INTRODUCTION

In Mexico, coronavirus disease 2019 (COVID-19) is currently the leading cause of death in pregnancy, overtaking obstetric hemorrhage and pre-eclampsia1. There are several reasons that could explain the higher rate of mortality among pregnant compared with non-pregnant individuals affected by COVID-19. First, recent studies have demonstrated that pregnancy is an independent risk factor for adverse outcome associated with COVID-19. Studies conducted in developing countries, such as Mexico, have demonstrated that severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-infected pregnant women have higher odds of pneumonia (odds ratio (OR), 1.99 (95% CI, 1.81–2.19)), intensive care unit (ICU) admission (OR, 2.25 (95% CI, 1.86–2.71)) and death (OR, 1.65 (95% CI, 1.30–2.09)), when compared with infected non-pregnant women of the same age2. Similar findings have been demonstrated in developed countries, such as the USA, in which Hispanic/Latin-American pregnant women with SARS-CoV-2 infection have experienced a 2.4-times higher risk of death compared with other ethnic groups3. Second, in the context of any pandemic, there is a strong influence of ethnicity and socioeconomic status on health-related outcomes among infected individuals4, 5. The higher rate of COVID-19-related mortality among Latin-American pregnant women raises the questions of why this particular population is more susceptible to developing adverse outcome associated with COVID-19, and which are the clinical and socioeconomic risk factors6. It is, therefore, of great importance to unveil all possible prognostic factors associated with COVID-19-related mortality, especially those related to socioeconomic inequality, as the identification of these risk factors could help guide the development of heath policies to protect vulnerable groups, which are defined not only by clinical and demographic criteria but also by socioeconomic indices, such as the poverty and social vulnerability indices.

The objective of this study was to evaluate the association of clinical characteristics and socioeconomic determinants with COVID-19-related mortality and severe morbidity among pregnant women.

METHODS Study design and participants

This study analyzed data from the Mexican National Registry of Coronavirus7, which is an ongoing prospective cohort based on information from the Mexican government and is updated weekly with data from 475 monitoring hospitals located across the 32 states of Mexico. Inclusion criteria were all pregnant women with a positive reverse-transcription quantitative polymerase chain reaction (RT-qPCR) test for SARS-CoV-2 in the Mexican National Registry of Coronavirus between 1 April 2020 and 31 July 2021. Maternal deaths due to other causes were excluded from the study population. The study protocol was approved by the General Hospital of Mexico (Dr Eduardo Liceaga), Mexico City, Mexico, under the ethics committee number CE/23020.

Data collection

Data on patients' medical history were collected and transferred from the Mexican National Registry of Coronavirus. Access to the Mexican National Registry is available at each hospital only by institutional approval. The following data were collected for each patient: age; pregestational diabetes mellitus; chronic obstructive pulmonary disease; asthma; immunosuppression; chronic hypertension; cardiovascular disease; obesity (defined as body mass index ≥ 30 kg/m2); chronic renal disease or other non-specified morbidity; smoking; pneumonia; ICU admission; and maternal death. To avoid bias due to missing data, we retrieved and analyzed data from the last update of the Mexican National Registry of Coronavirus, which contains complete information on the outcomes at each update. Socioeconomic determinants included ethnicity (including the proportion of women who were indigenous), access to private health services, social security for public health services, social lag indices or social vulnerability indices and the poverty index for Mexican states. Socioeconomic determinants were calculated based on postal code for every participant; information on postal code, town, state and country is part of the COVID-19 National Database.

Outcomes

The primary outcome was maternal death as a direct result of COVID-19, as labeled by the monitoring hospital. Maternal deaths due to other causes were excluded from the study population. A woman with COVID-19 was defined as any symptomatic patient with a positive RT-qPCR result for SARS-CoV-2. Secondary outcomes were severe pneumonia, ICU admission and intubation. Severe pneumonia due to COVID-19 was defined according to the definition of the American Thoracic Society criteria8, which includes either one major criterion (septic shock with need for vasopressors or respiratory failure requiring mechanical ventilation) or three or more minor criteria (respiratory rate ≥ 30 breaths/min, PaO2/FiO2 ratio ≤ 250, multilobar infiltrates, confusion/disorientation, uremia (blood urea nitrogen level ≥ 20 mg/dL), leukopenia (white blood cell count < 4000 cells/µL), thrombocytopenia (platelet count < 100 000/µL), hypothermia (core temperature < 36°C) or hypotension requiring aggressive fluid resuscitation). Need for ICU admission was determined using the quick Sequential Organ Failure Assessment (qSOFA) score, with a score of ≥ 2 points indicating need for ICU admission9. Viral sepsis was defined according to the Sepsis-3 International Consensus associated with SARS-CoV-2 infection10, 11. Data for severe pneumonia and ICU admission were missing in some cases; therefore, we included only cases with complete data in the calculation of these outcomes.

