Social and health system factors associated with maternal mortality in Eastern and Western China: Population health estimates using provincial-level data

Xiaojing Zeng; Dongjian Yang; Shiyang Li; Xiaolin Hua; Yanlin Wang; Jun Zhang; Zhiwei Liu

doi:10.1371/journal.pmed.1004837

. 2025 Dec 4;22(12):e1004837. doi: 10.1371/journal.pmed.1004837

Social and health system factors associated with maternal mortality in Eastern and Western China: Population health estimates using provincial-level data

Xiaojing Zeng ^1,^2,^#, Dongjian Yang ^1,^2,^#, Shiyang Li ³, Xiaolin Hua ¹, Yanlin Wang ^1,², Jun Zhang ^1,^3,^*, Zhiwei Liu ^1,^2,^*

Editor: Margaret E Kruk⁴

PMCID: PMC12677549 PMID: 41343599

Abstract

Background

Globally, maternal mortality is off track in achieving the Sustainable Development Goals by 2030. Over the past two decades, China has dramatically reduced maternal mortality in more developed (eastern) and less developed (western) regions. An understanding of the social and health system factors associated with maternal mortality in China may be helpful for countries attempting to meet the 2030 targets and beyond.

Methods and findings

We analyzed provincial-level data on maternal mortality and social and health system factors from the National Health Statistics Yearbooks and China Statistical Yearbooks from 2004 to 2020. We investigated the factors associated with maternal mortality before and after 2013, the year that a historic national program, Reducing Maternal Mortality and Eliminating Neonatal Tetanus, came to an end. Bayesian kernel machine regression was employed to analyze social and health system factors (urbanization rate, per capita disposable income, average years of schooling, number of health technical personnel in maternal and child healthcare, number of hospital beds for obstetrics and gynecology, local fiscal expenditure on healthcare, prenatal booking rate, antenatal care rate, and hospital delivery rate) as a mixture and identify the factors with larger posterior inclusion probability and a higher value of the exposure–response relationship for the total and cause-specific maternal mortality. In the East, an increase in hospital delivery rate correlated with the decrease in total maternal mortality [posterior mean and standard deviation (SD): −14.8(1.5)] before 2013, and the urbanization rate was negatively associated with total maternal mortality [posterior mean and SD: −3.9(0.6)] after 2013. Hospital delivery, urbanization, local fiscal expenditure on healthcare, and antenatal care were the factors associated with reduced cause-specific maternal mortality in the East. In the West, an increase in antenatal care rate was associated with reduced total maternal mortality, with the posterior mean and SD of −33.8(6.8) and −11.5(4.1) before and after 2013, respectively. Hospital delivery and antenatal care were the factors associated with reduced cause-specific maternal mortality in the West. The main limitation of this study was the data constraints in the national statistics.

Conclusions

Coverage of maternal care, health financing, and urbanization were the factors associated with the substantial reduction in maternal deaths in Eastern and Western China during 2004–2020. The improvement of the quantity and quality of antenatal care and hospital delivery may be a viable policy priority in less developed regions worldwide.

Author summary

Why was this study done?

Maternal mortality worldwide is not decreasing fast enough to reach global targets by 2030.
China has made major progress in reducing maternal deaths in both wealthier and poorer regions over the past two decades.
Learning from China’s experience may help low- and lower-middle-income countries accelerate progress in reducing maternal mortality toward global goals.

What did the researchers do and find?

We employed a Bayesian kernel machine regression model to identify the social and health system factors associated with decreased maternal mortality in China using nationwide provincial-level data from 2004 to 2020.
Hospital delivery rate, antenatal care rate, urbanization rate, local fiscal expenditure on healthcare, and per capita disposable income were associated with improvements in maternal survival in both wealthier and poorer regions in China.

What do these findings mean?

To reduce maternal deaths in less developed regions, it may be vital to increase both the utilization and quality of antenatal care and facility birth, supported by adequate health funding and social improvements.
The study’s main limitation was the unavailability of more direct and comprehensive indicators of non-biomedical factors.

Xiaojing Zeng and colleagues, in a population health estimates analysis, explore the distinct factors associated with declines in maternal mortality in Eastern and Western China, which may help inform efforts to reach global targets.

Introduction

Tackling maternal mortality remains a complex task globally. Although some advances have been made, the progress has been slow and uneven. The declining trends in maternal mortality experienced a stagnation between 2016 and 2020 on a global level, and the global maternal mortality in 2020 was 223 per 100,000 livebirths [1]. The current state is off track in achieving the Sustainable Development Goals (SDG) target 3.1 of a maternal mortality below 70 per 100,000 livebirths by 2030 [1]. Of all maternal deaths, nearly 95% occurred in low- and lower-middle-income countries (LMICs), and 68 countries still had a maternal mortality of over 100 maternal deaths per 100,000 livebirths in 2020 [1]. Multisectoral action is required for less developed regions to transition from high to low maternal mortality stages. Evidence is needed on why some countries achieved continuous improvement and fulfilled the SDGs.

The World Health Organization (WHO) ranked China as one of the top 10 low- and middle-income fast-track countries in women’s and children’s health. It has achieved a considerable and consistent reduction in maternal mortality from 94.7 deaths in 1990 to 15.7 deaths per 100,000 livebirths in 2022 [2–4]. Currently, China has fulfilled the aim of reducing the maternal mortality to less than 70 per 100,000 livebirths in the SDGs. The more developed Eastern China has achieved Healthy China 2030’s target level of 12 per 100,000 [2,3]. Although regional disparities also exist, and some western regions still have a long way to go, the gap has narrowed, with an 81% reduction in maternal mortality between 1996 and 2,018 in Western China [5]. These successes were not only attributable to the efforts aimed at addressing the leading biomedical causes of maternal mortality, but also the manifestations of the development in economic, political, and cultural systems [2]. The economic, political, and cultural superdeterminants may influence the social determinants of health and the performance of the healthcare system, which are the distal determinants of maternal health outcomes [6]. A comprehensive, quantitative analysis of the distal determinants contributing to the reduction of maternal mortality in the East and West may inform strategies for other less developed regions worldwide to achieve targets for 2030 and beyond.

Accordingly, guided by the recent review on superdeterminants of maternal mortality [6], we aimed to estimate the most contributing social and health system-related factors to the decline in total and cause-specific maternal mortality in China from 2004 to 2020. However, social and health system-related factors are often highly correlated, for example, between gross domestic product and life expectancy, and between population education levels and healthcare expenditure. Therefore, a new statistical method is required to address the collinearity of covariates. In recent years, Bayesian kernel machine regression (BKMR) has been increasingly used in environmental epidemiological studies since human beings are exposed to multiple chemicals simultaneously and these chemicals are often highly correlated [7,8]. BKMR is a non-parametric approach that combines Bayesian and statistical learning methods to estimate the joint exposure–response function iteratively using a Gaussian kernel function [7,8]. Under a hierarchical variable selection approach, BKMR addresses the possibility of collinearity by concurrently estimating both the importance of groups of highly correlated exposures or exposures from common sources, and the individual components within the group, making it more suitable for co-occurring exposure data than a traditional single-exposure model [8]. Here, we applied this method to social epidemiology as social and health system-related factors also concurrently impact pregnant women. We aimed to estimate the posterior inclusion probability (PIP) and the single-exposure risk measures of 9 social and health system factors, including urbanization rate, per capita disposable income (PCDI), average years of schooling, number of health technical personnel in maternal and child health (MCH), number of hospital beds for obstetrics and gynecology (Ob/Gyn), local fiscal expenditure on healthcare, prenatal booking rate, antenatal care rate, and hospital delivery rate, and identify the factors with higher values of exposure–response relationships for reduced maternal mortality in China, thereby drawing implications for LMICs around the world.

Methods

Data source

Our study used publicly available databases, including the Global Burden of Diseases, Injuries, and Risk Factors Study 2021 (GBD 2021), the National Health Statistics Yearbooks, and the China Statistical Yearbooks. Guided by the framework and the list of non-biomedical determinants of maternal mortality proposed by Souza and colleagues [6], we selected the relevant social and health system-related factors from our data source (S1 Table). We restricted the analysis to the period between 2004 and 2020 because data on total and cause-specific maternal mortality at the provincial level are available during this period.

The National Health Statistical Yearbook is an annual statistical compendium documenting the progression of China’s healthcare system and population health status [9]. Data on the maternal mortality, proportions of cause-specific maternal deaths, hospital delivery rate, antenatal care rate, prenatal booking rate, local fiscal expenditure on healthcare, number of health technical personnel in MCH care per 1,000 livebirths (hereafter referred to as number of MCH personnel), and number of hospital beds for Ob/Gyn per 1,000 livebirths (hereafter referred to as number of Ob/Gyn beds) in 31 provinces in mainland China from 2004 to 2020 were obtained from the National Health Statistics Yearbook (2005–2021). In the National Health Statistical Yearbook, the causes of maternal deaths were only classified as hemorrhage, coexisting medical conditions, hypertensive disorders in pregnancy (HDP), amniotic fluid embolism, and others. Due to a lack of detailed classification of causes of maternal deaths in the National Health Statistical Yearbook, we also extracted the data on the number and rate of deaths due to maternal disorders, including maternal hemorrhage, maternal sepsis and other maternal infections, maternal hypertensive disorders, obstructed labor and uterine rupture, abortion and miscarriage, ectopic pregnancy, indirect maternal deaths, late maternal deaths, maternal deaths aggravated by HIV/AIDS, and other direct maternal disorders, for females aged between 15 and 49 in China during 2004–2020 from GBD 2021 to identify the major causes of maternal mortality [10]. Indirect maternal deaths are due to existing diseases that are exacerbated by pregnancy. Coexisting medical conditions recorded in the National Health Statistical Yearbook specifically refer to internal medicine diseases that develop during pregnancy or pre-exist and are aggravated by pregnancy. The China Statistical Yearbook is an annual statistical publication that reflects China’s economic and social development [11]. Data derived from the China Statistical Yearbook (2005–2021) include PCDI, urbanization rate, and the population by education attainment level for females in 31 provinces on the mainland of China from 2004 to 2020.

Definition of variables

Maternal mortality refers to the number of maternal deaths per 100,000 livebirths [12]. The National Health Statistical Yearbook defines maternal deaths as any death of a woman while pregnant or within 42 days of the termination of the pregnancy from any cause related to the pregnancy or its management, but not from accidental or incidental causes [12]. Hospital delivery rate refers to the ratio of pregnant women delivered in hospitals to the number of livebirths. Antenatal care rate is the ratio of pregnant women who received at least one antenatal visit to the number of livebirths. Prenatal booking rate is the ratio of pregnant women who have maternal health records established by healthcare providers to the number of livebirths. Local fiscal expenditure on healthcare is the government’s expenditure on medical and healthcare services, medical subsidies, health administration, and family planning. The number of MCH personnel is the number of professional staff engaged in MCH healthcare services per 1,000 livebirths. The number of Ob/Gyn beds refers to the year-end number of beds in Ob/Gyn departments per 1,000 live births. Urbanization rate measures the proportion of a population residing in cities and towns. PCDI refers to the average income obtained by dividing the total disposable income by the resident population. To calculate the average years of schooling for females, the population by education attainment level for females was multiplied by the duration of the corresponding level first, and then the sum of these results was divided by the female population aged 6 and older.

Statistical analysis

The eastern and western regions of China still have a disparity in socioeconomic development. We used the Aihui-Tengchong line as the geographical dividing line proposed by the Chinese geographer Huanyong Hu to divide China into eastern and western regions based on population density, natural conditions, and social, economic and human activities [13]. Maternal mortality in the East and West was calculated as the weighted average of maternal mortality in the 24 eastern provinces (Anhui, Beijing, Chongqing, Fujian, Guangdong, Guangxi, Guizhou, Hainan, Hebei, Heilongjiang, Henan, Hubei, Hunan, Jiangsu, Jiangxi, Jilin, Liaoning, Shandong, Shanghai, Shanxi, Sichuan, Tianjin, Yunnan, Zhejiang) and 7 western provinces (Gansu, Inner Mongolia, Ningxia, Qinghai, Shaanxi, Tibet, Xinjiang), respectively. The cause-specific mortality rate for each province was estimated by multiplying the maternal mortality and the proportion of maternal deaths from a specific cause.

