BioMed Research International

BioMed Research International / 2019 / Article

Research Article | Open Access

Volume 2019 |Article ID 2594343 | 11 pages |

Skin Color and Severe Maternal Outcomes: Evidence from the Brazilian Network for Surveillance of Severe Maternal Morbidity

Academic Editor: Maria Barbolina
Received24 Feb 2019
Revised28 May 2019
Accepted15 Jul 2019
Published30 Jul 2019


Background. Taking into account the probable role that race/skin color may have for determining outcomes in maternal health, the objective of this study was to assess whether maternal race/skin color is a predictor of severe maternal morbidity. Methods. This is a secondary analysis of the Brazilian Network for Surveillance of Severe Maternal Morbidity, a national multicenter cross-sectional study of 27 Brazilian referral maternity hospitals. A prospective surveillance was performed to identify cases of maternal death (MD), maternal near miss (MNM) events, and potentially life-threatening conditions (PLTC), according to standard WHO definition and criteria. Among 9,555 women with severe maternal morbidity, data on race/skin color was available for 7,139 women, who were further divided into two groups: 4,108 nonwhite women (2,253 black and 1,855 from other races/skin color) and 3,031 white women. Indicators of severe maternal morbidity according to WHO definition are shown by skin color group. Adjusted Prevalence Ratios ( - 95%CI) for Severe Maternal Outcome (SMO=MNM+MD) were estimated according to sociodemographic/obstetric characteristics, pregnancy outcomes, and perinatal results considering race. Results. Among 7,139 women with severe maternal morbidity evaluated, 90.5% were classified as PLTC, 8.5% as MNM, and 1.6% as MD. There was a significantly higher prevalence of MNM and MD among white women. MNMR (maternal near miss ratio) was 9.37 per thousand live births (LB). SMOR (severe maternal outcome ratio) was 11.08 per 1000 LB, and MMR (maternal mortality ratio) was 170.4 per 100,000 LB. Maternal mortality to maternal near miss ratio was 1 to 5.2, irrespective of maternal skin color. Hypertension, the main cause of maternal complications, affected mostly nonwhite women. Hemorrhage, the second more common cause of maternal complication, predominated among white women. Nonwhite skin color was associated with a reduced risk of SMO in multivariate analysis. Conclusion. Nonwhite skin color was associated with a lower risk for severe maternal outcomes. This result could be due to confounding factors linked to a high rate of Brazilian miscegenation.

1. Background

Significant progress in reducing maternal mortality has been achieved in the past 15 years, in addition to a growing awareness about the burden of severe maternal morbidity. Standard definitions of potentially life-threatening conditions (PLTC) and maternal near miss (MNM) events have enabled our understanding of different conditions and delays related to the quality of maternal health (Box 1 provides the full list of WHO criteria for both conditions) [15].

The identification of risk factors can be effective in providing adequate prevention and surveillance of women with severe maternal morbidity, allowing early diagnosis and treatment of complications [1]. It is essential that global disparities are acknowledged among high-income and low-income settings to improve health care. Ethnic or racial inequalities also require investigation within nations [6, 7].

In general, white women have better maternal health outcomes [6] while black women usually have higher maternal mortality ratios [8]. It is more likely that black women are younger and lack a partner [6]. Furthermore, these women usually have a lower level of school education and less adequate prenatal care assistance [5, 8]. Recent data on severe maternal morbidity and different races also confirmed that black women have higher chances of developing MNM events [9]. Other minority ethnic groups have been investigated and associated with worse maternal outcomes within nations, such as White-Hispanics in New York, USA [10], and indigenous populations worldwide [11]. Not only race/skin color, but also other socioeconomic and demographic factors, preexisting conditions, and pregnancy complications are also associated with poor maternal and perinatal outcomes [12, 13].

Another relevant challenge in studying race/ethnicity lies in the extensive admixture of a population, for example, in a country like Brazil, where this information is clearly difficult to collect and interpret [14]. Therefore, our aim was to explore data on skin color variations and maternal morbidity from the Brazilian Network for Surveillance of Severe Maternal Morbidity study.

2. Methods

This study is a secondary data analysis from the Brazilian Network for Surveillance of Severe Maternal Morbidity, a multicenter cross-sectional study including 27 referral maternity hospitals (private, public, university, nonuniversity) in Brazil that had at least 1000 deliveries annually [15]. Briefly, the study was conducted from July 2009 to June 2010. A prospective surveillance was carried out to identify cases of maternal death (MD), maternal near miss (MNM), and potentially life-threatening conditions (PLTC), according to standard World Health Organization (WHO) definition and criteria (Box 1) [1]. Other particularities of the study methodology and main results have been previously published elsewhere [2, 16].

Sample size was calculated based on a previous pilot study that obtained a MD ratio of 70/100,000 live births and a MNM ratio estimated as maximum 20% of the severe maternal morbidity rate of 42/1000 live births [17]. Thus, the expected prevalence of 8 MNM/1000 births with a 95% confidence interval indicated that a surveillance of 75 thousand deliveries was required to achieve at least 100 MD and 600 MNM.

