventable maternal mortality (EPMM). 2015.

    Google Scholar 

  2. Ministério da Saúde, & Ministério da Saúde. Guia de vigilância epidemiológica. caderno 6: Aids, Hepatites virais, Sífilis congênita, Sífilis em gestantes. 2009. p. 1–22.

    Google Scholar 

  3. Brazil. Manual dos comitês de mortalidade materna. (2009). http://bvsms.saude.gov.br/bvs/publicacoes/comites_mortalidade_materna_3ed.pdf. Accessed 25 Sep 2023.

  4. van den Akker T, Nair M, Goedhart M, Schutte J, Schaap T, Knight M. Maternal mortality: direct or indirect has become irrelevant. Lancet Glob Health. 2017;5(12):e1181–2.

    Article  PubMed  Google Scholar 

  5. Motta CT, Moreira MR. O Brasil cumprirá o ODS 3.1 da Agenda 2030? Uma análise sobre a mortalidade materna, de 1996 a 2018. Ciênc Saúde Colet. 2021;26:4397–409.

    Article  Google Scholar 

  6. Leal LF, Malta DC, de Souza MFM, Vasconcelos AMN, Teixeira RA, Veloso GA, et al. Maternal mortality in Brazil, 1990 to 2019: a systematic analysis of the Global Burden of Disease Study 2019. Rev Soc Bras Med Trop. 2022;55:e0279.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Orellana J, Jacques N, Leventhal DGP, Marrero L, Morón-Duarte LS. Excess maternal mortality in Brazil: Regional inequalities and trajectories during the COVID-19 epidemic. PLoS One. 2022;17(10):e0275333.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. do Socorro Candeira Costa M, do Santos Figueiredo FW. Relationship between income inequality, socioeconomic development, vulnerability index, and maternal mortality in Brazil, 2017. BMC Public Health. 2021;21(1):1842.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Aggarwal R, Ranganathan P. Study designs: part 2 – descriptive studies. Perspect Clin Res. 2019;10(1):34–6.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Instituto Brasileiro de Geografia e Estatística. Malha Municipal. 2021. <https://www.ibge.gov.br/geociencias/organizacao-do-territorio/malhas-territoriais/15774-malhas.html?=&t=sobre>. Accessed 21 Sep 2023.

  11. OPENDATASUS. opendatasus.saude.gov.br. Accessed 15 Nov 2023.

  12. de Saldanha RF, Bastos RR, Barcellos C. Microdatasus: pacote para download e pré-processamento de microdados do Departamento de Informática do SUS (DATASUS). Cad Saúde Pública. 2019;16(35):e00032419.

    Article  Google Scholar 

  13. Laurenti R, de Jorge MHPM, Gotlieb SLD. Maternal mortality in Brazilian capitals: some characteristics and estimation of an adjustment factor. Rev Bras Epidemiol. 2004;7(4):449–60. https://doi.org/10.1590/S1415-790X2004000400008.

    Article  Google Scholar 

  14. Luizaga Carolina Terra de Moraes, Gotlieb Sabina Léa Davidson, Jorge Maria Helena Prado de Mello, Laurenti Ruy. Maternal deaths: revision of the correction factor for official data. Epidemiol Serv Health. 2010;19(1): 8–15. Available at: http://scielo.iec.gov.br/scielo.php?script=sci_arttext&pid=S1679-49742010000100002&lng=pt. https://doi.org/10.5123/S1679-49742010000100002. Cited 2024 Jan 09.

  15. da Silva BGC, Lima NP, da Silva SG, Antúnez SF, Seerig LM, Restrepo-Méndez MC, et al. Maternal mortality in Brazil from 2001 to 2012: temporal trends and regional differences. Rev Bras Epidemiol. 2016;19(3):484–93. https://doi.org/10.1590/1980-5497201600030002.

    Article  PubMed  Google Scholar 

  16. Fernandes AAT, Figueiredo Filho DB, da Rocha EC, da Nascimento WS. Read this paper if you want to learn logistic regression. Rev Sociol Polit. 2021;28:006.