Social vulnerability index

The social vulnerability index, also known as the social delay index, allows ranking of the states of Mexico from the highest to the lowest degree of social vulnerability at a given moment in time. Using the Dalenius–Hodges stratification method, the social vulnerability index in Mexico may be divided into five categories: very high vulnerability, high vulnerability, medium vulnerability, low vulnerability and very low vulnerability12. It provides a summary of four social deficiencies monitored by the National Council for Evaluation of Social Development Policy13: educational lag, access to health services, access to essential services in housing and quality of and space in housing.

Poverty in Mexican states

The poverty index is based on the CONEVAL methodology14. It considers the current per-capita income, average educational lag in a household, access to healthcare services, access to social security services, quality of and space in housing, access to quality and nutritious food, degree of social cohesion and degree of accessibility to a paved road. The poverty index in Mexico is divided into three categories: not poor, poor and extremely poor. The 2018 poverty report and its methodology were published on 31 July 2019 (MCS-ENIGH 2018 report)15.

Statistical analysis

Descriptive and inferential statistics were used. Quantitative variables were reported as mean ± SD; categorical variables were summarized as n (%). Univariate and multivariate log-binomial regression analyses were performed to establish the association of several clinical, demographic and social risk factors with the primary and secondary outcomes. The treatment effect used for this analysis was relative risk (RR) based on univariate analysis, as well as adjusted RR (aRR) based on multivariate analysis when statistically significant variables were identified on univariate analysis, including maternal age, pre-existing diabetes, obesity, hypertension, renal chronic disease, asthma and ethnicity.

Statistical analyses were performed using Stata version 16 (Stata Corp., College Station, TX, USA). A P-value < 0.05 was considered statistically significant.

RESULTS Description of cohort and characteristics of study population

A total of 13 062 consecutive SARS-CoV-2-positive pregnant women were included in the analysis. Mean age at diagnosis was 28.3 ± 6.0 years. Of those women, 176 (1.35%) died as a direct result of COVID-19, 1191 (9.12%) were diagnosed with severe pneumonia, 322 (2.47%) were admitted to the ICU and 185 (1.42%) were intubated.

On univariate analysis, pregnant women who died due to COVID-19, compared with those who did not die, were older and had higher rates of comorbidities, such as pre-existing diabetes, chronic hypertension, obesity, chronic renal disease and asthma (Table 1). In addition, women who died, compared with those who did not die, had a significantly higher rate of very high and high social vulnerability and of poverty and extreme poverty, while the rate of very low social vulnerability was lower in those who died.

Table 1. Characteristics of the study population of 13 062 SARS-CoV-2-positive pregnant women, according to whether maternal death due to COVID-19 occurred Characteristic Maternal survival (n = 12 886) Maternal death (n = 176) P Maternal age (years) 28.3 ± 6.00 31.7 ± 6.63 < 0.001 Maternal age < 0.001 < 35 years 10 699 (83.03) 106 (60.23) 35–39 years 1745 (13.54) 47 (26.70) ≥ 40 years 442 (3.43) 23 (13.07) Pre-existing diabetes 391 (3.03) 24 (13.64) < 0.001 Chronic hypertension 345 (2.68) 19 (10.80) < 0.001 Obesity 986 (7.65) 30 (17.05) < 0.001 Chronic renal disease 32 (0.25) 4 (2.27) < 0.001 Asthma 268 (2.08) 8 (4.55) 0.022 COPD 18 (0.14) 0 (0) 0.623 Immunosuppression 93 (0.72) 2 (1.14) 0.511 Cardiovascular disease 50 (0.39) 1 (0.57) 0.696 Smoker 225 (1.75) 3 (1.70) 0.74 Indigenous ethnicity 330 (2.56) 5 (2.84) 0.815 Type of HS used 0.450 Private 295 (2.29) 4 (2.27) Public (SW) 5247 (40.72) 65 (36.93) Public (not SW) 7344 (56.99) 107 (60.80) Social vulnerability Very high 1073 (8.33) 27 (15.34) 0.001 High 2021 (15.68) 37 (21.02) 0.043 Medium 2026 (15.72) 20 (11.36) 0.114 Low 4699 (36.47) 69 (39.20) 0.454 Very low 3067 (23.80) 23 (13.07) 0.001 Poverty Not poor 5431 (42.15) 45 (25.57) 0.001 Poor 5226 (40.56) 82 (46.59) 0.042 Extremely poor 2229 (17.30) 49 (27.84) 0.001 Data are given as mean ± SD or n (%). COPD, chronic obstructive pulmonary disease; HS, health service; SW, state worker. Risk factors for maternal death