Data collection was performed by investigators and research assistants allocated to each hospital. Medical charts of all women who fulfilled inclusion criteria were reviewed after hospital discharge or death. If there was any doubt about a diagnosis or procedure, the physician responsible for the case was sought for elucidation. For each case identified, a pretested and coded form was used. The form included 80 items, e.g., data on sociodemographic and economic characteristics; obstetric history; prenatal, childbirth, and postpartum periods; neonatal outcome, complications, and delays in medical care.

Data was entered in the OpenClinica® - version 2.5.5 - (Akaza Research, Waltham, MA, USA) electronic platform. The current article is an analysis of the occurrence of severe maternal morbidity among different ethnic/skin color categories. The study was approved by the National Research Ethics Council (CONEP) and Institutional Review Board of the participating locations. No contact was maintained with participants, only chart review; therefore a waiver of informed consent term was granted. The National Scientific Technological Development Council (CNPq) and the Department of Science and Technology of the Ministry of Health (DECIT) funded the study.

For the current analysis, data on ethnic/race/skin color variations were further explored. In the form, according to information retrieved from the medical chart and self-reported, the skin color was categorized as white, black, indigenous, yellow, and others. In the study, there were 31.7% of white women, 23.6% of black women, 0.2% of indigenous women, 0.3% of yellow women, 19.0 % of others (mainly those from a mixed group of black and white, known as pardo), and 25.3% of missing data. Although the rate of missing data appears to be relatively high, it was distributed equally between the groups and was due to lack of information in the clinical records. Due to the small number of indigenous, yellow, and other ethnicities, we chose to group these women along with the black ethnic group, representing the nonwhite group. Therefore, analysis was further conducted in two groups: whites and nonwhites (black, indigenous, yellow, and others).

First, we demonstrated the indicators of severe maternal morbidity defined by the WHO (Maternal Near Miss Ratio, Severe Maternal Outcomes Ratio, Maternal Near Miss to Maternal Mortality ratio, and Mortality Index) according to ethnic group. Next, we evaluated the distribution of pregnancy outcomes (degrees of maternal morbidity) according to the main causes of obstetric complications and skin color, using χ2 and Fisher’s exact test to evaluate any differences between groups. Furthermore, the risk of severe maternal outcome (SMO = MNM + MD) was estimated by skin color group, according to sociodemographic characteristics, obstetric characteristics, pregnancy outcome, and perinatal results (PR adj - 95%CI). Multiple analysis (Poisson regression) was used to assess conditions independently associated with SMO (MNM or MD) resulting from pregnancies with severe maternal morbidity, including also skin color as a predictor. Data analysis was conducted by using SPSS software version 23 (IBM, Armonk, NY, EUA) and Stata software version 11.2 (StataCorp, College Station, TX, EUA).

3. Results

During the study period, 82,388 women underwent surveillance in 27 participating hospitals, resulting in 82,144 live births. Severe pregnancy-related complications occurred in 9,555 women. Data on skin color was available for 7,139 women (missing data: 25%). All 7,139 women were further divided into major skin color groups: white, black, and others, according to the severity of morbidity (Figure 1). The majority of cases were described as white (42.5%), followed by black (31.6%) and others (25.9%). For further analysis of skin color variations, two groups were considered: a group of white women and a group of nonwhite women (black + others). Considering the continuum of morbidity, there were significant differences among the two groups, with a significantly higher proportion of MNM and MD among white women (Table 1).


9,5559.37/1000 LB11.08 /1000 LB

With information on ethnicity/skin color6,415

TotalMaternal near miss: mortality ratio: MNM: 1 MD = 5.2: 1
 Non-whiteMaternal near miss: mortality ratio: MNM: 1 MD =5.3: 1
 WhiteMaternal near miss: mortality ratio: MNM: 1 MD = 5.2: 1

TotalMortality index: MI = MD/(MNM+MD) = 0.154 = 15.4%
 Non-whiteMortality index: MI = MD/(MNM+MD) = 0.159 = 15.9%
 WhiteMortality index: MI = MD/(MNM+MD) = 0.160 = 16.0%

TotalMaternal mortality ratio: MMR= MD/LB X100.000 = 170.4/100 000 LB

LB: 82,144; Deliveries: 82,388; a: indicators cannot be estimated for categories of skin color because this information was not available for the total number of livebirths; b: p=0.044 (white X non-white)
PLTC: potentially life-threatening condition; MNM: maternal near miss; MD: maternal death; MI: Mortality Index is the proportion of women with near miss who died; SMOR: Severe Maternal Outcome Ratio is the proportion of all women delivering a live newborn who had a maternal near miss event or died

Allowing for additional recommended health indicators for maternal morbidity and mortality, there were no statistical differences in the maternal near miss to mortality ratio and mortality index between groups (Table 1).