The following data were missing for all provinces: local fiscal expenditure on healthcare between 2004 and 2006, PCDI in 2004, and the number of health technical personnel in maternal and child healthcare per 1,000 livebirths in 2009. For missing data imputation, we fitted a linear regression model for year and each of these three factors in each province. We used linear regression estimate imputation because the socio-economic missing value has a correlation with time and is at the beginning of the time [14]. We calculated the Spearman correlation coefficients among these social and health system factors (S1 Fig). Based on the correlation patterns and the sociologically-informed properties of these factors, we partitioned the 9 factors into 4 groups for the hierarchical variable selection in BKMR: 1) social determinants, including urbanization rate, PCDI, and average years of schooling; 2) health resources, including the number of MCH personnel and the number of Ob/Gyn beds; 3) health financing, including local fiscal expenditure on healthcare; and 4) coverage of maternal care, including prenatal booking rate, antenatal care rate, and hospital delivery rate.

The R package “bkmr” was used to perform the BKMR model before and after 2013, the year that a historic national program, Reducing Maternal Mortality and Eliminating Neonatal Tetanus, came to an end. In China, the annualized rate of reduction in maternal mortality was 4.9% from 1996 to 2000, which is slower than the Millennium Development Goal (MDG) 5 target of a 5.5% annual decline [15,16]. To meet the target pace in MDG 5, the Chinese government implemented the Reducing Maternal Mortality and Eliminating Neonatal Tetanus Program from 2000 to 2013 [2,17]. The program began in 378 counties across 12 western provinces, expanded to 2,288 counties in all central and western provinces by 2009, and was nationalized in 2009.

All the social and health system factors were standardized for the BKMR analysis. The kmbayes() function was used to fit the BKMR models with 10,000 Markov chain Monte Carlo iterations. Each province was assigned an identifier, so that a random intercept was included in the model to account for the correlation between repeated measures within the same province. The embedded hierarchical variable selection was performed to estimate the PIP of each of the 4 groups (group PIP, GroupPIP), and each factor within a group (conditional PIP, CondPIP). Both GroupPIP and CondPIP are indices that range from 0 to 1. A larger GroupPIP suggests that this group is relatively more likely to be associated with the outcome than other groups. Similarly, a larger ConPIP suggests that the factor is more likely to be associated with the outcome within the group.

To summarize the contribution of each single exposure to the response, we used the SingVarRiskSummaries() function to calculate the exposure–response relationship associated with a change in a single factor from its 75th percentile to its 25th percentile when other factors are fixed to a specific quantile. We used the following criteria to identify the factors associated with decreased maternal mortality: 1) Among the four groups of social and health system factors, we first selected the groups with a GroupPIP ≥ 0.9; 2) For a group with a GroupPIP ≥ 0.9, the factor with the highest CondPIP was selected; 3) For the factor with the highest CondPIP within a group, significant and negative estimate of the single-exposure risk measures should be observed.

Several sensitivity analyses were performed to test the robustness of our results. 1) BKMR is less interpretable than traditional semi-parametric and parametric statistical models in providing a direct epidemiological interpretation with parsimonious parametric inference. Therefore, we further applied linear mixed-effects models to investigate the parameter inference for the factors identified by BKMR. The effect of each factor was estimated using univariate and multivariate linear mixed-effects models weighted by the number of livebirths in each province for each year. In the multivariate mixed-effects models, the other 8 social and health system factors were adjusted for when the effect of each factor was calculated. 2) Considering the potential impact of inflation on the analytical results, we conducted the BKMR analysis with government health expenditure adjusted for inflation. Government health expenditure was adjusted using 2012 as the base year for reference and the yearly inflation in the International Financial Statistics database [18]. 3) We used another method, multivariate imputation by chained equations (MICE) algorithm, to impute the missing values of exposures in 2004–2006 and 2009. Five imputed datasets were generated and modeled separately to re-examine the effects of these social and health system factors between 2004 and 2012. Results were pooled to obtain the final effect parameters. All the analyses were conducted in R 4.3.6. This study is reported as per the Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER) statement (S1 Checklist).

Results

Trends in maternal mortality in Eastern and Western China from 2004 to 2020

In both the East and West, maternal mortality declined substantially, and the urban-rural gap in maternal mortality also narrowed between 2004 and 2020 (Fig 1). However, regional differences in maternal mortality persisted between the East and West during this period (S2 Fig). The characteristics of social and health system-related factors in the East and West before and after 2013 are shown in Table 1.

Fig 1 — The dashed line represents the year 2013 when the Reducing Maternal Mortality and Eliminating Neonatal Tetanus Program (2000–2013) ended.

Table 1. Characteristics of social and health-system factors in Eastern and Western China during 2004-2012 and 2013–2020.

Factor	Eastern China		Western China
Factor	2004–2012	2013–2020	2004–2012	2013–2020
Hospital delivery rate (%)	98.7 [91.7, 99.7]	100 [99.9, 100]	89.3 [77.2, 97.4]	99.7 [98.6, 100]
Antenatal care rate (%)	92.9 [88.8, 96.5]	97.2 [96.0, 98.3]	92.4 [85.7, 95.9]	97.1 [94.7, 97.7]
Prenatal booking rate (%)	92.3 [85.5, 96.6]	96.2 [94.2, 98.3]	91.9 [84.3, 96.1]	95.7 [92.7, 97.7]
Local fiscal expenditure on healthcare (100 million Chinese Yuan)	147 [90, 219]	444 [302, 617]	54 [34, 116]	203 [104, 313]
Urbanization rate (%)	45.8 [35.9, 55.9]	58.4 [52.7, 67.4]	39.8 [29.5, 44.9]	52.1 [46.3, 58.3]
PCDI (10 thousand Chinese Yuan)	1.0 [0.7, 1.4]	2.4 [1.9, 3.0]	0.73 [0.55, 0.98]	1.9 [1.5, 2.2]
Average years of schooling for females (years)	8.0 [7.5, 8.9]	8.8 [8.4, 9.2]	7.7 [6.5, 8.2]	8.5 [7.6, 9.0]
Number of Ob/Gyn beds per 1,000 livebirths	21.1 [16.9, 34.6]	30.7 [25.3, 40.0]	23.3 [20.0, 27.8]	34.9 [28.1, 42.1]
Number of MCH personnel per 1,000 livebirths	13.9 [11.0, 18.9]	21.4 [15.2, 29.8]	12.9 [9.1, 22.1]	19.4 [12.0, 26.7]

Open in a new tab

Note: Data are median [Q1, Q3]. MCH, maternal and child health; Ob/Gyn, obstetrics and gynecology; PCDI, per capita disposable income.

Based on the data from the GBD 2021 study, maternal hemorrhage, indirect maternal deaths, other direct maternal disorders, and maternal hypertensive disorders remained the leading causes of maternal mortality in China from 2004 to 2020 (Fig 2 and S2 and S3 Tables). We further utilized provincial-level data from the National Health Statistics Yearbooks to estimate the cause-specific mortality rates in the East and West. Due to the lack of a detailed classification of maternal disorders in the provincial-level data comparable to that in the GBD study, we focused on hemorrhage, coexisting medical conditions, and HDP in the following analysis. These major causes of maternal mortality have been steadily declining in both the East and West from 2004 to 2020 (Fig 3).

Fig 2 — The top four causes are labeled with the mortality rate (left panel) and fraction (right panel).

Fig 3 — HDP, hypertensive disorders in pregnancy.

Factors associated with total and cause-specific maternal mortality in Eastern and Western China

Using BKMR, we identified five factors associated with the decreased maternal mortality based on the results of GroupPIP, CondPIP, and single-exposure risk measures (S4–S7 Tables and S3–S6 Figs): hospital delivery rate, antenatal care rate, urbanization rate, local fiscal expenditure on healthcare, and PCDI (Fig 4A–4D).

Fig 4 — The asterisk * and dagger † symbols in the cell represent that the factor contributes the most to the exposure–response relationship when all other factors are fixed at their 25th and 75th percentiles, respectively. White cell means that the factor is not identified as the component associated with reduced maternal mortality in the mixture. PCDI, per capita disposable income.

For the reduction of total maternal mortality before 2013, the factors with larger GroupPIP and CondPIP, and significant single-exposure risk measures in the East were hospital delivery rate, urbanization rate, and local fiscal expenditure on healthcare (Figs 4A and S3 and S4 Table). When all other factors were at a lower level (25th percentile), an increase in hospital delivery rate from 91.7% to 99.7% was associated with a higher value of the exposure–response relationship with a posterior mean and standard deviation (SD) of −14.8(1.5) (S3 Fig). When all other factors reached a high level (75th percentile), an increase in urbanization rate from 35.9% to 55.9% and an increase in local fiscal expenditure on healthcare from 9 billion to 21.9 billion Chinese Yuan were associated with larger responses in the decrease in maternal mortality [posterior mean and SD for urbanization rate: −5.8(2.6); posterior mean and SD for local fiscal expenditure on healthcare: −5.1(1.4)]. Before 2013, the factors identified by BKMR in the western region were PCDI and antenatal care rate (Figs 4A and S4 and S5 Tables). When other factors were at a lower level, an increase in PCDI from 5,500 to 9,800 Chinese Yuan [posterior mean and SD: −43.4 (8.1)] and an increase in antenatal care rate from 85.7% to 95.9% [posterior mean and SD: −33.8 (6.8)] were associated with larger single-exposure risk measures. An increase in antenatal care rate from 85.7% to 95.9% was also associated with a higher response in the decrease in maternal mortality when other factors were at a higher level [posterior mean and SD: −21.4(9.6)] (S4 Fig). After 2013, the urbanization rate and hospital delivery rate had higher GroupPIP and CondPIP, and significant single-exposure risk measures for the reduction of total maternal mortality in the East. An increase in the urbanization rate from 52.7% to 67.4% was associated with a larger response when considering the single-exposure risk measures, with a posterior mean and SD of −3.9(0.6) when other factors were at a lower lever (Figs 4A and S5 and S6 Table). In the West after 2013, an increase in antenatal care rate from 94.7% to 97.7% was associated with a larger response when considering the single-exposure risk measures, with a posterior mean and SD of −11.5(4.1) when other factors were at a lower level (Figs 4A and S6 and S7 Table).

The reduction of deaths due to hemorrhage was mainly correlated with hospital delivery rate, urbanization rate, and local fiscal expenditure on healthcare in the East during 2004–2012, with hospital delivery rate having a higher single-exposure risk measure, and hospital delivery rate and urbanization rate during 2013–2020, with urbanization rate having a higher single-exposure risk measure in the mixture of factors (Figs 4B, S3 and S5 and S4 and S6 Tables). In the West, however, antenatal care rate and hospital delivery rate were the factors with larger GroupPIP and CondPIP, and significant single-exposure risk measures before and after 2013, respectively (Figs 4B, S4 and S6 and S5 and S7 Tables).

For the reduction of maternal mortality caused by coexisting medical diseases in the East, the urbanization rate and local fiscal expenditure on healthcare were the factors associated with higher values of the exposure–response relationship before and after 2013, respectively (Figs 4C, S3 and S5 and S4 and S6 Tables). Before 2013, hospital delivery rate was also a factor associated with the decreased maternal mortality due to coexisting medical diseases in both the East and the West (Figs 4C and S3–S4 and S4–S5 Tables).

Hospital delivery rate and local fiscal expenditure on healthcare were the factors with larger GroupPIP and CondPIP, and significant single-exposure risk measures for the decline in deaths due to HDP in the East before 2013, and hospital delivery rate was associated with a larger single-exposure risk measure (Figs 4D and S3 and S4 Table). Antenatal care rate was the factor associated with a higher value of the exposure–response relationship in the West during 2004–2012, and in the East during 2013–2020 (Figs 4D, S4 and S5, and S5–S6 Tables).

Sensitivity analyses

S8 and S9 Tables show the results of linear mixed-effects models for the factors identified by BKMR, including hospital delivery rate, antenatal care rate, urbanization rate, local fiscal expenditure on healthcare, and PCDI. Consistent with the results in the primary analyses, hospital delivery rate was negatively associated with total and cause-specific maternal mortality in the East before 2013. When adjusting for the other eight factors, each additional percentage of hospital delivery rate was associated with nearly one percent reduction in total maternal mortality and maternal death due to hemorrhage. In the West before 2013, each additional percentage of antenatal care rate was associated with 2% and 1% of reduction in maternal death due to hemorrhage and HDP, respectively, when adjusting for the other eight factors. The observations for the urbanization rate and PCDI also aligned with the primary results.

The results of BKMR analysis, with government health expenditure adjusted for inflation, are shown in S7–S11 Figs and S10–S13 Tables. The results of GroupPIP, CondPIP, and single-exposure risk measures did not change substantially. When using datasets with imputations for missing data by MICE, the associations between factors and outcomes remained robust (S12–S14 Figs and S14–S15 Tables).