Risk factors associated with maternal death due to COVID-19 are shown in Table 2. Maternal age, as either a continuous (aRR, 1.08 (95% CI, 1.05–1.10)) or categorical variable, was associated with maternal death; women aged 35–39 years (aRR, 3.16 (95% CI, 2.34–4.26)) or 40 years or older (aRR, 4.07 (95% CI, 2.65–6.25)) had a higher risk for mortality, as compared with those aged < 35 years. Other risk factors associated with maternal mortality were pre-existing diabetes, chronic hypertension and obesity. Chronic renal disease and asthma were not significant on multivariate analysis.

Table 2. Risk factors for COVID-19-related maternal death and severe disease among 13 062 SARS-CoV-2-positive pregnant women Maternal death (n = 176) Severe pneumonia (n = 1191) ICU admission (n = 322) Intubation (n = 185) Risk factor aRR (95% CI) P aRR (95% CI) P aRR (95% CI) P aRR (95% CI) P MA (in years) 1.08 (1.05–1.10) < 0.001 1.03 (1.02–1.05) < 0.001 1.02 (1.01–1.04) 0.005 1.06 (1.04–1.09) < 0.0001 < 35 years Reference — Reference — Reference — Reference — 35–39 years 3.16 (2.34–4.26) < 0.001 1.57 (1.39–1.77) < 0.001 1.35 (1.07–1.69) 0.012 1.98 (1.49–2.65) < 0.0001 ≥ 40 years 4.07 (2.65–6.25) < 0.001 1.79 (1.44–2.23) < 0.001 1.17 (0.74–1.84) 0.505 2.26 (1.43–3.55) < 0.0001 Pre-existing diabetes 2.66 (1.65–4.27) < 0.001 1.35 (1.07–1.69) 0.011 1.08 (0.71–1.74) 0.707 0.71 (0.37–1.35) 0.297 Chronic hypertension 1.75 (1.02–3.00) 0.042 1.74 (1.39–2.17) < 0.001 0.99 (0.62–1.58) 0.973 1.05 (0.58–1.91) 0.880 Obesity 2.15 (1.46–3.17) < 0.0001 1.35 (1.14–1.59) < 0.001 1.17 (0.85–1.61) 0.321 1.37 (0.92–2.04) 0.122 Chronic renal disease 1.59 (0.52–4.79) 0.414 1.86 (0.89–3.82) 0.699 0.57 (0.08–3.77) 0.556 1.15 (0.17–7.79) 0.785 Asthma 1.62 (0.74–3.55) 0.228 1.16 (0.82–1.62) 0.400 1.00 (0.52–1.94) 0.998 1.23 (0.56–2.71) 0.600 COPD 1 — 1 — 1 — 1 — Immunosuppression 1.56 (0.39–6.12) 0.522 1.81 (1.15–2.84) 0.011 2.01 (1.03–3.94) 0.040 2.57 (1.13–5.84) 0.024 Cardiovascular disease 1.42 (0.21–9.83) 0.721 0.86 (0.33–2.19) 0.744 1.73 (0.49–6.13) 0.396 1.62 (0.26–9.94) 0.599 Smoker 0.93 (0.30–2.89) 0.905 0.94 (0.62–1.44) 0.791 1.08 (0.51–2.29) 0.842 1.21 (0.47–3.12) 0.684 Indigenous ethnicity 0.86 (0.35–2.10) 0.740 1.21 (0.91–1.62) 0.192 0.55 (0.25–1.22) 0.144 1.25 (0.59–2.63) 0.559 Using private health services 0.82 (0.30–2.20) 0.695 1.09 (0.78–1.52) 0.608 1.06 (0.62–1.83) 0.825 0.71 (0.30–1.69) 0.447 Using public health services (state worker) 0.82 (0.61–1.12) 0.218 1.04 (0.94–1.17) 0.431 0.78 (0.63–0.96) 0.021 0.91 (0.69–1.21) 0.525 Social vulnerability* Very high 1.88 (1.26–2.80) 0.002 3.00 (2.43–3.69) < 0.001 1.29 (0.89–1.86) 0.165 1.16 (0.70–1.95) 0.342 High 1.49 (1.04–2.13) 0.028 2.62 (2.17–3.18) < 0.001 0.85 (0.58–1.26) 0.440 1.11 (0.68–1.80) 0.673 Medium 0.76 (0.48–1.20) 0.237 1.03 (0.81–1.32) 0.753 0.92 (0.59–1.42) 0.729 0.98 (0.55–1.76) 0.188 Low 1.07 (0.79–1.45) 0.653 2.13 (1.79–2.54) < 0.001 1.40 (1.05–1.86) 0.021 1.40 (0.95–2.07) 0.085 Very low 0.47 (0.30–0.73) 0.001 1 — 1 — 1 — Poverty Not poor Reference — Reference — Reference — Reference — Poor 1.53 (1.09–2.15) 0.014 1.66 (1.46–1.88) < 0.0001 1.17 (0.93–1.47) 0.185 1.08 (0.79–1.47) 0.618 Extremely poor 1.83 (1.32–2.53) < 0.0001 1.63 (1.44–1.84) < 0.0001 0.92 (0.71–1.19) 0.511 1.01 (0.72–1.42) 0.939 Adjusted relative risk (aRR) values were calculated using log-binomial regression adjusted for the following confounders: maternal age (MA), pre-existing diabetes, obesity, hypertension, chronic renal disease, asthma and ethnicity. COPD, chronic obstructive pulmonary disease; ICU, intensive care unit.