The distribution of pregnancy outcomes according to major causes of obstetric complications was also evaluated by skin color. It revealed that the nonwhite group had a significantly higher proportion of MD due to hypertension, infection, and combined complications than the white group, despite the lower number of PLTC and MNM cases. For hypertension and clinical/surgical complications (which included the burden of the H1N1 influenza pandemic that occurred during study period), in the continuum of severity total numbers were higher among black women and other minority ethnic groups, with a similar or increased proportion of MD among white women (Table 2).

n (%)
n (%)
n (%)

Hemorrhage 1532 (23.7)253 (41.6)29 (25.0)0.009
 Non-white676 (44.1)112 (44.3)14 (48.3)<0.001
 White856 (55.9)141 (55.7)15 (51.7)<0.001
Hypertension 4550 (70.4)263 (43.2)30 (25.9)<0.001
 Non-white2927 (64.3)141 (53.6)16 (53.3)<0.001
 White1623 (35.7)122 (46.4)14 (46.7)<0.001
Infection47 (0.7)36 (5.9)5 (4.3)<0.001
 Non-white19 (40.4)24 (66.7)4 (80.0)<0.001
 White28 (59.6)12 (33.3)1 (20.0)<0.001
Clinical /Surgical605 (9.3)151 (24.8)58 (50.0)<0.001
 Non-white310 (51.2)64 (42.4)25 (43.1)<0.001
 White295 (48.8)87 (57.6)33 (56.9)<0.001
More than one319 (4.9)92 (15.1)4 (3.4)<0.001
 Non-white168 (52.7)50 (54.3)3 (75.0)<0.001
 White151 (47.3)42 (45.7)1 (25.0)<0.001

PLTC: potentially life-threatening condition; MNM: maternal near miss; MD: maternal death
They are not mutually exclusive; the sum of categories for each group can be higher than a 100% 
p-values comparing the proportions among groups adjusted for the effect of cluster design using χ2 or exact tests
P-values in italic mean they are statistically significant

In order to deeply understand factors associated with a worse maternal outcome within the continuum of severity, we further compared cases of SMO (MNM+MD) to cases of PLTC (less severity) among the skin color groups previously defined. This comparison was performed for previous clinical conditions, sociodemographic characteristics, and maternal and perinatal results. Both groups showed significant increases in the prevalence of SMO cases among previous cardiac and thyroid diseases. Women from the nonwhite group had less previous chronic hypertension and obesity, while there was a higher incidence of renal disease, sickle cell disease, and HIV/AIDS among SMO cases of nonwhite group. White women with SMO had a higher prevalence of other relevant conditions: low weight, neurologic disease, cancer, and drug addiction (Table 3).

Previous conditionsNon-white womenPWhite womenp

Any of below1677 (48.5)153 (48.9)0.9361393 (54.8)181 (52.8)0.592
Chronic Hypertension646 (18.7)39 (12.5)0.019429 (16.9)51 (14.9)0.491
Obesity922 (26.7)50 (16.0)0.024802 (31.6)49 (14.3)<0.001
Low weight8 (0.2)00.44410 (0.4)6 (1.7)0.014
Diabetes74 (2.1)9 (2.9)0.43467 (2.6)18 (5.2)0.076
Smoking175 (5.1)22 (7.0)0.340198 (7.8)28 (8.2)0.846
Cardiac disease89 (2.6)17 (5.4)0.02569 (2.7)18 (5.2)0.030
Respiratory disease72 (2.1)10 (3.2)0.206106 (4.2)20 5.8)0.166
Renal diseases30 (0.9)12 (3.8)<0.00143 (1.7)10 (2.9)0.226
Sickle cell disease26 (0.8)10 (3.2)<0.00114 (0.6)2 (0.6)0.954
HIV/AIDS35 (1.0)6 (1.9)0.01632 (1.3)8 (2.3)0.073
Thyroid diseases29 (0.8)8 (2.6)0.00945 (1.8)12 (3.5)0.028
Neurologic diseases34 (1.0)4 (1.3)0.65331 (1.2)8 (2.3)0.037
Collagenoses14 (0.4)3 (1.0)0.11820 (0.8)6 (1.7)0.098
Cancer7 (0.2)2 (0.6)0.1469 (0.4)6 (1.7)0.002
Drug addiction38 (1.1)8 (2.6)0.09627 (1.1)15 (4.4)<0.001
Others (not specified)132 (3.8)37 (11.8)<0.001149 (5.9)49 (14.3)<0.001

PLTC: potentially life-threatening condition; SMO: severe maternal outcomes; MNM: maternal near miss; MD: maternal death; p-values comparing PLTC and SMO groups adjusted for the effect of the cluster design using χ2 or exact tests; P-values in italic mean they are statistically significant

The risks of SMO were estimated, according to sociodemographic characteristics of women for each skin color group, showing that increased maternal age was the most significant factor associated with severity in both groups. Among the nonwhite group, the lack of a partner was protective. In white women, a higher education was a protection from severity of morbidity. Overweight and obesity reduced the risk of SMO, irrespective of skin color (Table 4). Furthermore, the risks of SMO were estimated according to obstetric characteristics of women for each group, revealing that increased risk occurred among multiparous women (especially parity ≥ 3 in the nonwhite group), inadequate antenatal care, preterm hospital admission, or presence of delays (Table 5) for both groups.