Discussion

In this study, we conducted a mixture analysis to quantify the exposure–response relationships of social and health system-related factors and the decline of maternal mortality in Eastern and Western China from 2004 to 2020. We found that hospital delivery, antenatal care, urbanization, local fiscal expenditure on healthcare, and PCDI were the factors associated with the remarkable progress in maternal survival in both the eastern and western regions during this period. In addition, we observed the differential impact of different factors on cause-specific maternal mortality in the eastern and western regions before and after 2013, which may inform strategies for other less developed areas in the coming decades.

In our mixture analysis considering the simultaneous effects of multiple distal factors, the coverage of maternal care, including antenatal care and hospital delivery, was a key component associated with the decline in maternal deaths in China. As an indispensable part of the continuum of maternal healthcare, antenatal care reduces maternal mortality through timely detection and management for chronic conditions and pregnancy complications, particularly for individuals at increased risk of developing complications during labor and delivery [19]. A minimum of eight antenatal visits is recommended by WHO to reduce perinatal mortality [19]. In-hospital delivery reflects the accessibility of hospitals and the availability of skilled attendants [20]. With more skilled health workforce and more advanced equipment and infrastructure, hospital-based care can offer opportunities for early recognition of labor complications, facilitate timely treatment and emergency services, and provide better protection against mortality [21,22]. High-quality facility birth has been recommended globally as a strategy to reduce maternal mortality, especially in less developed regions where many women give birth at home with a non-certified birth attendant rather than in hospitals with professional health staff [23,24]. Using simulation modeling, Ward and colleagues found that for countries with maternal mortality levels between 140 and 1,000 per 100,000 livebirths, the most effective policy to reduce maternal mortality between 2030−2050 is increasing facility births, which is projected to prevent 22.6%−51.5% of maternal deaths [25].

Although the coverage of antenatal care and facility delivery increased in LMICs over the past several decades, 11% of women in 54 Countdown to 2030 priority countries had no antenatal care during pregnancy during 2012–2018, and 31 LMICs still had less than 80% utilization of facility delivery services in 2018 [26,27]. In countries with a maternal mortality of ≥700 per 100,000 livebirths, the coverage of at least one antenatal care visit and institutional birth was only 66% and 36% during 2000–2020, respectively [28]. Qualitative evidence in LMICs has found that the cost of care, distance to the facility, perceived quality of care, and social and cultural context are all barriers to accessing antenatal care and facility-based delivery [29–32].

It should also be noted that without ensuring adequate quality of care, increased coverage of maternal care does not necessarily improve maternal survival [33–36]. Increasing the proportion of women delivering in medical facilities is estimated to reduce total maternal death, death from hypertensive disorders and from hemorrhage by merely 29%, 29% and 53%, respectively [33]. Instead, an integrated strategy to simultaneously increase facility births, improve the availability of clinical services and quality of care at facilities, and improve linkages to care could substantially improve maternal health outcomes and reduce total maternal death, death from hypertensive disorders and from hemorrhage by 51%, 60% and 79%, respectively [33]. In addition to the recommendation on the number of antenatal visits, WHO also emphasizes the interventions to improve the quality of antenatal services in LMICs [19].

In China, a comprehensive strategy consisting of multidimensional interventions was employed in the Reducing Maternal Mortality and Eliminating Neonatal Tetanus Program to improve both the quantity and quality of maternal care and in-hospital delivery [2,17,37–39]. First, hospital delivery was subsidized by the government, and on average, women in rural areas who delivered in a hospital received a ¥500 subsidy directly at the time of discharge. Second, public awareness campaigns utilizing posters, cartoons, slogans, and other formats were organized, and primary healthcare staff and village cadres worked together to provide health education and advocate for maternal care and hospital-based childbirth. Third, 24-hour obstetric emergency centers were built at the county level and pregnant women screened as high risk in the primary health system were referred to hospitals. The functions of traditional birth attendants were transformed from assisting home delivery to supporting antenatal care, referrals and the advocacy of hospital delivery. Fourth, equipment for emergency and routine obstetric care was provided to township health centers, and experts from provincial and municipal hospitals were dispatched to county and township hospitals to train local health staff. In addition to this program, a series of regulations and services were also instituted to improve the quality of antenatal care. The Ministry of Health in China required a minimum of five antenatal visits, including one visit in the first trimester, two in the second trimester, and two in the third trimester [40]. In 2009, the Chinese government launched the National Essential Public Health Services Program to provide universal coverage of essential health services nationwide, including five antenatal examinations free of charge at the point of care [17]. The cumulative effect of these programs and services resulted in a 68.7% increase in the hospital delivery rate in the project area, and the coverage of at least five antenatal visits increased from 38% in 2003 to 57% in 2,011 in Western China [2,9,17]. A subnational analysis revealed that the annualized rate of decline in maternal mortality improved year-over-year after the program’s implementation and reached a mean county-level rate of 9.3% during 2005–2015 [15]. Maternal mortality in the project area decreased by 67.1% from 2000 to 2013 [2]. The effect of hospital delivery and antenatal care on maternal survival has persisted in recent decades, as shown in our mixture analysis after 2,013 in both the East and West.

Using a mixture analysis, we found that the effects of hospital delivery and antenatal care on maternal survival were paralleled by the contribution of social determinants and health financing. Social determinants, such as income levels, can affect maternal health through changes in women’s decisions to seek care. Women living in urban areas have more access to high-quality and high-acuity healthcare during pregnancy and childbirth [41–43]. Financing is a pivotal factor contributing to universal health coverage, and allocating adequate resources and effective healthcare financing is one of the cross-cutting actions proposed by WHO for ending preventable maternal deaths [2,44]. Unfortunately, recent reports published by the World Bank Group revealed a steady decline in per capita government health spending from 2020 to 2023 and troubling trends in the prioritization of health across LMICs, and the expected spending trajectories in LMICs remain below the minimum needed to meet the SDGs by 2030 [45,46]. These financial shortfalls underscore the urgent need to increase government commitment to health and reform financing policies [45,46].

For LMICs with a high burden of maternal mortality, lessons could be learnt from China’s remarkable success. Here, we identified the factors associated with higher values of the exposure–response relationship in order to suggest the priorities that can be considered in policy and program planning. This is particularly relevant for resource-limited settings, where simultaneous optimization of multiple factors may not be feasible. The improvement of the quantity and quality of facility births and antenatal care might be the leverage point to accelerate the decline in maternal mortality. In China, antenatal attendance and safe birth in health facilities were underpinned by strong government commitment to maternal health and a series of long-term strategies, featuring in the subsidies and health education to reduce the financial and cultural barriers that delay the decision to seek care, and the green channel service for obstetric emergencies and the professionalization of obstetric healthcare to reduce the delays in reaching a health facility and receiving adequate care [17,47]. Additionally, a hierarchical healthcare network spanning from national to village levels might be considered in LMICs to establish a well-functioning referral system and foster close working partnerships between hospitals and primary healthcare facilities.

A major strength of our study is the quantification of the impact of non-medical distal factors associated with maternal survival in the East and West over a nearly 20-year timescale. We included a list of socioeconomic, financial and system-related factors at the national and subnational levels. These factors are closely linked to the policies and strategies China has adopted over the past few decades, from which lessons can be extracted to share with less developed regions in the world. In addition to total maternal mortality, we also measured the distal factors for reduced maternal deaths due to major biomedical causes to provide insights to guide targeted intervention programs for each maternal disorder. Furthermore, we utilized mixture analysis to simultaneously estimate the effects of these co-occurring factors, which can address the issue of collinearity and rank the values of exposure–response relationships for these factors.

Our analysis is limited by the unavailability of some direct indicators of non-medical factors. For example, although urbanization may represent the development of healthcare services and transportation, it is not a direct measure of the distance to a health facility. However, we used multiple indicators in addition to the urbanization rate, including the number of medical personnel and hospital beds, as well as the rate for prenatal booking, antenatal care and hospital delivery, to reflect the accessibility of health services. Besides, data on other non-medical factors are unavailable, such as gender and social bias against women. A global analysis including more distal factors is warranted to rank the values of exposure–response relationships for these co-occurring contributors to decreased maternal mortality. Second, data at the provincial level outside the 2004–2020 range are unavailable in the current study. However, due to the rapid shifts in economic, political and cultural contexts, the social and system factors before the MDG era might have little applicability in the following decades. Besides, according to Souza and colleagues’ obstetric transition model [6], China is now in the category of Stage 4B, which represents a very low maternal mortality (maternal mortality <20). Therefore, the experience in China after 2020 would be impractical for countries with a high maternal mortality burden. Third, data on rural and urban areas at the provincial level are unavailable for most of the factors. Therefore, we are not able to analyze the contributing factors by rural and urban areas in the East and West. Fourth, both too much, too soon and too little, too late are associated with maternal mortality and have been found to coexist in China and other LMICs [48]. However, due to the lack of data reflecting over-medicalization, we were unable to examine the impact of insufficient and excessive medicalization on maternal mortality in China in the past decades. Fifth, due to the data constraints in the National Health Statistics Yearbooks, our analysis only provides the data for maternal deaths from coexisting medical diseases rather than the indirect causes, including both previously existing conditions and non-communicable diseases. It is of great importance to conduct further investigations of the distal factors associated with indirect maternal deaths, which remain the second largest cause of maternal deaths globally [49]. Finally, this study relied on national annual statistics, which may be subject to reporting biases and regional inconsistencies in data collection and data quality. This limitation should be considered when interpreting the results.

Globally, maternal mortality is currently not on track to meet the SDG target, with a disproportionate burden in LMICs [1]. Although the social and health system factors are all crucial to maternal mortality, investigating the values of exposure–response relationships for these factors can provide practical guidance for policy and program planning in these developed regions. For other less developed regions around the world, enhancing both the quantity and quality of antenatal care and facility births may be considered a policy priority. Health financing and social determinants are also crucial in ensuring universal coverage of maternal health services in these regions. Actions targeting these non-biomedical, distal factors are necessary to hold any hope of achieving the SDG target 3.1.

Supporting information

S1 Table. Social and health system-related factors of maternal mortality.

Note: Ob/Gyn, obstetrics and gynecology.

(DOCX)

pmed.1004837.s001.docx^{(16.9KB, docx)}

S2 Table. Cause-specific maternal mortality rate in China, 2004–2020.

(DOCX)

pmed.1004837.s002.docx^{(15.6KB, docx)}

S3 Table. Cause-specific maternal mortality fractions in China, 2004–2020.

(DOCX)

pmed.1004837.s003.docx^{(15.8KB, docx)}

S4 Table. Group and conditional posterior inclusion probabilities for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection.

Note: GroupPIP, group posterior inclusion probabilities; CondPIP, conditional posterior inclusion probabilities; MCH, maternal and child health; Ob/Gyn, obstetrics and gynecology; PCDI, per capita disposable income.

(DOCX)

pmed.1004837.s004.docx^{(15KB, docx)}

S5 Table. Group and conditional posterior inclusion probabilities for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection.

(DOCX)

pmed.1004837.s005.docx^{(15.2KB, docx)}

S6 Table. Group and conditional posterior inclusion probabilities for each factor in Eastern China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection.

(DOCX)

pmed.1004837.s006.docx^{(15.1KB, docx)}

S7 Table. Group and conditional posterior inclusion probabilities for each factor in Western China, 2013––2020, using Bayesian Kernel Machine Regression hierarchical variable selection.

(DOCX)

pmed.1004837.s007.docx^{(15.1KB, docx)}

S8 Table. Effects of factors on maternal mortality, using univariate linear mixed-effects models weighted by the number of livebirths.

Note: Effects are shown as estimates and the 95% confidence intervals in the linear mixed effects models.

(DOCX)

pmed.1004837.s008.docx^{(15.9KB, docx)}

S9 Table. Effects of factors on maternal mortality, using multivariate linear mixed-effects models weighted by the number of livebirths.

Note: Effects are shown as estimates and the 95% confidence intervals in the linear mixed effects model, adjusting for the other eight factors. For example, the effect of hospital delivery rate was estimated when urbanization rate, per capita disposable income, average years of schooling, number of health technical personnel in maternal and child health, number of hospital beds for obstetrics and gynecology, local fiscal expenditure on healthcare, prenatal booking rate, and antenatal care rate were adjusted for.