Women with very high social vulnerability had an 88% increased risk of death due to COVID-19 (aRR, 1.88 (95% CI, 1.26–2.80)). Women with high social vulnerability also had a higher risk of death (aRR, 1.49 (95% CI, 1.04–2.13)), while women with very low social vulnerability had a 53% decreased risk (aRR, 0.47 (95% CI, 0.30–0.73)). Being poor (aRR, 1.53 (95% CI, 1.09–2.15)) or extremely poor (aRR, 1.83 (95% CI, 1.32–2.53)) were also significant risk factors for maternal death due to COVID-19.

Risk factors for severe pneumonia, intensive care unit admission and intubation

Risk factors associated with severe pneumonia, ICU admission and intubation are shown in Table 2. Maternal age, as either a continuous or categorical variable, pre-existing diabetes, chronic hypertension, immunosuppression and obesity were risk factors for severe pneumonia. Maternal age and immunosuppression were risk factors for ICU admission and intubation.

Women with very high, high or low social vulnerability had a higher risk of severe pneumonia (aRR, 3.00 (95% CI, 2.43–3.69), 2.62 (95% CI, 2.17–3.18) and 2.13 (95% CI, 1.79–2.54), respectively), which reflects that, overall, the risk of severe pneumonia remained stable across social vulnerability groups. Poor and extremely poor pregnant women had a higher risk for severe pneumonia (aRR, 1.66 (95% CI, 1.46–1.88) and 1.63 (95% CI, 1.44–1.84)). Women with low social vulnerability had a higher risk of ICU admission, while women using public health services had a lower risk of being admitted to the ICU. The relationship between socioeconomic determinants and comorbidities is shown in Table S1.

DISCUSSION Main findings

The findings of this study in SARS-CoV-2-positive pregnant women demonstrated that: (i) advanced maternal age, pre-existing diabetes, chronic hypertension and obesity were associated with COVID-19-related maternal mortality; (ii) similar to maternal death, risk factors associated with severe pneumonia were advanced maternal age, pre-existing diabetes, chronic hypertension, immunosuppression and obesity; advanced maternal age and immunosuppression were also risk factors for ICU admission and intubation, while women using public health services had a lower risk of being admitted to the ICU; and (iii) women with high social vulnerability had an increased risk of maternal death and severe pneumonia, while very low social vulnerability was associated with a reduced risk of maternal death, but low social vulnerability was associated with higher risks of severe pneumonia and ICU admission. Similarly, being poor and extremely poor were significant risk factors for maternal death and severe pneumonia.