Characteristics Non-white women White Women

Age (years)
 10-19507160.82 0.60-1.12]564370.98 0.77-1.24]
 30-3911510981.19 0.90-1.57]1308051.20 0.94-1.53]
 40-49241641.60 [1.10-2.33]251211.48 [1.04-2.11]
Marital status
 With partner1931849123223801
 Without8715850.55 [0.36-0.83]11610460.69 0.44-1.10]
 No/Primary11815260.92 0.36-2.36]1318720.81 0.47-1.37]
 High9213340.83 0.35-1.93]11010680.58 [0.36-0.92]
 Low weight222510.80 0.44-1.44]572451.38 0.91-2.08]
 Overweight284750.55 [0.34-0.90]344270.54 [0.40-0.73]
 Obesity265430.45 [0.26-0.78]315080.42 [0.27-0.65]

PLTC: potentially life-threatening condition; MNM: maternal near miss; MD: maternal death; = prevalence ratio adjusted for the effect of the cluster design
Missing information for: 394+257, 897+628, 2279+1175 cases; PR in italic mean they are statistically significant

Characteristics Non-white women White Women

Coverage for PN
 Public27429320.85 0.28-2.60]27720300.63 0.39-1.03]
 Private4650.58 0.15-2.24]7720.47 0.15-1.42]
 Social security4361281201
 No PN care211920.99 0.34-2.82]251310.85 0.31-2.34]
 1-214913851.73 [1.35-2.21]16510421.24 [1.07-1.45]
 ≥3795112.38 [1.71-3.32]553241.32 [0.88-1.97]
Prenatal care (visits)
 No282571.86 [1.18-2.92]361532.31 [1.15-4.65]
 1-512212431.69 [1.35-2.10]1127771.53 [1.20-1.96]
 6 or more931663112313701
Gestational age at termination of pregnancy
 < 22 weeks261774.06 [1.83-8.99]421653.67 [1.80-7.48]
 22-27321695.04 [2.88-8.83]511644.29 [2.45-7.50]
 28-33846843.46 [2.42-4.95]825002.55 [1.64-3.96]
 34-36607542.33 [1.43-3.80]624622.14 [1.42-3.22]
 ≥ 3758177917112121
 Still pregnant7411412.47 [8.05-19.30]58867.28 [4.02-13.17]
 With any delay23719611.97 [1.45-2.69]24011122.25 [1.69-3.00]
 No delay931609112414471

Missing information for: 580+341; 12+7; 702+460; 97+76; 208+108 cases
= prevalence ratio adjusted for the effect of the cluster design
PR in italic mean they are statistically significant

For characteristics of delivery and perinatal outcomes, results were consistently worse (increased risk of severity) among the nonwhite group. There was an overall increased higher risk of preterm deliveries, 5-minute Apgar scores below 7, low birthweight, stillbirth, and neonates with intensive care unit admission or transfer among women with SMO among white compared to nonwhite women. The nonwhite group also had a higher risk of neonatal death among SMO cases (Table 6).

Characteristics Non-white women White Women

GA at delivery
 Preterm (<37 weeks)17614482.67 [1.93-3.71]19010172.12 [1.54-2.93]
 Still pregnant282062.95 [1.76-4.97]432062.33 [1.24-4.38]
Mode of pregnancy termination
 Vaginal birth708051637491
 Cesarean section22025540.99 0.64-1.53]23815401.73 0.88-3.37]
 Abortion/ectopic241861.43 0.56-3.64]321522.24 0.79-6.34]
 Still pregnant282061.50 0.90-2.50]432082.21 0.88-5.55]
Apgar score min
 <727874.71 [3.28-6.77]31803.22 [2.24-4.63]
Birth weight
 <2.500 g13612272.33 [1.69-3.22]1488411.99 [1.38-2.87]
 ≥2.500 g922058111414031
Neonatal condition at birth
 Live birth2043212124922051
 Still birth531075.55 [3.88-7.94]34753.07 [1.88-5.04]
Neonatal outcomes
 Admitted or transferred726282.24 [1.56-3.20]795381.49 [1.02-2.17]
 Neonatal death8652.38 [1.52-3.75]14672.01 [0.98-4.15]

Missing information for: 226+132; 15+6; 732+623; 595+525; 532+468; 768+598 cases
= prevalence ratio adjusted for the effect of the cluster design; PR in italic mean they are statistically significant

Multiple regression analysis identified conditions independently associated with SMO. The nonwhite group significantly reduced the risk of severity in these women (Table 7).