    Article  Google Scholar 

  17. Vatcheva KP, Lee M, McCormick JB, Rahbar MH. Multicollinearity in regression analyses conducted in epidemiologic studies. Epidemiology (Sunnyvale). 2016;6(2):227.

    Article  PubMed  Google Scholar 

  18. Ferreira MES, Coutinho RZ, Queiroz BL. Morbimortalidade materna no Brasil e a urgência de um sistema nacional de vigilância do near miss materno. Cad Saúde Pública. 2023;7(39):e00013923.

    Article  Google Scholar 

  19. dos Santos APB, Vicente CR, Cola JP, Tanaka LF, Garbin JRT, Dell’Antonio LS, et al. The impact of COVID-19 on maternal death and fetal death, a cohort study in Brazil. PLoS One. 2023;18(8):e0290343.

    Article  Google Scholar 

  20. Souza ASR, Amorim MMR. Mortalidade materna pela COVID-19 no Brasil. Rev Bras Saude Mater Infant. 2021;24(21):253–6.

    Article  Google Scholar 

  21. Dahlke JD, Mendez-Figueroa H, Maggio L, Hauspurg AK, Sperling JD, Chauhan SP, et al. Prevention and management of postpartum hemorrhage: a comparison of 4 national guidelines. Am J Obstet Gynecol. 2015;213(1):76.e1-76.e10.

    Article  PubMed  Google Scholar 

  22. Bromfield SG, Ma Q, DeVries A, Inglis T, Gordon AS. The association between hypertensive disorders during pregnancy and maternal and neonatal outcomes: a retrospective claims analysis. BMC Pregnancy Childbirth. 2023;23(1):514.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Sahle SG, Weldemariam S, Mehari MA, Abraha TA. Determinants of puerperal sepsis among postpartum mothers in Mekelle city public hospitals, Tigray, Ethiopia, 2021: a case control study. BMC Women’s Health. 2023;23(1):502.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Say L, Chou D, Gemmill A, Tunçalp Ö, Moller AB, Daniels J, et al. Global causes of maternal death: a WHO systematic analysis. Lancet Glob Health. 2014;2(6):e323–33.

    Article  PubMed  Google Scholar 

  25. Baratieri T, Natal S. Ações do programa de puerpério na atenção primária: uma revisão integrativa. Ciênc Saúde Colet. 2019;28(24):4227–38.

    Article  Google Scholar 

  26. Guimarães RM, Reis LGC, de Souza Mendes Gomes MA, Magluta C, de Freitas CM, Portela MC. Tracking excess of maternal deaths associated with COVID-19 in Brazil: a nationwide analysis. BMC Pregnancy Childbirth. 2023;23(1):22.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Codeço CT, Dal’Asta AP, Rorato AC, Lana RM, Neves TC, Andreazzi CS, et al. Epidemiology, biodiversity, and technological trajectories in the Brazilian Amazon: from Malaria to COVID-19. Front Public Health. 2021;9. Available from: https://www.frontiersin.org/articles/10.3389/fpubh.2021.647754. Cited 2022 Oct 21.

  28. Firoz T, Vidler M, Makanga PT, Boene H, Chiaú R, Sevene E, et al. Community perspectives on the determinants of maternal health in rural southern Mozambique: a qualitative study. Reprod Health. 2016;13(2):112.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Chen L, Feng P, Shaver L, Wang Z. Maternal mortality ratio in China from 1990 to 2019: trends, causes and correlations. BMC Public Health. 2021;21(1):1536.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Chinn JJ, Eisenberg E, Dickerson SA, King RB, Chakhtoura N, Lim IAL, et al. Maternal mortality in the United States: research gaps, opportunities, and priorities. Am J Obstet Gynecol. 2020;223(4):486-492.e6.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Costa ML, Souza RT, Pacagnella RC, Bento SF, Ribeiro-do-Valle CC, Luz AG, et al. Facing the COVID-19 pandemic inside maternities in Brazil: A mixed-method study within the REBRACO initiative. PLoS One. 2021;16(7): e0254977.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Suarez-Balcazar Y, Buckingham S, Rusch DB, Charvonia A, Young RI, Lewis RK, et al. Reproductive justice for black, indigenous, women of color: uprooting race and colonialism. Am J Community Psychol. n/a(n/a). Available from: https://onlinelibrary.wiley.com/doi/abs/10.1002/ajcp.12650. Cited 2023 Nov 25.