Comparison with existing literature

It has been demonstrated recently that pregnancy per se constitutes a risk factor for complications in women with SARS-CoV-2 infection during the reproductive period16. Our findings are in agreement with those of previous studies involving non-pregnant and pregnant individuals, in which advanced age, diabetes, chronic hypertension and obesity were found to be risk factors for severe COVID-19 and mortality2, 3, 16. Zambrano et al.3 showed that Latin-American pregnant women living in the USA, who belong to a vulnerable minority group, are at a higher risk of COVID-19-related mortality; therefore, in this study, we explored the interplay between comorbidities and socioeconomic determinants contributing to severe COVID-19 and maternal mortality. Although the possible mechanisms leading to higher maternal mortality related to COVID-19 in developing countries are not clear, some investigators have demonstrated that low socioeconomic status is associated with a higher risk of severe maternal morbidity17, 18. Previous studies have shown that minority women receive disproportionately delayed or inadequate prenatal care19, and a systematic review identified strong evidence for the impact of race, ethnicity, insurance and education on maternal mortality and severe morbidity20.

Measures of socioeconomic disadvantage, such as social vulnerability and poverty indices, are also associated with an increased risk of complications related to SARS-CoV-2 infection21, 22. The significant association of low social vulnerability with a higher risk of ICU admission may be related to socioeconomic status and access to healthcare, since highly vulnerable women may not have access to adequate healthcare and, therefore, their probability of being admitted to the ICU is lower. On the other hand, women with low social vulnerability, who, by definition, have better access to healthcare, have a higher risk of ICU admission, accompanied by a lower risk of death. During the pandemic, this was observed in the health services; public hospitals were overcrowded, while ICU beds were readily available to people who were able to pay for care at a private hospital. Another explanation for the higher risk of intubation and ICU admission in women with low social vulnerability could be the higher incidence of chronic hypertension in this subgroup; on the other hand, despite the higher incidence of obesity in the very high vulnerability group, multivariate logistic regression analysis showed that high social vulnerability as well as obesity are independent predictors of maternal mortality. Our finding that lower socioeconomic status is associated with a higher incidence of SARS-CoV-2-related severe pneumonia and maternal death supports existing evidence that densely populated communities living in poverty have an increased risk of sustained community transmission of various infectious diseases, including SARS-CoV-223, 24. On the other hand, pregnant women with very low social vulnerability had a reduced risk of mortality due to COVID-19. Overall, our findings suggest a possible causal relationship of education, access to health services, basic infrastructure, quality of and space in housing and household assets with mortality in pregnant women with SARS-CoV-2 infection, which could explain why the rate of maternal mortality is higher in developing countries and in minority groups in developed countries who have limited access to health services.

In symptomatic pregnant women with COVID-19 in Mexico, the rate of maternal mortality is 1.5%2. A recent small cohort of 793 patients from Greece, Turkey, the UK and Austria showed a similar 1.3% incidence of death25. However, Zambrano et al.3 described a 0.14% mortality rate in symptomatic pregnant women with SARS-CoV-2 infection in the USA, and Mullins et al.26 found a 0.5% mortality rate in a UK registry and a 0.2% mortality rate in a USA registry, which differ substantially from the rate observed in this study. We speculate that the excess mortality is due to the factors explored in the current study, which are related to lower socioeconomic status, access to healthcare services, housing, education and household assets.

Strengths and limitations

This study reports results based on one of the largest consecutive cohorts of pregnant women with SARS-CoV-2 infection and is the first to demonstrate that comorbidities, including pre-existing diabetes, chronic hypertension, obesity and social determinants, such as poverty and social vulnerability indices, are significant risk factors for COVID-19-related mortality and morbidity during pregnancy. The advantage of a population-based cohort is that it minimizes bias by allowing calculation of real-population estimates in an unselected population. Cohort studies with low numbers or an overselected population tend to overestimate effect size. Our large dataset, including 176 maternal deaths, allowed us to estimate the effect size of each risk factor with a robust CI, thus reducing bias. No other single consecutive cohort of pregnant women with COVID-19 has included such a large number of maternal deaths with sufficient data to calculate robust effect sizes. Another strength of this study is the prospective acquisition of data across the whole country, allowing representative data from a desired population, which is often a limitation of hospital-based cohorts.

A limitation of this study is the amount of missing data on severe pneumonia and ICU admission, which is compensated for by the large number of included participants, allowing us to calculate robust effect sizes for these outcomes. Another limitation is the lack of data on perinatal outcome, such as fetal growth restriction, pre-eclampsia, preterm birth, stillbirth and neonatal death, due to the population-based origin of the information used for the analysis. However, the data included in this study provide sufficient information to allow us to understand the most important risk factors and social determinants associated with the main outcomes related to COVID-19, which are death, severe pneumonia, intubation and ICU admission.

Clinical implications