Model/ Condition95% CI p

Model [n = 4,981] for Severe Maternal Outcome
Other conditions2.532.09–3.05<0.001
Any delay1.721.44–2.05<0.001
Gestational age at admission (<37 weeks or postpartum)2.831.99–4.01<0.001
Number of previous deliveries (≥1)1.331.10–1.62 0.005
Skin color/ethnicity (non-white)0.610.44–0.85 0.005
Marital status (Without partner)0.560.38–0.83 0.005
Renal diseases1.891.16–3.05 0.012
Neoplasms1.791.09–2.96 0.024
Sickle cell diseases2.011.09–3.69 0.026
Schooling (up to primary)1.221.02–1.47 0.033
Chronic hypertension0.710.51–0.99 0.049

Analysis considering cluster design (center)
Predictors entering the models: age (years); marital status (with partner: 0/ without: 1); schooling (up to primary: 1/ high school or higher: 0); skin color/ethnicity (White: 0/ non-white: 1); financial support for prenatal care (public: 0/ other: 1); number of previous deliveries (0/ ≥1: 1); number of prenatal visits (< 6: 1/ ≥6: 0); gestational age at admission (< 37 weeks or postpartum: 1/ ≥37 sem.: 0); any delay (Yes: 1/ No: 0); previous pathological conditions (Yes: 1/No: 0); chronic hypertension (Yes: 1/No: 0); obesity (Yes: 1/No: 0); low weight (Yes: 1/No: 0); diabetes (Yes: 1/No: 0); smoking (Yes: 1/No: 0); cardiac diseases (Yes: 1/No: 0); respiratory diseases (Yes: 1/No: 0); renal diseases (Yes: 1/No: 0); sickle cell diseases (Yes: 1/No: 0); HIV/Aids (Yes: 1/No: 0); thyroid diseases (Yes: 1/No: 0); neurological diseases/epilepsy (Yes: 1/No: 0); collagenoses (Yes: 1/No: 0); neoplasms (Yes: 1/No: 0); drug addiction (Yes: 1/No: 0); other conditions (Yes: 1/No: 0).
BMI was not included in the models due to the high number of missing values

4. Discussion

The current analysis highlights race/skin color variations among cases of maternal morbidity, associated with severity of morbidity. Overall, skin color was not associated with an increased maternal morbidity. The most significant finding was the higher proportion of maternal deaths among the white group, along with increased overall cases of hemorrhage and infection among white women. This group also had worse results in cases of previous chronic hypertension, renal disease, sickle cell disease, and HIV/AIDS and worse perinatal outcomes, when compared to nonwhite women. Nonwhite skin color reduced the risk of SMO in the multivariate analysis of factors independently associated with increased severity.

Before going ahead with these results and their potential implications for the knowledge on this topic, it is necessary to remember the limitations and difficulties we have in obtaining information on ethnicity/skin color in Brazil, first, because according to national rules, the skin color should be autoreported and it is not necessarily recorded in clinical records from all hospitals. This could be an explanation why we had relatively high missing rates for skin color in this study, although similarly distributed according to the severity of the morbid condition. Second, there was a huge miscegenation in Brazil, which made difficult the classification in a specific category.

According to data from the Center for Disease Control and Prevention (CDC), the nonwhite group had more PLTC, in comparison to the white group. However, white women had more severe complications (MNM and MD), and black women had a 4-fold increased chance of dying from pregnancy-related causes compared to white women [18]. In the current study, the maternal mortality ratio was 170.4/100,000 LB, while according to national vital statistics from the same period, the maternal mortality ratio in Brazil was 72/100,000 LB in 2009 and 68.2/100,000 LB in 2010 for the entire country [19]. The 2009 H1N1 influenza A pandemic occurred at the same time period as our study. Maternity hospitals participating in the study are also referral centers; therefore our study had a higher mortality ratio [20].

Other examples from low-income settings present even higher rates of maternal mortality [3, 4, 7]. A study conducted in a hospital in Nigeria showed a maternal mortality ratio of 1908/100 000 LB, which was 11-fold higher than results found in the current study, and a severe maternal outcome ratio of 218/1000 LB, a value that was 20-fold higher than our findings [21].

Underlying social-demographic conditions and previous clinical comorbidities are likely to influence maternal and perinatal outcomes [5, 8]. Low level of school education is considered a risk factor for SMO [5]. In the current study, white women with complete secondary school education had a lower risk of developing SMO. Multiparous nonwhite women had a higher risk of SMO, in accordance with the literature [2].

Regardless of skin color, women of advanced maternal age had a higher risk of developing severe complications (SMO). Similar reports in the literature have described that maternal age over 40 years increases the odds of complications [2227]. CDC data indicated that black women at advanced maternal age had a higher mortality rate [18]. On the other hand, obesity and overweight were protective factors for SMO.

A previous report on data from Brazil showed that women with underlying disorders had a higher risk of near miss events and nonwhites had more chronic diseases than whites [5]. In our study, the nonwhite group who developed SMO had a lower prevalence of HIV/AIDS and chronic hypertension. In this group, the complication that caused a higher rate of PLTC was mostly hypertension. Nonwhite women have a greater trend towards hypertension diagnosed at the beginning of pregnancy and also of developing preeclampsia [22, 23]. Although white women had more PLTC due to infection, nonwhite women had more severe complications and deaths related to infection, in our sample.

A main concern when considering risk factors for worse outcomes is to identify such conditions and provide adequate care, awareness, and prompt diagnosis of complications, in order to avoid delays. These delays can be organized as related to the delay in deciding to seek care by the individual and/or family (called phase I delay); delay in reaching an adequate health care facility (phase II); and delay in receiving adequate care at the facility (phase III) [15, 28]. The present study shows once more that severe maternal outcome is associated with the occurrence of delays, among nonwhite color and even more among white women.