  33. Noursi S, Saluja B, Richey L. Using the ecological systems theory to understand black/white disparities in maternal morbidity and mortality in the United States. J Racial Ethn Health Disparities. 2021;8(3):661–9.

    Article  PubMed  Google Scholar 

  34. Pícoli RP, de Cazola LHO, Lemos EF. Maternal mortality according to race/skin color in Mato Grosso do Sul, Brazil, from 2010 to 2015. Rev Bras Saude Mater Infant. 2017;17:729–37.

    Article  Google Scholar 

  35. de Souza SD, de Oliveira MM, Andreucci CB, Nakamura-Pereira M, Knobel R, Katz L, et al. Disproportionate impact of Coronavirus disease 2019 (COVID-19) among pregnant and postpartum black women in Brazil through structural racism lens. Clin Infect Dis. 2021;72(11):2068–9.

    Article  Google Scholar 

  36. Rodrigues CB, Thomaz EBAF, Batista RFL, Riggirozzi P, de Moreira DSO, Gonçalves LLM, et al. Prenatal care and human rights: addressing the gap between medical and legal frameworks and the experience of women in Brazil. PLoS One. 2023;18(2):e0281581.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Wang E, Glazer KB, Howell EA, Janevic TM. Social determinants of pregnancy-related mortality and morbidity in the United States: a systematic review. Obstet Gynecol. 2020;135(4):896–915.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Roozbeh N, Nahidi F, Hajiyan S. Barriers related to prenatal care utilization among women. Saudi Med J. 2016;37(12):1319–27.

    Article  PubMed Central  Google Scholar 

  39. Bedaso A, Adams J, Peng W, Sibbritt D. Prevalence and determinants of low social support during pregnancy among Australian women: a community-based cross-sectional study. Reprod Health. 2021;18(1):158.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Bartlett L, LeFevre A, Zimmerman L, Saeedzai SA, Turkmani S, Zabih W, et al. Progress and inequities in maternal mortality in Afghanistan (RAMOS-II): a retrospective observational study. Lancet Glob Health. 2017;5(5):e545–55.

    Article  PubMed  Google Scholar 

  41. Oguntunde O, Nyenwa J, Yusuf F, Dauda DS, Salihu A, Sinai I. Factors associated with the knowledge of obstetric danger signs, and perceptions of the need for obstetric care among married young women in northern Nigeria. Afr J Prim Health Care Fam Med. 2021;13(1):2557.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Andualem EM, Beyene DT, Jemaneh TM, Abose DW. Relationships of nurses with midwives in their working area and Associated Factors at Jimma Medical Center, Jimma, Southwest Ethiopia. SAGE Open Nurs. 2022;25(8):23779608221146664.