(DOCX)

pmed.1004837.s009.docx^{(16KB, docx)}

S10 Table. Group and conditional posterior inclusion probabilities for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

(DOCX)

pmed.1004837.s010.docx^{(15KB, docx)}

S11 Table. Group and conditional posterior inclusion probabilities for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

(DOCX)

pmed.1004837.s011.docx^{(15KB, docx)}

S12 Table. Group and conditional posterior inclusion probabilities for each factor in Eastern China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

(DOCX)

pmed.1004837.s012.docx^{(15KB, docx)}

S13 Table. Group and conditional posterior inclusion probabilities for each factor in Western China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

(DOCX)

pmed.1004837.s013.docx^{(15KB, docx)}

S14 Table. Group and conditional posterior inclusion probabilities for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with missing data imputed by MICE.

(DOCX)

pmed.1004837.s014.docx^{(15KB, docx)}

S15 Table. Group and conditional posterior inclusion probabilities for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with missing data imputed by MICE.

(DOCX)

pmed.1004837.s015.docx^{(15KB, docx)}

S16 Table. Data inputs for BKMR.

(XLSX)

pmed.1004837.s016.xlsx^{(75.5KB, xlsx)}

S1 Fig. Spearman correlation coefficient matrix for the non-medical factors.

Note: The number in the cells represent the Spearman correlation coefficient. The circles are filled clockwise as pie charts to indicate the positive values of the correlation coefficients, and the color intensity represents the strength or the correlation. The asterisks in the circles indicate the significant levels with * representing a P-value of < 0.05, ** representing a P-value of < 0.01, and *** representing a P-value of < 0.001. Ob/Gyn beds: Number of hospital beds for obstetrics and gynecology per 1,000 livebirths; MCH personnel: Number of health technical personnel in maternal and child healthcare per 1,000 livebirths; Years of schooling: Average years of schooling for females; PCDI: per capita disposable income; Hospital delivery: Hospital delivery rate; Local fiscal expenditure: Local fiscal expenditure on healthcare; Antenatal care: Antenatal care rate; Prenatal booking: Prenatal booking rate

(TIFF)

pmed.1004837.s017.tiff^{(908.5KB, tiff)}

S2 Fig. Map of the provincial-level maternal mortality in mainland China in 2004, 2011 and 2020.

Note: The source of the basemap is: https://cloudcenter.tianditu.gov.cn/administrativeDivision.

(TIFF)

pmed.1004837.s018.tiff^{(1.1MB, tiff)}

S3 Fig. Single-exposure risk measures for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection.

Note: The plot compares the exposure–response relationships associated with a change in a single exposure from the 75th percentile to the 25th percentile, when the other exposures are fixed at their 25th, 50th, and 75th percentiles. MCH, maternal and child health; Ob/Gyn, obstetrics and gynecology; PCDI, per capita disposable income; HDP, hypertensive disorders in pregnancy; q.fixed, quantiles at which to fix the remaining exposures.

(TIFF)

pmed.1004837.s019.tiff^{(28.2MB, tiff)}

S4 Fig. Single-exposure risk measures for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection.

(TIFF)

pmed.1004837.s020.tiff^{(28.3MB, tiff)}

S5 Fig. Single-exposure risk measures for each factor in Eastern China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection.

(TIFF)

pmed.1004837.s021.tiff^{(28.7MB, tiff)}

S6 Fig. Single-exposure risk measures for each factor in Western China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection.

(TIFF)

pmed.1004837.s022.tiff^{(28.4MB, tiff)}

S7 Fig. Factors associated with decreased total and cause-specific maternal mortality in Eastern and Western China during 2004–2012 and 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

Note: The asterisk * and dagger † symbols in the cell represent that the factor contributes the most to the exposure–response relationship when all other factors are fixed at their 25th and 75th percentiles, respectively. White cell means that the factor is not identified as the component associated with reduced maternal mortality in the mixture. CMD, coexisting medical diseases; HDP, hypertensive disorders in pregnancy; PCDI, per capita disposable income.

(TIFF)

pmed.1004837.s023.tiff^{(9.7MB, tiff)}

S8 Fig. Single-exposure risk measures for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

(TIFF)

pmed.1004837.s024.tiff^{(28MB, tiff)}

S9 Fig. Single-exposure risk measures for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

(TIFF)

pmed.1004837.s025.tiff^{(28.3MB, tiff)}

S10 Fig. Single-exposure risk measures for each factor in Eastern China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

(TIFF)

pmed.1004837.s026.tiff^{(28.5MB, tiff)}

S11 Fig. Single-exposure risk measures for each factor in Western China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection with fiscal expenditure adjusted for inflation.

(TIFF)

pmed.1004837.s027.tiff^{(28.5MB, tiff)}

S12 Fig. Factors associated with decreased total and cause-specific maternal mortality in Eastern and Western China during 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with missing data imputed by MICE.

(TIFF)

pmed.1004837.s028.tiff^{(6.6MB, tiff)}

S13 Fig. Single-exposure risk measures for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with missing data imputed by MICE.

(TIFF)

pmed.1004837.s029.tiff^{(28.5MB, tiff)}

S14 Fig. Single-exposure risk measures for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with missing data imputed by MICE.

(TIFF)

pmed.1004837.s030.tiff^{(28.6MB, tiff)}

S1 Checklist. Checklist of information that should be included in new reports of global health estimates.

(DOCX)

pmed.1004837.s031.docx^{(31KB, docx)}

Abbreviation:

BKMR: Bayesian kernel machine regression
LMICs: lower-middle-income countries
MCH: maternal and child health
MDG: Millennium Development Goal
PCDI: per capita disposable income
PIP: posterior inclusion probability
SDG: Sustainable Development Goals
WHO: World Health Organization

Data Availability

Data is publicly available at the National Health Commission website (http://www.nhc.gov.cn/mohwsbwstjxxzx/tjzxtjcbw/tjsj_list.shtml), National Bureau of Statistics website (https://www.stats.gov.cn/sj/ndsj/), and Global Health Data Exchange (GHDx) online website (https://vizhub.healthdata.org/gbd-results/). All data inputs for BKMR are also provided in S16 Table.

Funding Statement

This study was supported in part by the Shanghai Sailing Program (24YF2750300) (https://stcsm.sh.gov.cn and the Postdoctoral Fellowship Program of China Postdoctoral Science Foundation (GZC20251538) (https://www.chinapostdoctor.org.cn/home) (X.Z.) and the Gates Foundation (INV-088623) (https://www.gatesfoundation.org) (J.Z.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.World Health Organization. Trends in maternal mortality 2000 to 2020: estimates by WHO, UNICEF, UNFPA, World Bank Group and UNDESA/Population Division. Geneva: World Health Organization; 2023. [Google Scholar]
2.Qiao J, Wang Y, Li X, Jiang F, Zhang Y, Ma J, et al. A Lancet Commission on 70 years of women’s reproductive, maternal, newborn, child, and adolescent health in China. Lancet. 2021;397(10293):2497–536. doi: 10.1016/S0140-6736(20)32708-2 [DOI] [PubMed] [Google Scholar]
3.Shi H, Chen L, Wei Y, Chen X, Zhao Y. Improving maternal healthcare further in China at a time of low maternal mortality. BMJ. 2024;386:e078640. doi: 10.1136/bmj-2023-078640 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Partnership for Maternal, Newborn and Child Health, World Health Organization, the World Bank Group, Alliance for Health Policy and Systems Research. Success Factors for Women’s and Children’s Health: Policy and programme highlights from 10 fast-track countries. Geneva: World Health Organization; 2014. [Google Scholar]
5.Department of Maternal and Child Health. Report on the development of maternal and child health in China (2019). National Health Commission of China. 2019 May 27 [Cited 2025 Apr 9]. Available from: https://www.nhc.gov.cn/fys/c100077/201905/142632c029194e42a3cca755e336db80.shtml#.
6.Souza JP, Day LT, Rezende-Gomes AC, Zhang J, Mori R, Baguiya A, et al. A global analysis of the determinants of maternal health and transitions in maternal mortality. Lancet Glob Health. 2024;12(2):e306–16. doi: 10.1016/S2214-109X(23)00468-0 [DOI] [PubMed] [Google Scholar]
7.Yu L, Liu W, Wang X, Ye Z, Tan Q, Qiu W, et al. A review of practical statistical methods used in epidemiological studies to estimate the health effects of multi-pollutant mixture. Environ Pollut. 2022;306:119356. doi: 10.1016/j.envpol.2022.119356 [DOI] [PubMed] [Google Scholar]
8.Lazarevic N, Barnett AG, Sly PD, Knibbs LD. Statistical Methodology in Studies of Prenatal Exposure to Mixtures of Endocrine-Disrupting Chemicals: A Review of Existing Approaches and New Alternatives. Environ Health Perspect. 2019;127(2):26001. doi: 10.1289/EHP2207 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.National Health Commission. National Health Statistics Yearbook. National Health Commission. 2025 Jan 24 [Cited 2025 Mar 31]. Available from: http://www.nhc.gov.cn/mohwsbwstjxxzx/tjzxtjcbw/tjsj_list.shtml
10.Institute for Health Metrics and Evaluation (IHME). GBD Results. Seattle, WA: IHME, University of Washington. 2024. [Cited 2024 Oct 31]. Available from: https://vizhub.healthdata.org/gbd-results/ [Google Scholar]
11.National Bureau of Statistics. China Statistical Yearbook. National Bureau of Statistics. 2025 Mar 27 [Cited 2025 Apr 8]. Available from: https://www.stats.gov.cn/sj/ndsj/
12.National Health Commission. National Health Statistics Yearbook 2021. Beijing: Peking Union Medical College Press; 2021. [Google Scholar]
13.Guo H, Wang X, Wu B, Li X. Cognizing population density demarcative line (Hu Huanyong-Line) based on space technology. Bull Chin Acad Sci. 2016;31(12):1385–94. [Google Scholar]
14.Wubetie HT. Missing data management and statistical measurement of socio-economic status: application of big data. J Big Data. 2017;4(1). doi: 10.1186/s40537-017-0099-y [DOI] [Google Scholar]
15.Liang J, Li X, Kang C, Wang Y, Kulikoff XR, Coates MM. Maternal mortality ratios in 2852 Chinese counties, 1996-2015, and achievement of Millennium Development Goal 5 in China: a subnational analysis of the Global Burden of Disease Study 2016. Lancet. 2019;393(10168):241–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.World Health Organization, World Bank, United Nations Population Fund & United Nations Children’s Fund. Maternal mortality in 2005: estimates developed by WHO, UNICEF, UNFPA, and The World Bank. Geneva: World Health Organization; 2007. [Google Scholar]
17.Gao Y, Zhou H, Singh NS, Powell-Jackson T, Nash S, Yang M, et al. Progress and challenges in maternal health in western China: a Countdown to 2015 national case study. Lancet Glob Health. 2017;5(5):e523–36. doi: 10.1016/S2214-109X(17)30100-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.International Monetary Fund. Inflation, consumer prices (annual %) - China. World Bank Group. 2020 [Cited 2025 Sep 18]. Available from: https://data.worldbank.org/indicator/FP.CPI.TOTL.ZG?end=2020&locations=CN&start=2004&view=chart
19.World Health Organization. WHO recommendations on antenatal care for a positive pregnancy experience. Geneva: World Health Organization; 2016. [PubMed] [Google Scholar]
20.Li J, Liang J, Wang J, Ren Z, Yang D, Wang Y, et al. Spatiotemporal trends and ecological determinants in maternal mortality ratios in 2,205 Chinese counties, 2010-2013: A Bayesian modelling analysis. PLoS Med. 2020;17(5):e1003114. doi: 10.1371/journal.pmed.1003114 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Gage AD, Fink G, Ataguba JE, Kruk ME. Hospital delivery and neonatal mortality in 37 countries in sub-Saharan Africa and South Asia: An ecological study. PLoS Med. 2021;18(12):e1003843. doi: 10.1371/journal.pmed.1003843 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Chaka EE, Mekurie M, Abdurahman AA, Parsaeian M, Majdzadeh R. Association between place of delivery for pregnant mothers and neonatal mortality: a systematic review and meta-analysis. Eur J Public Health. 2020;30(4):743–8. doi: 10.1093/eurpub/ckz060 [DOI] [PubMed] [Google Scholar]
23.Chinkhumba J, De Allegri M, Muula AS, Robberstad B. Maternal and perinatal mortality by place of delivery in sub-Saharan Africa: a meta-analysis of population-based cohort studies. BMC Public Health. 2014;14:1014. doi: 10.1186/1471-2458-14-1014 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Filippi V, Ronsmans C, Campbell OMR, Graham WJ, Mills A, Borghi J, et al. Maternal health in poor countries: the broader context and a call for action. Lancet. 2006;368(9546):1535–41. doi: 10.1016/S0140-6736(06)69384-7 [DOI] [PubMed] [Google Scholar]
25.Ward ZJ, Atun R, King G, Sequeira DMello B, Goldie SJ. Global maternal health country typologies: A framework to guide policy. PLOS Glob Public Health. 2024;4(11):e0003867. doi: 10.1371/journal.pgph.0003867 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Hasan MM, Magalhaes RJS, Fatima Y, Ahmed S, Mamun AA. Levels, trends, and inequalities in using institutional delivery services in low- and middle-income countries: a stratified analysis by facility type. Glob Health Sci Pract. 2021;9(1):78–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Jiwani SS, Amouzou-Aguirre A, Carvajal L, Chou D, Keita Y, Moran AC, et al. Timing and number of antenatal care contacts in low and middle-income countries: Analysis in the Countdown to 2030 priority countries. J Glob Health. 2020;10(1):010502. doi: 10.7189/jogh.10.010502 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Boerma T, Campbell OMR, Amouzou A, Blumenberg C, Blencowe H, Moran A, et al. Maternal mortality, stillbirths, and neonatal mortality: a transition model based on analyses of 151 countries. Lancet Glob Health. 2023;11(7):e1024–31. doi: 10.1016/S2214-109X(23)00195-X [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Bohren MA, Hunter EC, Munthe-Kaas HM, Souza JP, Vogel JP, Gülmezoglu AM. Facilitators and barriers to facility-based delivery in low- and middle-income countries: a qualitative evidence synthesis. Reprod Health. 2014;11(1):71. doi: 10.1186/1742-4755-11-71 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Finlayson K, Downe S. Why do women not use antenatal services in low- and middle-income countries? A meta-synthesis of qualitative studies. PLoS Med. 2013;10(1):e1001373. doi: 10.1371/journal.pmed.1001373 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Alam CE, Abou-Abbas L, Ramadan MS, Asmar MK. Exploring the barriers to accessing antenatal care at the primary health care center level of a tertiary hospital in Lebanon: a qualitative study. BMC Health Serv Res. 2025;25(1):304. doi: 10.1186/s12913-025-12444-y [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Deo KK, Paudel YR, Khatri RB, Bhaskar RK, Paudel R, Mehata S, et al. Barriers to Utilization of Antenatal Care Services in Eastern Nepal. Front Public Health. 2015;3:197. doi: 10.3389/fpubh.2015.00197 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Ward ZJ, Atun R, King G, Sequeira Dmello B, Goldie SJ. A simulation-based comparative effectiveness analysis of policies to improve global maternal health outcomes. Nat Med. 2023;29(5):1262–72. doi: 10.1038/s41591-023-02311-w [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Fink G, Cohen J. Delivering quality: safe childbirth requires more than facilities. Lancet Glob Health. 2019;7(8):e990–1. doi: 10.1016/S2214-109X(19)30193-7 [DOI] [PubMed] [Google Scholar]
35.Nihal S, Shekhar C. An assessment of adequate quality antenatal care and its determinants in India. BMC Pregnancy Childbirth. 2024;24(1):698. doi: 10.1186/s12884-024-06806-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Mandiwa C, Namondwe B. Assessment of quality of antenatal care services and associated factors in Malawi: Insights from a nationwide household survey. PLoS One. 2024;19(6):e0305294. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Yu C. Analysis of the effects of implementing the Reducing maternal mortality and eliminating neonatal tetanus program. Maternal and Child Health Care of China. 2004;14:28–9. [Google Scholar]
38.Peng Y. Implementing the reducing maternal mortality and eliminating neonatal tetanus program to achieve the goals of the outlines of women’ and children’s development. Maternal and Child Health Care of China. 2000;15. [Google Scholar]
39.Wu C, Yang G, Wang W, Yang M. Interim assessment and analysis report on the reducing maternal mortality and eliminating neonatal tetanus program. Maternal and Child Health Care of China. 2002;:19–20. [Google Scholar]
40.Ministry of Health of the People’s Republic of China. Notice of “Management Methods for the Maternal Health Care Service” and “Standards For Maternal Health Care Service” Issued by Ministry of Health of the People’s Republic of China. Ministry of Health of the People’s Republic of China 2011 Jul 8 [Cited 2024 Oct 31]. Available from: https://www.nhc.gov.cn/zwgk/wtwj/201304/84ec2d5d37a54398ace49df48ca91c83.shtml
41.Jeong W, Jang S-I, Park E-C, Nam JY. The Effect of Socioeconomic Status on All-Cause Maternal Mortality: A Nationwide Population-Based Cohort Study. Int J Environ Res Public Health. 2020;17(12):4606. doi: 10.3390/ijerph17124606 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.ACOG Committee. ACOG Committee Opinion No. 586: Health disparities in rural women. Obstet Gynecol. 2014;123(2 Pt 1):384–8. [DOI] [PubMed] [Google Scholar]
43.Kozhimannil KB, Interrante JD, Henning-Smith C, Admon LK. Rural-urban differences in severe maternal morbidity and mortality in the US, 2007-15. Health Aff (Millwood). 2019;38(12):2077–85. [DOI] [PubMed] [Google Scholar]
44.World Health Organization. Strategies towards ending preventable maternal mortality (EPMM). Geneva: World Health Organization. 2015. [Google Scholar]
45.Kurowski C, Schmidt M, Kumar A, Mieses J, Gabani J. Government Health Spending Trends Through 2023: Peaks, Declines, and Mounting Risks. World Bank. 2024. [Google Scholar]
46.Kurowski C, Schmidt M, Evans DB, Tandon A, Eozenou PHV, Cain JS. Government health spending outlook - projections through 2029 diverging fiscal pressures, uneven constraints. World Bank. 2024. [Google Scholar]
47.Thaddeus S, Maine D. Too far to walk: maternal mortality in context. Soc Sci Med. 1994;38(8):1091–110. doi: 10.1016/0277-9536(94)90226-7 [DOI] [PubMed] [Google Scholar]
48.Miller S, Abalos E, Chamillard M, Ciapponi A, Colaci D, Comandé D, et al. Beyond too little, too late and too much, too soon: a pathway towards evidence-based, respectful maternity care worldwide. Lancet. 2016;388(10056):2176–92. doi: 10.1016/S0140-6736(16)31472-6 [DOI] [PubMed] [Google Scholar]
49.Cresswell JA, Alexander M, Chong MYC, Link HM, Pejchinovska M, Gazeley U, et al. Global and regional causes of maternal deaths 2009-20: a WHO systematic analysis. Lancet Glob Health. 2025;13(4):e626–34. doi: 10.1016/S2214-109X(24)00560-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS Med. doi: 10.1371/journal.pmed.1004837.r001