A study in the United Kingdom showed that starting prenatal care at a later gestational age or no prenatal care may be associated with maternal death in ethnic minority groups [29]. In contrast, a study in Holland did not show any association between late prenatal care or no prenatal care and maternal death [30]. Another study from the United Kingdom showed that pregnancy in black women is diagnosed later, delaying initiation of prenatal care and consequently delaying access to prenatal care [31]. Nevertheless, in our study the lack of prenatal care was a risk factor for SMO among whites. Regardless of ethnicity, the quality of prenatal care is fundamental. On the other hand, women who have many medical prenatal visits usually have some pathological condition or risk justifying this excessive number [5]. Quality assessment of prenatal care is not easy, although there is a suspicion that it may be directly influenced by the social class and ethnic/skin color group of the health service user.

Considering perinatal outcomes, nonwhite group who developed SMO had a higher risk of delivering preterm infants. Furthermore, SMO was associated with a higher occurrence of fetal death, 5-minute Apgar scores lower than 7, low birthweight (<2.500g), need for hospital admission, or transfer of the newborn infant or neonatal death, and these risk estimates were slightly higher for nonwhite than for white women. Two previous Brazilian studies have identified that black women have a higher incidence of low birthweight infants [5, 19].

This network study may have some limitations in determining skin color, since data was captured from medical charts. Information may have been extracted by woman self-report or defined by the medical provider of obstetric care or even the hospital personnel who registered obstetric care management. Furthermore, due to miscegenation in the country, some characteristics that are specific to the black people may be missing, as we observed in some studies conducted in countries with a lower proportion of miscegenation.

In our study, women from the white group were the most affected by SMO, a result that is surprising taking into account all the previously available evidence. We tried to explore the possible interrelationships with other variables that could explain these results, including the difficulty of classifying race/skin color by the women themselves, what could also be seen as a limitation of the study. However, another important point must be also considered: the data came from a prospective surveillance performed during a one-year period in 27 referral maternity hospitals in the country, mainly university tertiary hospitals. We cannot exclude the possibility that a differential access to these facilities was experienced according to race/ethnicity, with higher system and personal difficulties for nonwhite women having access to these more resourceful facilities in dealing with their pregnancy complications. This reflects the importance of studies addressing different risk factors for worse outcomes in low and middle-income settings. Local characteristics, such as significant miscegenation, need to be considered when studying skin color.


CDC:Center for Disease Control and Prevention
LB:Live births
MD:Maternal death
MI:Mortality Index is the proportion of women with near miss who died
MNM:Maternal near miss
MD:Maternal death
PLTC:Potentially life-threatening condition
:Prevalence ratio adjusted for the effect of the cluster design
SMO:Severe maternal outcomes
SMOR:Severe maternal outcome ratio
WHO:World Health Organization.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Ethical Approval

Research protocol was approved by the Institutional Review Board of the coordinating institution on May 2009 (Document CEP 027/2009). The need for an individual informed consent was waived, considering data were collected only from clinical records after discharge or death of each participating woman.


This manuscript is part of the PhD thesis of Karayna G Fernandes under the tutorial of Jose G Cecatti, presented to the Postgraduate Program of Obstetrics and Gynecology from the University of Campinas, Brazil. The content is solely the responsibility of the authors and does not necessarily represent the official views of CNPq. It did not influence the content of the manuscript.

Conflicts of Interest

All authors declare that they have not conflicts of interest.

Authors’ Contributions

The idea for the study and this specific analytical approach arose in a group discussion among J. G. Cecatti, M. A. Parpinelli, and K. G. Fernandes. Analyses were planned and performed by J. G. Cecatti, S. M. Haddad, M. L. Costa, and K. G. Fernandes. The first version of the manuscript was drafted by K. G. Fernandes and then complemented with suggestions from all the others and mainly J. G. Cecatti and M. L. Costa. All authors contributed to the development of the study protocol and approved the final version of the manuscript.


This study was funded by CNPq/DECIT (The National Research Council and the Department of Science and Technology of the Brazilian Ministry of Health), grant number 402702/2008-5. We would like to thank The Brazilian Network for the Surveillance of Severe Maternal Morbidity study group that also included Rodolfo C Pacagnella, Rodrigo S. Camargo, Vilma Zotareli, Lúcio T. Gurgel, Fernanda G Surita, Iracema M Calderon, Simone P Gonçalves, Olímpio B Moraes Filho, Simone A Carvalho, Francisco E Feitosa, George N Chaves, Ione R Brum, Gloria C Saint’Ynes, Carlos A Menezes, Patricia N Santos, Everardo M Guanabara, Elson J Almeida Jr, Joaquim L Moreira, Maria R Sousa, Frederico A Peret, Liv B Paula, Luiza E Schmaltz, Cleire Pessoni, Leila Katz, Adriana Bione, Antonio C Barbosa Lima, Edilberto A Rocha Filho, Melania M Amorim, Debora F Leite, Ivelyne Radaci, Marilia G Martins, Frederico Barroso, Fernando C Oliveira Jr, Denis J Nascimento, Cláudio S Paiva, Moises D Lima, Djacyr M Freire, Roger D Rohloff, Simone M Rodrigues, Sergio M Costa, Lucia C Pfitscher, Adriana G Luz, Daniela Guimaraes, Gustavo Lobato, Marcos Nakamura-Pereira, Eduardo Cordioli, Alessandra Peterossi, Cynthia D Perez, Jose C Peraçoli, Roberto A Costa, Nelson L Maia Filho, Jacinta P Matias, Silvana M Quintana, Elaine C Moises, Fátima A Lotufo, Luiz E Carvalho, Carla B Andreucci, Elvira A Zanette, Márcia M Aquino, Maria H Ohnuma, Rosiane Mattar, and Felipe F Campanharo.