    Google Scholar 

  43. Tesfaye G, Loxton D, Chojenta C, Assefa N, Smith R. Magnitude, trends and causes of maternal mortality among reproductive aged women in Kersa health and demographic surveillance system, eastern Ethiopia. BMC Women’s Health. 2018;18(1):198.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Carvalho OMC, Junior ABV, Augusto MCC, Leite ÁJM, Nobre RA, Bessa OAAC, et al. Delays in obstetric care increase the risk of neonatal near-miss morbidity events and death: a case-control study. BMC Pregnancy Childbirth. 2020;20(1):437.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Byerley BM, Haas DM. A systematic overview of the literature regarding group prenatal care for high-risk pregnant women. BMC Pregnancy Childbirth. 2017;17(1):329.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Keag OE, Norman JE, Stock SJ. Long-term risks and benefits associated with cesarean delivery for mother, baby, and subsequent pregnancies: systematic review and meta-analysis. PLoS Med. 2018;15(1):e1002494.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Tekelab T, Chojenta C, Smith R, Loxton D. Factors affecting utilization of antenatal care in Ethiopia: a systematic review and meta-analysis. PLoS One. 2019;14(4):e0214848.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Kumar P, Sharma H. Prevalence and determinants of socioeconomic inequality in cesarean section deliveries in Bangladesh: an analysis of cross-sectional data from Bangladesh Demographic Health Survey, 2017–18. BMC Pregnancy Childbirth. 2023;23(1):492.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Zhang SX, Yang M, Zheng JX, Zhang BQ, Pan CH, Tian LG. Comprehensive strategies and measures to control COVID-19. Infect Dis Poverty. 2022;11(1):67.

    Article  PubMed  PubMed Central  Google Scholar 

  50. Brazil, Conselho Nacional de Saúde. Resolução n° 466, de 12 de dezembro de 2012. Diretrizes e normas regulamentadoras de pesquisas envolvendo seres humanos. Diário Oficial da União [da] República Federativa do Brasil. 2013;150(112 Seção 1):59–62.

Download references

Acknowledgements

We thank PROPESP/UFPA (PAPQ) for promoting the quality of the publications.

Funding

Not applicable. The paper was self-financed.

Author information

Authors and Affiliations

  1. Urogenital System Clinical and Experimental Research Unit, Institute of Health Sciences, Federal University of Pará (UFPA), Belém, PA, 66075-110, Brazil

    Eric Renato Lima Figueiredo, Cibele Nazaré Câmara Rodrigues & João Simão de Melo-Neto

  2. Laboratory of Epidemiology and Geoprocessing of Amazon, State University of Pará (UEPA), Belém, PA, 66113-010, Brazil

    Claudia do Socorro Carvalho Miranda

  3. Laboratory and Observatory in Surveillance and Social Epidemiology, Federal University of the South and Southeast of Pará (Unifesspa), Marabá, PA, 68500-000, Brazil

    Ana Cristina Viana Campos

  4. Ceres Medical School (FACERES), São José Do Rio, Preto, PA, 15090-305, Brazil

    Fabiana de Campos Gomes

Authors
  1. Eric Renato Lima Figueiredo
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Claudia do Socorro Carvalho Miranda
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Ana Cristina Viana Campos
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Fabiana de Campos Gomes
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Cibele Nazaré Câmara Rodrigues
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. João Simão de Melo-Neto
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

ERLF, CSCM, ACVC and JSMN designed the study. ERLF collected the data. ERLF and JSMN analyzed the data. ERLF, CSCM, ACVC, and JSMN interpreted the data. FCG and CNCR drafted the manuscript. ERLF, CSCM, ACVC, FCG, CNCR, and JSMN revised the manuscript. ERLF, CNCR and JSMN accept responsibility for the integrity of the data analysis. All authors agree to be accountable for all the aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The author (s) read and approved the final manuscript.

Corresponding author

Correspondence to João Simão de Melo-Neto.

Ethics declarations

Ethics approval and consent to participate

The data used in this study have unrestricted access and are therefore exempt from the evaluation of the Research Ethics Committee, in accordance with Resolution No. 466 of the National Health Council, dated December 12, 2012 [50].

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lima Figueiredo, E.R., do Socorro Carvalho Miranda, C., Viana Campos, A.C. et al. Influence of sociodemographic and obstetric factors on maternal mortality in Brazil from 2011 to 2021. BMC Women's Health 24, 84 (2024). https://doi.org/10.1186/s12905-024-02925-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12905-024-02925-3

Keywords

BMC Women's Health

ISSN: 1472-6874

Contact us