Decision Letter 0

Heather Van Epps

6 Aug 2025

Dear Dr Zhang,

Thank you for submitting your manuscript entitled "Social and Health System Determinants of Maternal Mortality in More and Less Developed Regions of China: Implications for SDG 2030" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff as well as by an academic editor with relevant expertise and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

For clinical studies, please upload a copy of your trial study protocol as a supporting information file. The study protocol should be the version submitted for approval to the institutional review board or ethics committee, should include any amendments to the study protocol, as well as the date of their approval by the institutional review or ethics committee. Please also detail any deviations from the study protocol in the Methods section of your manuscript. The editors will consider the protocol and study conduct prior to a final decision for external review.

Please re-submit your manuscript within two working days, i.e. by Aug 08 2025 11:59PM.

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Heather Van Epps, PhD

Consulting Editor

PLOS Medicine

PLoS Med. doi: 10.1371/journal.pmed.1004837.r002

Decision Letter 1

Alison Farrell

12 Sep 2025

Dear Dr Zhang,

Many thanks for submitting your manuscript "Social and Health System Determinants of Maternal Mortality in More and Less Developed Regions of China: Implications for SDG 2030" (PMEDICINE-D-25-02747R1) to PLOS Medicine. The paper has been reviewed by subject experts and a statistician; their comments are included below and can also be accessed here: [LINK]

As you will see, the reviewers find the work interesting and timely and have provide suggestions to improve the analyses, presentation and potential impact. After discussing the paper with the editorial team and an academic editor with relevant expertise, I'm pleased to invite you to revise the paper in response to the reviewers' comments. We plan to send the revised paper to some or all of the original reviewers, and we cannot provide any guarantees at this stage regarding publication.

In addition to these revisions, you may need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests shortly. If you do not receive a separate email within a few days, please assume that checks have been completed, and no additional changes are required.

When you upload your revision, please include a point-by-point response that addresses all of the reviewer and editorial points, indicating the changes made in the manuscript and either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please also be sure to check the general editorial comments at the end of this letter and include these in your point-by-point response. When you resubmit your paper, please include a clean version of the paper as the main article file and a version with changes tracked as a marked-up manuscript. It may also be helpful to check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper.

We ask that you submit your revision by Sep 29 2025 11:59PM. However, if this deadline is not feasible, please contact me by email, and we can discuss a suitable alternative.

Don't hesitate to contact me directly with any questions.

Best regards,

Alison

Alison Farrell, PhD

Senior Editor

PLOS Medicine

afarrell@plos.org

-----------------------------------------------------------

Comments from the reviewers:

Reviewer #1: This article uses Bayesian kernel machine regression to perform variable selection on various provincial-level social determinants of health to examine associations with longitudinal maternal mortality markers. In general, this was a very well-written article, with strong visualizations, and utilizing a statistical methodology that has been used before in similar settings. I appreciated the clarity in which the variables were defined, and the background provided on China's impressive improvements in maternal mortality and public health were helpful and provided great context for the study. However, there is potential room for improvement, as outlined in the suggestions below. I am not necessarily asking for entire reanalyses of the data or even that the suggestions be run, but would like to engage with the authors and understand more regarding the following points.

I have two main comments: the first being that I didn't think the GP model was particularly necessary (in that use of this model and the interpretations that came out of it might be obfuscating simpler relationships), and the second being that some of the most relevant epidemiological interpretations weren't actually provided in terms of direct interpretations of exposure/response relatonships. There were only nine predictors considered in the study, with "exposure group 2" comprising only a single variable! I'm curious what a model that simply used all predictors or one with sociologically-informed variable selection with groups would look like; my guess is pretty similar in terms of coefficient estimates to the one presented after the complex statistical machinery.

As mentioned, another potential area for improvement would be in interpretability of the model. Although it is perhaps of interest to know that an increase in hospital relivery rate from Q1 to Q3 "contributed the most to the exposure-response relationship," this provides only part of the picture. Is it possible that the antenatal care rate going from Q1 to Q3 also "contributes" quite a great deal to this relationship? What are the estimated posterior means and 95% credible sets there? Additionally, it would have been helpful to have some sort of real-world idea of what going from Q1 to Q3 entailed for this variable in the text instead of having to look it up in Table 1.

On this note, the real-world implications of Q1 to Q3 changes also vary quite a bit. For instance, in Western China from 2004-2012, Q1-Q3 was 77.2 - 97.4 for hospital delivery rate, whereas in Eastern China from 2013-2020 that same Q1-Q3 was 99.9-100. To me, this seems to drive the story - there simply wasn't any meaningful variability in Eastern China in that time period, thus driving the storyline that hospital delivery "wasn't important" in explaining exposure/outcome relationships for that model. Would it not simply be more relevant to present results from a simpler model? Even a straightforward mixed-effects longitudinal model would be able to quantify something more directly relevant to public health/sociological understanding, something along the lines of "for this region in this time period, adjusting for xyz, each additional percentage of hospital delivery % was associated with an ##% reduction in maternal death due to hemorrhage" or something, instead of identifying the "most important variable" when going to Q1-Q3, knowing that this difference might be quite different across variables or even irrelevant in other cases (cf. the example earlier in this paragraph).

An observation regarding the modeling procedure is that fitting the Eastern/Western separately removes ability to borrow information (which the Bayesian paradigm is great for) across provinces in the other region, essentially fitting interactions across all variables. This is fine if this was the intention, but I also wonder what "one big model" with all provinces and regions would have looked like together, perhaps with terms/interactions corresponding to region. As a personal curiosity, I also wonder what an analysis weighted by population would look like - as is, Qinghai has the same weight as Sichuan in the model, though on absolute public health terms regarding mortality, it seems that we might care much more about Sichuan or that its contribution should be weighted much more. It may not be worth running the model, but a sensitivity analysis (perhaps of a simple mixed-effects model) may be interesting from a scientific point of view.

Finally, use of words like "attributable" or "most important" may unintentionally imply a causal relationship with these social determinants of health and reduced maternal mortality despite these being retrospective observational data. I would strongly encourage you to use more associational language in order not to inadvertently mislead less savvy readers.