  1. L. Say, J. P. Souza, and R. C. Pattinson, “Maternal near miss—towards a standard tool for monitoring quality of maternal health care,” Best Practice & Research Clinical Obstetrics & Gynaecology, vol. 23, no. 3, pp. 287–296, 2009. View at: Publisher Site | Google Scholar
  2. J. G. Cecatti, M. L. Costa, S. M. Haddad et al., “Network for Surveillance of Severe Maternal Morbidity: a powerful national collaboration generating data on maternal health outcomes and care,” BJOG: An International Journal of Obstetrics & Gynaecology, vol. 123, no. 6, pp. 946–953, 2016. View at: Publisher Site | Google Scholar
  3. The World Bank, Trends in Maternal Mortality: 1990 to 2015. In: World Health Organization; 2010, p. 1–55,
  4. WHO, WHO Maternal mortality, World Health Organization; 2016,
  5. A. L. Martins, “Near miss and black women,” Saúde e Sociedade, vol. 25, no. 3, pp. 573–588, 2016. View at: Publisher Site | Google Scholar
  6. M. Knight, J. J. Kurinczuk, P. Spark, P. Brocklehurst, and UKOSS, “Inequalities in maternal health: national cohort study of ethnic variation in severe maternal morbidities,” BMJ (Clinical research ed.), vol. 338, article b542, 2009. View at: Google Scholar
  7. M. J. Small, T. K. Allen, and H. L. Brown, “Global disparities in maternal morbidity and mortality,” Seminars in Perinatology, vol. 41, no. 5, pp. 318–322, 2017. View at: Publisher Site | Google Scholar
  8. M. J. Tucker, C. J. Berg, W. M. Callaghan, and J. Hsia, “The black-white disparity in pregnancy-related mortality from 5 conditions: differences in prevalence and case-fatality rates,” American Journal of Public Health, vol. 97, no. 2, pp. 247–251, 2007. View at: Publisher Site | Google Scholar
  9. M. Nair, J. J. Kurinczuk, M. Knight, and V. J. Vitzthum, “Ethnic variations in severe maternal morbidity in the uk– a case control study,” PLoS ONE, vol. 9, no. 4, article e95086, 2014. View at: Publisher Site | Google Scholar
  10. E. A. Howell, N. N. Egorova, T. Janevic, A. Balbierz, J. Zeitlin, and P. L. Hebert, “Severe maternal morbidity among hispanic women in new york city,” Obstetrics & Gynecology, vol. 129, no. 2, pp. 285–294, 2017. View at: Publisher Site | Google Scholar
  11. L. J. Kirmayer and G. Brass, “Addressing global health disparities among Indigenous peoples,” The Lancet, vol. 388, no. 10040, pp. 105-106, 2016. View at: Publisher Site | Google Scholar
  12. A. A. Creanga, C. J. Berg, C. Syverson, K. Seed, F. C. Bruce, and W. M. Callaghan, “Race, ethnicity, and nativity differentials in pregnancy-related mortality in the united states,” Obstetrics & Gynecology, vol. 120, no. 2, Part 1, pp. 261–268, 2012. View at: Publisher Site | Google Scholar
  13. E. A. Howell, N. Egorova, A. Balbierz, J. Zeitlin, and P. L. Hebert, “Black-white differences in severe maternal morbidity and site of care,” American Journal of Obstetrics & Gynecology, vol. 214, no. 1, pp. 122.e1–122.e7, 2016. View at: Publisher Site | Google Scholar
  14. R. V. Santos, G. O. da Silva, and S. Gibbon, “Pharmacogenomics, human genetic diversity and the incorporation and rejection of color/race in Brazil,” BioSocieties, vol. 10, no. 1, pp. 48–69, 2015. View at: Publisher Site | Google Scholar
  15. R. C. Pacagnella, J. G. Cecatti, M. A. Parpinelli et al., “Delays in receiving obstetric care and poor maternal outcomes: results from a national multicentre cross-sectional study,” BMC Pregnancy and Childbirth, vol. 14, no. 1, article 159, 2014. View at: Publisher Site | Google Scholar
  16. S. M. Haddad, J. G. Cecatti, M. A. Parpinelli et al., “From planning to practice: building the national network for the surveillance of severe maternal morbidity,” BMC Public Health, vol. 11, article 283, 2011. View at: Publisher Site | Google Scholar
  17. J. P. Souza, J. G. Cecatti, M. A. Parpinelli, S. J. Serruya, and E. Amaral, “Appropriate criteria for identification of near-miss maternal morbidity in tertiary care facilities: a cross sectional study,” BMC Pregnancy and Childbirth, vol. 7, article 20, 2007. View at: Publisher Site | Google Scholar