As minor suggestions, I would have appreciated a bit more information regarding the imputation approach for missing determinants of health (what specific model was used? How much data was missing? Why not an imputation method that allows for valid inference, e.g., MICE, instead of deterministic imputation according to a model, thus under-representing variability both in terms of the variable itself but also through the fact that imputation had to be performed?). Furthermore, the term "posterior SD" does not seem to apply (e.g., in line 259) - my guess is that you are presenting 95% HPD credible intervals for the posterior mean. The term "interquartile range" is used incorrectly; the IQR specifically refers to the difference between Q1 and Q3, and so I recommend simply reporting [Q1 - Q3] instead. It does not appear that fiscal expenditure is adjusted for inflation - the nominal values reported mean quite different things, and perhaps would look less stark if adjusted to, say, 2010 RMB. Finally, the text in the figures were quite hard to read - please ensure that they are available at sufficient DPI in the final submission.

Reviewer #2: Manuscript title: Social and Health System Determinants of Maternal Mortality in More and Less Developed Regions of China: Implications for SDG 2030

Journal: PLOS Medicine

________________________________________

Summary

This manuscript analyzes maternal mortality in China from 2004-2020, using provincial-level data and Bayesian kernel machine regression (BKMR) to evaluate the relative contributions of social and health system determinants. The authors highlight hospital delivery, antenatal care, health financing, income, and urbanization as key factors behind China's remarkable reduction in maternal mortality, and frame lessons for low- and middle-income countries (LMICs) still striving toward SDG target 3.1.

The topic is timely and highly relevant for global health policy. The innovative application of BKMR to maternal health research is a clear methodological strength, and the East/West and pre-/post-2013 comparisons are well thought out. However, the manuscript would benefit from sharpening its narrative focus, simplifying presentation of results, and expanding the discussion of policy implications — particularly around quality of care. In addition, the quality of data collected from the national statistics especial in the 2010s may be problematic.

________________________________________

Major Comments

The manuscript has put overemphasis on descriptive results, which have been widely reported over the past decade. For example, Figures 1-3 and much of the Results reiterate well-documented declines in maternal mortality. These sections could be streamlined to allow more focus on the determinants analysis, which is the paper's main value-added.

The study used many data from the national annual statistics books published by the Chinese central and local governments. However, we all know that the quality of such routine data has been problematic. For example, China has got a huge population of rural-to-urban migrant workers (over 200 million, based on some reports). The health statistics in one province may not include those migrant populations who do not have official resident status. Such a practice was very common in the early 2000s in China. I am wondering how the authors addressed such a challenge. In addition, different provinces/regions may have different working definition of "urbanization rate". Is the indicator only the use of official residence status? Or the calculation has considered the rural-to-urban population living in cities, but registered in the rural areas?

It should also give clearer interpretation of BKMR Findings: while GroupPIP and CondPIP are presented, their practical meaning is not sufficiently explained. For a broad PLOS Medicine readership, please clarify: what does a CondPIP of 0.9 mean in policy terms? How large are these effects compared to known biomedical interventions?

The discussion highlights antenatal care and facility births, but the issue of quality of care is underdeveloped. WHO has emphasized that coverage alone is insufficient; over-medicalization (e.g., rising cesarean section rates in China) and quality deficits are important considerations. A more balanced discussion would strengthen the global relevance.

________________________________________

Minor Comments

Abstract: Too dense. Simplify by foregrounding the five main determinants identified (hospital delivery, antenatal care, financing, urbanization, income) and reduce descriptive statistics.

Figures: Figure 4 is central but overloaded. Consider splitting into separate panels (total vs. cause-specific mortality) for clarity.

Terminology: Ensure consistent use of "maternal mortality ratio (MMR)" vs. "maternal mortality" throughout. Clarify terms like "coexisting medical conditions" vs. "indirect causes."

Discussion Structure: Could be tightened by reducing repetition and focusing on three big global lessons (coverage + quality; system/urbanization context).

Reviewer #3: The study aimed to estimate the most contributing social determinants and health system-related factors to the decline in total and cause-specific maternal mortality in China from 2004 to 2020, by use of Bayesian kernel machine regression (BKMR) to tackle the collinearity of covariates. The data originated from Global Burden of Diseases, Injuries, and Risk Factors Study 2021 (GBD 2021), the National Health Statistics Yearbooks, and the China Statistical Yearbooks. This study is of significant importance. It will provide important experience for other regions to achieving MMR decline target in SDGs. Overall, the manuscript is well written. I just have minor comments.

1.The dependent variable in this study is time-dependent. Please explain how this issue is addressed in statistical analysis.

2.Have authors tried to analyze the contributing determinants by rural and urban areas?

Any attachments provided with reviews can be seen via the following link: [LINK]

--------------------------------------------------------- ---

General editorial requests:

(Note: not all will apply to your paper, but please check each item carefully)

* Please include an Author Summary after the Abstract. See:https://journals.plos.org/plosmedicine/s/revising-your-manuscript

* We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. Please do not add or remove authors without first discussing this with the handling editor. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

* Please upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, we strongly recommend that you use PLOS's NAAS tool (https://ngplosjournals.pagemajik.ai/artanalysis) to test your figure files. NAAS can convert your figure files to the TIFF file type and meet basic requirements (such as print size, resolution), or provide you with a report on issues that do not meet our requirements and that NAAS cannot fix.

After uploading your figures to PLOS's NAAS tool - https://ngplosjournals.pagemajik.ai/artanalysis, NAAS will process the files provided and display the results in the "Uploaded Files" section of the page as the processing is complete.

If the uploaded figures meet our requirements (or NAAS is able to fix the files to meet our requirements), the figure will be marked as "fixed" above. If NAAS is unable to fix the files, a red "failed" label will appear above.

When NAAS has confirmed that the figure files meet our requirements, please download the file via the download option, and include these NAAS processed figure files when submitting your revised manuscript.

* Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information (web or email address) for obtaining the data. Please note that a study author cannot be the contact person for the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

* We expect all researchers with submissions to PLOS in which author-generated code underpins the findings in the manuscript to make all author-generated code available without restrictions upon publication of the work. In cases where code is central to the manuscript, we may require the code to be made available as a condition of publication. Authors are responsible for ensuring that the code is reusable and well documented. Please make any custom code available, either as part of your data deposition or as a supplementary file. Please add a sentence to your data availability statement regarding any code used in the study, e.g. "The code used in the analysis is available from Github [URL] and archived in Zenodo [DOI link]" Please review our guidelines at https://journals.plos.org/plosmedicine/s/materials-software-and-code-sharing and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse. Because Github depositions can be readily changed or deleted, we encourage you to make a permanent DOI'd copy (e.g. in Zenodo) and provide the URL.

* Please ensure that the study is reported according to the appropriate guideline (e.g. STROBE) and include the completed checklist as Supporting Information. When completing the checklist, please use section and paragraph numbers, rather than page numbers. Please add the following statement, or similar, to the Methods: "This study is reported as per [XXXX] guideline (S1 Checklist)."

FORMATTING - GENERAL

* Abstract: Please structure your abstract using the PLOS Medicine headings (Background, Methods and Findings, Conclusions). Please combine the Methods and Findings sections into one section.

* At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Ideally each sub-heading should contain 2-3 single sentence, concise bullet points containing the most salient points from your study. In the final bullet point of 'What Do These Findings Mean?', please include the main limitations of the study in non-technical language. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary.

* Please express the main results with 95% CIs as well as p values. When reporting p values please report as p<0.001 and where higher as the exact p value p=0.002, for example. Throughout, suggest reporting statistical information as follows to improve clarity for the reader "22% (95% CI [13%,28%]; p</=)". Please be sure to define all numerical values at first use.

* Please include page numbers and line numbers in the manuscript file. Use continuous line numbers (do not restart the numbering on each page).

* Please cite the reference numbers in square brackets. Citations should precede punctuation.

FIGURES AND TABLES

* Please provide titles and legends for all figures and tables (including those in Supporting Information files).

* Please define all abbreviations used in each figure/table (including those in Supporting Information files).

* Please consider avoiding the use of red and green in order to make your figure more accessible to those with color blindness.

SUPPLEMENTARY MATERIAL

* Please note that supplementary material will be posted as supplied by the authors. Therefore, please amend it according to the relevant comments outlined here.

* Please cite your Supporting Information as outlined here: https://journals.plos.org/plosmedicine/s/supporting-information

REFERENCES

* PLOS uses the numbered citation (citation-sequence) method and first six authors, et al.

* Please ensure that journal name abbreviations match those found in the National Center for Biotechnology Information (NCBI) databases (http://www.ncbi.nlm.nih.gov/nlmcatalog/journals), and are appropriately formatted and capitalised.

* Where website addresses are cited, please include the complete URL and specify the date of access (e.g. [accessed: 12/06/2023]).

* Please also see https://journals.plos.org/plosmedicine/s/submission-guidelines#loc-references for further details on reference formatting.

OBSERVATIONAL STUDIES

* Abstract: Please include the study design, population and setting, number of participants, years during which the study took place (enrollment and follow up), length of follow up, and main outcome measures.

* Please ensure that the study is reported according to the STROBE (or appropriate STOBE extension) guideline (available from: https://www.equator-network.org/reporting-guidelines/strobe) and include the completed STROBE (or STROBE extension) checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist)." When completing the checklist, please use section and paragraph numbers, rather than page numbers.

* [FOR POPULATION HEALTH/REGISTRY STUDIES] Please ensure that the study is reported according to the RECORD guideline (available from https://www.record-statement.org) and include the completed checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Reporting of Studies Conducted using Observational Routinely-Collected Data (RECORD) guideline (S1 Checklist)." When completing the checklist, please use section and paragraph numbers, rather than page numbers.

* [FOR POPULATION HEALTH ESTIMATES] Please ensure that the study is reported according to the GATHER statement (available from https://www.equator-network.org/reporting-guidelines/gather-statement) and include the completed checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER) statement (S1 Checklist)." When completing the checklist, please use section and paragraph numbers, rather than page numbers.

* [FOR MEDIATION ANALYSES] We recommend that the study is reported according to the AGReMA statement (https://agrema-statement.org/#:~:text=AGReMA%20is%20an%20evidence%2D%20and,randomised%20trials%20and%20observational%20studies) and include the completed checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Guideline for Reporting Mediation Analyses (AGReMA) statement (S1 Checklist)." When completing the checklist, please use section and paragraph numbers, rather than page numbers.

* For all observational studies, in the manuscript text, please indicate: (1) the specific hypotheses you intended to test, (2) the analytical methods by which you planned to test them, (3) the analyses you actually performed, and (4) when reported analyses differ from those that were planned, transparent explanations for differences that affect the reliability of the study's results. If a reported analysis was performed based on an interesting but unanticipated pattern in the data, please be clear that the analysis was data driven.

* Please state in the Methods section whether the study had a prospective protocol or analysis plan. If a prospective analysis plan (from your funding proposal, IRB or other ethics committee submission, study protocol, or other planning document written before analyzing the data) was used in designing the study, please include the relevant document(s) with your revised manuscript as a Supporting Information file to be published alongside your study and cite it in the Methods section. A legend for this file should be included at the end of your manuscript. If no such document exists, please make sure that the Methods section transparently describes when analyses were planned, and when/why any data-driven changes to analyses took place. Changes in the analysis, including those made in response to peer review comments, should be identified as such in the Methods section of the paper, with rationale.

MODELLING STUDIES

The following list is derived from Geoffrey P Garnett, Simon Cousens, Timothy B Hallett, Richard Steketee, Neff Walker. Mathematical models in the evaluation of health programmes. (2011) Lancet DOI:10.1016/S0140-6736(10)61505-X:

* If pertinent, please provide a diagram that shows the model structure, including how the natural history of the disease is represented, the process and determinants of disease acquisition, and how the putative intervention could affect the system.

* Please provide a complete list of model parameters, including clear and precise descriptions of the meaning of each parameter, together with the values or ranges for each, with justification or the primary source cited and important caveats about the use of these values noted.

* Please provide a clear statement about how the model was fitted to the data, including goodness-of-fit measure, the numerical algorithm used, which parameter varied, constraints imposed on parameter values, and starting conditions.

* For uncertainty analyses, please state the sources of uncertainties quantified and not quantified [can include parameter, data, and model structure].

* Please provide sensitivity analyses to identify which parameter values are most important in the model. Uncertainty estimates seek to derive a range of credible results on the basis of an exploration of the range of reasonable parameter values. The choice of method should be presented and justified.

* Please discuss the scientific rationale for the choice of model structure and identify points where this choice could influence conclusions drawn. Please also describe the strength of the scientific basis underlying the key model assumptions.

* For studies that develop a prediction model or evaluate its performance, please ensure that the study is reported according to the TRIPOD statement (https://www.equator-network.org/reporting-guidelines/tripod-statement) and include the completed checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis Or Diagnosis (TRIPOD) statement (S1 Checklist)." For studies using machine learning, please use the TRIPOD-AI checklist. When completing the checklist, please use section and paragraph numbers, rather than page numbers.

PLoS Med. 2025 Dec 4;22(12):e1004837. doi: 10.1371/journal.pmed.1004837.r003

Author response to Decision Letter 2

28 Sep 2025

Attachment

Submitted filename: 20250928 Response letter.docx

pmed.1004837.s032.docx^{(74.6MB, docx)}

PLoS Med. doi: 10.1371/journal.pmed.1004837.r004

Decision Letter 2

Alison Farrell

7 Nov 2025

Dear Dr. Zhang,

Thank you very much for re-submitting your manuscript "Social and Health System Determinants of Maternal Mortality in More and Less Developed Regions of China: Implications for SDG 2030" (PMEDICINE-D-25-02747R2) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by 2 reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The reviewers have a few remaining comments that we ask you to address in a point-by-point response. Please note in particular that we require that you to remove all causal assertions from the manuscript and frame your results as associative, as per the comments of reviewer 1. Please remove 'important' from the Abstract, qualify determinants (i.e. determinants of what) and see below for other Abstract requirements.

Please also note that we require you to respond to the comments of reviewer 2, and please refer to this reviewer's comments from the original round of review, and provide a response in full to the points originally raised.

Please ensure that the manuscript includes a definition of Eastern and Western China.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Nov 14 2025 11:59PM.

Sincerely,

Alison Farrell, Ph.D.

Senior Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

* Please confirm that your title complies with PLOS Medicine's style. Your title must be nondeclarative and not a question. It should begin with main concept if possible. "Effect of" should be used only if causality can be inferred, i.e., for an RCT. Please place the study design ("A randomized controlled trial," "A retrospective study," "A modelling study," etc.) in the subtitle (ie, after a colon).

* Please confirm that your abstract complies with our requirements, including format (three sections: Background, Methods and Findings, and Conclusions) and providing all the information relevant to this study type https://journals.plos.org/plosmedicine/s/submission-guidelines#loc-abstract

* Please ensure that the Introduction ends with a clear description of the study question or hypothesis.

* Please ensure that all abbreviations are defined at first use throughout the text.

* Please confirm that all numbers presented in the abstract are present and identical to numbers presented in the main manuscript text.

* Please review your text for claims of novelty or primacy (e.g. 'for the first time') and remove this language. In addition, please check that any use of statistical terms (such as trend or significant) are supported by the data, and if not please remove them.

* In the last sentence of the Abstract Methods and Findings section, please describe the main limitation(s) of the study's methodology.

* In the abstract, please include the important dependent variables that are adjusted for in the analyses.

* Please convert any stacked bar charts to another data representation for example a table, or other type of graph. If that is not possible, please explain why and provide a Table containing the data in the bar chart (e.g. in Fig. 2).

In Fig. 3, please lighten the dark blue colour (Hemorrhage) for improved clarity.

* Where data points are discrete, please ensure that they are depicted in the figures as discrete data and not as a continuous line, e.g. in Fig. 1.

* Please provide the unadjusted comparisons as well as the adjusted comparisons in all relevant Tables

* Please specify the variables controlled for in all relevant Tables

Comments from Reviewers:

Reviewer #1: The authors have comprehensively addressed many of my original concerns and those of other reviewers, particularly regarding some of the additional analyses they present. In general, I am satisfied with these revisions, but did still have two lingering comments - one that I would like to see addressed, and another that might serve merely as a suggestion.

Most importantly, the language used in the model still inappropriately makes causal assertions. for instance, the word "determinant" implies that these social factors are determinants of mortality or health outcome, which is a causal association. Furthermore, the word "important" is being used as a proxy for magnitude of effect, which might not necessarily connote "importance" in terms of the concept we regularly think of. Some of the suggested language is indeed appropriate (e.g., directly stating that something was "associated with a higher value of the exposure-response relationship,"), and so I encourage the authors to use such language throughout, instead of anything that may - incorrectly or not - be construed as implying a causal relationship (e.g., "important determinant for").

As a potential suggestion, I wonder whether the current "sensitivity analyses" might actually serve better as the main analysis. For instance, the mixed model that directly addresses the question "an x increase in y is associated with a z change in w, while adjusting for the other abc..." is directly interpretable and directly addresses research questions of interest regarding the magnitude of relationships. The sensitivity analysis might then be the BKMR, which answers the potential challenge to the analysis of "we know some of these are correlated - what might "get chosen" if we only had to choose a subset of these individual predictors."

Regardless, I am satisfied with the response to original review and the revised manuscript. The only real comment I have is on the lingering use of causal language.

Reviewer #2: The authors of the manuscript has addressed most of my comments in the revised version. However, they failed to address one major comment, that is, the quality of data derived from the national health statistics published in the 2010s. The answer to my comment (No 2) was vague and did not show their understanding of potential implications for the use of such poor quality data. I understand that it is a very tough question that may not be easily addressed. Should the authors not be able to address this issue, they should at least put it as one of main limitations of the study.

Any attachments provided with reviews can be seen via the following link:

[LINK]

PLoS Med. 2025 Dec 4;22(12):e1004837. doi: 10.1371/journal.pmed.1004837.r005

Author response to Decision Letter 3

12 Nov 2025

Attachment

Submitted filename: 20251112 Response letter.docx

pmed.1004837.s033.docx^{(406.4KB, docx)}

PLoS Med. doi: 10.1371/journal.pmed.1004837.r006

Decision Letter 3

Alison Farrell

14 Nov 2025

Dear Dr Zhang,

On behalf of my colleagues and the Academic Editor, Margaret Kruk, I am pleased to inform you that we have agreed to publish your manuscript "Social and health system factors associated with maternal mortality in more and less developed regions of China: Population health estimates using provincial-level data" (PMEDICINE-D-25-02747R3) in PLOS Medicine.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

Please also address these editorial requests:

Title: replace "more and less developed regions of" with "Eastern and Western" so that the title does not age as development changes occur in China

Line 36: capitalize Eastern and Western

Line 56: delete 'the remarkable progress' and replace with 'improvements'

Line 59: revise "it is vital" to "it may be vital"

Line 337: delete 'and' before 'an'

Add the URLs for funders websites to the funding information.

Please check capitalization of Eastern and Western throughout (should be capitalized when next to China).

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Alison Farrell, Ph.D.

Senior Editor

PLOS Medicine

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Social and health system-related factors of maternal mortality.

Note: Ob/Gyn, obstetrics and gynecology.

(DOCX)

pmed.1004837.s001.docx^{(16.9KB, docx)}

S2 Table. Cause-specific maternal mortality rate in China, 2004–2020.

(DOCX)

pmed.1004837.s002.docx^{(15.6KB, docx)}

S3 Table. Cause-specific maternal mortality fractions in China, 2004–2020.

(DOCX)

pmed.1004837.s003.docx^{(15.8KB, docx)}

S4 Table. Group and conditional posterior inclusion probabilities for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection.

(DOCX)

pmed.1004837.s004.docx^{(15KB, docx)}

S5 Table. Group and conditional posterior inclusion probabilities for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection.

(DOCX)

pmed.1004837.s005.docx^{(15.2KB, docx)}

S6 Table. Group and conditional posterior inclusion probabilities for each factor in Eastern China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection.

(DOCX)

pmed.1004837.s006.docx^{(15.1KB, docx)}

S7 Table. Group and conditional posterior inclusion probabilities for each factor in Western China, 2013––2020, using Bayesian Kernel Machine Regression hierarchical variable selection.

(DOCX)

pmed.1004837.s007.docx^{(15.1KB, docx)}

S8 Table. Effects of factors on maternal mortality, using univariate linear mixed-effects models weighted by the number of livebirths.

Note: Effects are shown as estimates and the 95% confidence intervals in the linear mixed effects models.

(DOCX)

pmed.1004837.s008.docx^{(15.9KB, docx)}

S9 Table. Effects of factors on maternal mortality, using multivariate linear mixed-effects models weighted by the number of livebirths.

(DOCX)

pmed.1004837.s009.docx^{(16KB, docx)}

(DOCX)

pmed.1004837.s010.docx^{(15KB, docx)}

(DOCX)

pmed.1004837.s011.docx^{(15KB, docx)}

(DOCX)

pmed.1004837.s012.docx^{(15KB, docx)}

(DOCX)

pmed.1004837.s013.docx^{(15KB, docx)}

(DOCX)

pmed.1004837.s014.docx^{(15KB, docx)}

(DOCX)

pmed.1004837.s015.docx^{(15KB, docx)}

S16 Table. Data inputs for BKMR.

(XLSX)

pmed.1004837.s016.xlsx^{(75.5KB, xlsx)}

S1 Fig. Spearman correlation coefficient matrix for the non-medical factors.

(TIFF)

pmed.1004837.s017.tiff^{(908.5KB, tiff)}

S2 Fig. Map of the provincial-level maternal mortality in mainland China in 2004, 2011 and 2020.

Note: The source of the basemap is: https://cloudcenter.tianditu.gov.cn/administrativeDivision.

(TIFF)

pmed.1004837.s018.tiff^{(1.1MB, tiff)}

S3 Fig. Single-exposure risk measures for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection.

(TIFF)

pmed.1004837.s019.tiff^{(28.2MB, tiff)}

S4 Fig. Single-exposure risk measures for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection.

(TIFF)

pmed.1004837.s020.tiff^{(28.3MB, tiff)}

S5 Fig. Single-exposure risk measures for each factor in Eastern China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection.

(TIFF)

pmed.1004837.s021.tiff^{(28.7MB, tiff)}

S6 Fig. Single-exposure risk measures for each factor in Western China, 2013–2020, using Bayesian Kernel Machine Regression hierarchical variable selection.

(TIFF)

pmed.1004837.s022.tiff^{(28.4MB, tiff)}

(TIFF)

pmed.1004837.s023.tiff^{(9.7MB, tiff)}

(TIFF)

pmed.1004837.s024.tiff^{(28MB, tiff)}

(TIFF)

pmed.1004837.s025.tiff^{(28.3MB, tiff)}

(TIFF)

pmed.1004837.s026.tiff^{(28.5MB, tiff)}

(TIFF)

pmed.1004837.s027.tiff^{(28.5MB, tiff)}

(TIFF)

pmed.1004837.s028.tiff^{(6.6MB, tiff)}

S13 Fig. Single-exposure risk measures for each factor in Eastern China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with missing data imputed by MICE.

(TIFF)

pmed.1004837.s029.tiff^{(28.5MB, tiff)}

S14 Fig. Single-exposure risk measures for each factor in Western China, 2004–2012, using Bayesian Kernel Machine Regression hierarchical variable selection with missing data imputed by MICE.

(TIFF)

pmed.1004837.s030.tiff^{(28.6MB, tiff)}

S1 Checklist. Checklist of information that should be included in new reports of global health estimates.

(DOCX)

pmed.1004837.s031.docx^{(31KB, docx)}

Attachment

Submitted filename: 20250928 Response letter.docx

pmed.1004837.s032.docx^{(74.6MB, docx)}

Attachment

Submitted filename: 20251112 Response letter.docx

pmed.1004837.s033.docx^{(406.4KB, docx)}

Data Availability Statement

[pmed.1004837.ref001] 1.World Health Organization. Trends in maternal mortality 2000 to 2020: estimates by WHO, UNICEF, UNFPA, World Bank Group and UNDESA/Population Division. Geneva: World Health Organization; 2023. [Google Scholar]

[pmed.1004837.ref002] 2.Qiao J, Wang Y, Li X, Jiang F, Zhang Y, Ma J, et al. A Lancet Commission on 70 years of women’s reproductive, maternal, newborn, child, and adolescent health in China. Lancet. 2021;397(10293):2497–536. doi: 10.1016/S0140-6736(20)32708-2 [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref003] 3.Shi H, Chen L, Wei Y, Chen X, Zhao Y. Improving maternal healthcare further in China at a time of low maternal mortality. BMJ. 2024;386:e078640. doi: 10.1136/bmj-2023-078640 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref004] 4.Partnership for Maternal, Newborn and Child Health, World Health Organization, the World Bank Group, Alliance for Health Policy and Systems Research. Success Factors for Women’s and Children’s Health: Policy and programme highlights from 10 fast-track countries. Geneva: World Health Organization; 2014. [Google Scholar]

[pmed.1004837.ref005] 5.Department of Maternal and Child Health. Report on the development of maternal and child health in China (2019). National Health Commission of China. 2019 May 27 [Cited 2025 Apr 9]. Available from: https://www.nhc.gov.cn/fys/c100077/201905/142632c029194e42a3cca755e336db80.shtml#.