  18. J. Chang, L. D. Elam-Evans, C. J. Berg et al., “Pregnancy-related mortality surveillance-United States, 1991-1999,” MMWR. Surveillance Summaries : Morbidity and Mortality Weekly Report. Surveillance Summaries / CDC, vol. 52, no. 2, pp. 1–8, 2003. View at: Google Scholar
  19. DATASUS, Maternal Mortality Ratio, 2017,
  20. L. C. Pfitscher, J. G. Cecatti, R. C. Pacagnella et al., “Severe maternal morbidity due to respiratory disease and impact of 2009 H1N1 influenza A pandemic in Brazil: results from a national multicenter cross-sectional study,” BMC Infectious Diseases, vol. 16, no. 1, article 220, 2016. View at: Publisher Site | Google Scholar
  21. I. I. Mbachu, C. Ezeama, K. Osuagwu, O. S. Umeononihu, C. Obiannika, and N. Ezeama, “A cross sectional study of maternal near miss and mortality at a rural tertiary centre in southern nigeria,” BMC Pregnancy and Childbirth, vol. 17, no. 1, article 251, 2017. View at: Publisher Site | Google Scholar
  22. G. Ghosh, J. Grewal, T. Männistö et al., “Racial/ethnic differences in pregnancy-related hypertensive disease in nulliparous women,” Ethnicity & Disease, vol. 24, no. 3, pp. 283–289, 2014. View at: Google Scholar
  23. V. L. Burt, P. Whelton, E. J. Roccella et al., “Prevalence of hypertension in the us adult population: results from the third national health and nutrition examination survey, 1988-1991,” Hypertension, vol. 25, no. 3, pp. 305–313, 1995. View at: Publisher Site | Google Scholar
  24. M. Laopaiboon, P. Lumbiganon, N. Intarut et al., “Advanced maternal age and pregnancy outcomes: a multicountry assessment,” BJOG: An International Journal of Obstetrics & Gynaecology, vol. 121, no. s1, pp. 49–56, 2014. View at: Publisher Site | Google Scholar
  25. F. Oliveira Jr, M. Costa, J. Cecatti, J. Pinto e Silva, and F. Surita, “Maternal morbidity and near miss associated with maternal age: the innovative approach of the 2006 Brazilian demographic health survey,” Clinics, vol. 68, no. 7, pp. 922–927, 2013. View at: Publisher Site | Google Scholar
  26. J. P. Souza, J. G. Cecatti, A. Faundes et al., “Maternal near miss and maternal death in the World Health Organization's 2005 global survey on maternal and perinatal health,” Bulletin of the World Health Organization, vol. 88, no. 2, pp. 113–119, 2010. View at: Publisher Site | Google Scholar
  27. J. J. Zwart, J. M. Richters, F. Öry, J. I. P. de Vries, K. W. M. Bloemenkamp, and J. van Roosmalen, “Severe maternal morbidity during pregnancy, delivery and puerperium in the Netherlands: a nationwide population-based study of 371 000 pregnancies,” BJOG: An International Journal of Obstetrics & Gynaecology, vol. 115, no. 7, pp. 842–850, 2008. View at: Publisher Site | Google Scholar
  28. S. Thaddeus and D. Maine, “Too far to walk: maternal mortality in context,” Social Science & Medicine, vol. 38, no. 8, pp. 1091–1110, 1994. View at: Publisher Site | Google Scholar
  29. L. Bowyer, The Confidential Enquiry into Maternal and Child Health (CEMACH). Saving Mothers’ Lives: Reviewing Maternal Deaths to Make Childhood Safer—2003-2005, CEMACH, London, UK, 2007. View at: Publisher Site
  30. J. van Roosmalen, “Substandard care in immigrant versus indigenous maternal deaths in The Netherlands,” BJOG: An International Journal of Obstetrics & Gynaecology, vol. 109, no. 2, pp. 212-213, 2002. View at: Publisher Site | Google Scholar
  31. R. Rowe, H. Magee, M. Quigley, P. Heron, J. Askham, and P. Brocklehurst, “Social and ethnic differences in attendance for antenatal care in England,” Public Health, vol. 122, no. 12, pp. 1363–1372, 2008. View at: Publisher Site | Google Scholar

Copyright © 2019 K. G. Fernandes et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

More related articles

730 Views | 281 Downloads | 0 Citations
 PDF  Download Citation  Citation
 Download other formatsMore
 Order printed copiesOrder

Related articles

We are committed to sharing findings related to COVID-19 as quickly and safely as possible. Any author submitting a COVID-19 paper should notify us at to ensure their research is fast-tracked and made available on a preprint server as soon as possible. We will be providing unlimited waivers of publication charges for accepted articles related to COVID-19. Sign up here as a reviewer to help fast-track new submissions.