[pmed.1004837.ref006] 6.Souza JP, Day LT, Rezende-Gomes AC, Zhang J, Mori R, Baguiya A, et al. A global analysis of the determinants of maternal health and transitions in maternal mortality. Lancet Glob Health. 2024;12(2):e306–16. doi: 10.1016/S2214-109X(23)00468-0 [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref007] 7.Yu L, Liu W, Wang X, Ye Z, Tan Q, Qiu W, et al. A review of practical statistical methods used in epidemiological studies to estimate the health effects of multi-pollutant mixture. Environ Pollut. 2022;306:119356. doi: 10.1016/j.envpol.2022.119356 [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref008] 8.Lazarevic N, Barnett AG, Sly PD, Knibbs LD. Statistical Methodology in Studies of Prenatal Exposure to Mixtures of Endocrine-Disrupting Chemicals: A Review of Existing Approaches and New Alternatives. Environ Health Perspect. 2019;127(2):26001. doi: 10.1289/EHP2207 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref009] 9.National Health Commission. National Health Statistics Yearbook. National Health Commission. 2025 Jan 24 [Cited 2025 Mar 31]. Available from: http://www.nhc.gov.cn/mohwsbwstjxxzx/tjzxtjcbw/tjsj_list.shtml

[pmed.1004837.ref010] 10.Institute for Health Metrics and Evaluation (IHME). GBD Results. Seattle, WA: IHME, University of Washington. 2024. [Cited 2024 Oct 31]. Available from: https://vizhub.healthdata.org/gbd-results/ [Google Scholar]

[pmed.1004837.ref011] 11.National Bureau of Statistics. China Statistical Yearbook. National Bureau of Statistics. 2025 Mar 27 [Cited 2025 Apr 8]. Available from: https://www.stats.gov.cn/sj/ndsj/

[pmed.1004837.ref012] 12.National Health Commission. National Health Statistics Yearbook 2021. Beijing: Peking Union Medical College Press; 2021. [Google Scholar]

[pmed.1004837.ref013] 13.Guo H, Wang X, Wu B, Li X. Cognizing population density demarcative line (Hu Huanyong-Line) based on space technology. Bull Chin Acad Sci. 2016;31(12):1385–94. [Google Scholar]

[pmed.1004837.ref014] 14.Wubetie HT. Missing data management and statistical measurement of socio-economic status: application of big data. J Big Data. 2017;4(1). doi: 10.1186/s40537-017-0099-y [DOI] [Google Scholar]

[pmed.1004837.ref015] 15.Liang J, Li X, Kang C, Wang Y, Kulikoff XR, Coates MM. Maternal mortality ratios in 2852 Chinese counties, 1996-2015, and achievement of Millennium Development Goal 5 in China: a subnational analysis of the Global Burden of Disease Study 2016. Lancet. 2019;393(10168):241–52. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref016] 16.World Health Organization, World Bank, United Nations Population Fund & United Nations Children’s Fund. Maternal mortality in 2005: estimates developed by WHO, UNICEF, UNFPA, and The World Bank. Geneva: World Health Organization; 2007. [Google Scholar]

[pmed.1004837.ref017] 17.Gao Y, Zhou H, Singh NS, Powell-Jackson T, Nash S, Yang M, et al. Progress and challenges in maternal health in western China: a Countdown to 2015 national case study. Lancet Glob Health. 2017;5(5):e523–36. doi: 10.1016/S2214-109X(17)30100-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref018] 18.International Monetary Fund. Inflation, consumer prices (annual %) - China. World Bank Group. 2020 [Cited 2025 Sep 18]. Available from: https://data.worldbank.org/indicator/FP.CPI.TOTL.ZG?end=2020&locations=CN&start=2004&view=chart

[pmed.1004837.ref019] 19.World Health Organization. WHO recommendations on antenatal care for a positive pregnancy experience. Geneva: World Health Organization; 2016. [PubMed] [Google Scholar]

[pmed.1004837.ref020] 20.Li J, Liang J, Wang J, Ren Z, Yang D, Wang Y, et al. Spatiotemporal trends and ecological determinants in maternal mortality ratios in 2,205 Chinese counties, 2010-2013: A Bayesian modelling analysis. PLoS Med. 2020;17(5):e1003114. doi: 10.1371/journal.pmed.1003114 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref021] 21.Gage AD, Fink G, Ataguba JE, Kruk ME. Hospital delivery and neonatal mortality in 37 countries in sub-Saharan Africa and South Asia: An ecological study. PLoS Med. 2021;18(12):e1003843. doi: 10.1371/journal.pmed.1003843 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref022] 22.Chaka EE, Mekurie M, Abdurahman AA, Parsaeian M, Majdzadeh R. Association between place of delivery for pregnant mothers and neonatal mortality: a systematic review and meta-analysis. Eur J Public Health. 2020;30(4):743–8. doi: 10.1093/eurpub/ckz060 [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref023] 23.Chinkhumba J, De Allegri M, Muula AS, Robberstad B. Maternal and perinatal mortality by place of delivery in sub-Saharan Africa: a meta-analysis of population-based cohort studies. BMC Public Health. 2014;14:1014. doi: 10.1186/1471-2458-14-1014 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref024] 24.Filippi V, Ronsmans C, Campbell OMR, Graham WJ, Mills A, Borghi J, et al. Maternal health in poor countries: the broader context and a call for action. Lancet. 2006;368(9546):1535–41. doi: 10.1016/S0140-6736(06)69384-7 [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref025] 25.Ward ZJ, Atun R, King G, Sequeira DMello B, Goldie SJ. Global maternal health country typologies: A framework to guide policy. PLOS Glob Public Health. 2024;4(11):e0003867. doi: 10.1371/journal.pgph.0003867 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref026] 26.Hasan MM, Magalhaes RJS, Fatima Y, Ahmed S, Mamun AA. Levels, trends, and inequalities in using institutional delivery services in low- and middle-income countries: a stratified analysis by facility type. Glob Health Sci Pract. 2021;9(1):78–88. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref027] 27.Jiwani SS, Amouzou-Aguirre A, Carvajal L, Chou D, Keita Y, Moran AC, et al. Timing and number of antenatal care contacts in low and middle-income countries: Analysis in the Countdown to 2030 priority countries. J Glob Health. 2020;10(1):010502. doi: 10.7189/jogh.10.010502 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref028] 28.Boerma T, Campbell OMR, Amouzou A, Blumenberg C, Blencowe H, Moran A, et al. Maternal mortality, stillbirths, and neonatal mortality: a transition model based on analyses of 151 countries. Lancet Glob Health. 2023;11(7):e1024–31. doi: 10.1016/S2214-109X(23)00195-X [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref029] 29.Bohren MA, Hunter EC, Munthe-Kaas HM, Souza JP, Vogel JP, Gülmezoglu AM. Facilitators and barriers to facility-based delivery in low- and middle-income countries: a qualitative evidence synthesis. Reprod Health. 2014;11(1):71. doi: 10.1186/1742-4755-11-71 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref030] 30.Finlayson K, Downe S. Why do women not use antenatal services in low- and middle-income countries? A meta-synthesis of qualitative studies. PLoS Med. 2013;10(1):e1001373. doi: 10.1371/journal.pmed.1001373 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref031] 31.Alam CE, Abou-Abbas L, Ramadan MS, Asmar MK. Exploring the barriers to accessing antenatal care at the primary health care center level of a tertiary hospital in Lebanon: a qualitative study. BMC Health Serv Res. 2025;25(1):304. doi: 10.1186/s12913-025-12444-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref032] 32.Deo KK, Paudel YR, Khatri RB, Bhaskar RK, Paudel R, Mehata S, et al. Barriers to Utilization of Antenatal Care Services in Eastern Nepal. Front Public Health. 2015;3:197. doi: 10.3389/fpubh.2015.00197 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref033] 33.Ward ZJ, Atun R, King G, Sequeira Dmello B, Goldie SJ. A simulation-based comparative effectiveness analysis of policies to improve global maternal health outcomes. Nat Med. 2023;29(5):1262–72. doi: 10.1038/s41591-023-02311-w [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref034] 34.Fink G, Cohen J. Delivering quality: safe childbirth requires more than facilities. Lancet Glob Health. 2019;7(8):e990–1. doi: 10.1016/S2214-109X(19)30193-7 [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref035] 35.Nihal S, Shekhar C. An assessment of adequate quality antenatal care and its determinants in India. BMC Pregnancy Childbirth. 2024;24(1):698. doi: 10.1186/s12884-024-06806-3 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref036] 36.Mandiwa C, Namondwe B. Assessment of quality of antenatal care services and associated factors in Malawi: Insights from a nationwide household survey. PLoS One. 2024;19(6):e0305294. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref037] 37.Yu C. Analysis of the effects of implementing the Reducing maternal mortality and eliminating neonatal tetanus program. Maternal and Child Health Care of China. 2004;14:28–9. [Google Scholar]

[pmed.1004837.ref038] 38.Peng Y. Implementing the reducing maternal mortality and eliminating neonatal tetanus program to achieve the goals of the outlines of women’ and children’s development. Maternal and Child Health Care of China. 2000;15. [Google Scholar]

[pmed.1004837.ref039] 39.Wu C, Yang G, Wang W, Yang M. Interim assessment and analysis report on the reducing maternal mortality and eliminating neonatal tetanus program. Maternal and Child Health Care of China. 2002;:19–20. [Google Scholar]

[pmed.1004837.ref040] 40.Ministry of Health of the People’s Republic of China. Notice of “Management Methods for the Maternal Health Care Service” and “Standards For Maternal Health Care Service” Issued by Ministry of Health of the People’s Republic of China. Ministry of Health of the People’s Republic of China 2011 Jul 8 [Cited 2024 Oct 31]. Available from: https://www.nhc.gov.cn/zwgk/wtwj/201304/84ec2d5d37a54398ace49df48ca91c83.shtml

[pmed.1004837.ref041] 41.Jeong W, Jang S-I, Park E-C, Nam JY. The Effect of Socioeconomic Status on All-Cause Maternal Mortality: A Nationwide Population-Based Cohort Study. Int J Environ Res Public Health. 2020;17(12):4606. doi: 10.3390/ijerph17124606 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004837.ref042] 42.ACOG Committee. ACOG Committee Opinion No. 586: Health disparities in rural women. Obstet Gynecol. 2014;123(2 Pt 1):384–8. [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref043] 43.Kozhimannil KB, Interrante JD, Henning-Smith C, Admon LK. Rural-urban differences in severe maternal morbidity and mortality in the US, 2007-15. Health Aff (Millwood). 2019;38(12):2077–85. [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref044] 44.World Health Organization. Strategies towards ending preventable maternal mortality (EPMM). Geneva: World Health Organization. 2015. [Google Scholar]

[pmed.1004837.ref045] 45.Kurowski C, Schmidt M, Kumar A, Mieses J, Gabani J. Government Health Spending Trends Through 2023: Peaks, Declines, and Mounting Risks. World Bank. 2024. [Google Scholar]

[pmed.1004837.ref046] 46.Kurowski C, Schmidt M, Evans DB, Tandon A, Eozenou PHV, Cain JS. Government health spending outlook - projections through 2029 diverging fiscal pressures, uneven constraints. World Bank. 2024. [Google Scholar]

[pmed.1004837.ref047] 47.Thaddeus S, Maine D. Too far to walk: maternal mortality in context. Soc Sci Med. 1994;38(8):1091–110. doi: 10.1016/0277-9536(94)90226-7 [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref048] 48.Miller S, Abalos E, Chamillard M, Ciapponi A, Colaci D, Comandé D, et al. Beyond too little, too late and too much, too soon: a pathway towards evidence-based, respectful maternity care worldwide. Lancet. 2016;388(10056):2176–92. doi: 10.1016/S0140-6736(16)31472-6 [DOI] [PubMed] [Google Scholar]

[pmed.1004837.ref049] 49.Cresswell JA, Alexander M, Chong MYC, Link HM, Pejchinovska M, Gazeley U, et al. Global and regional causes of maternal deaths 2009-20: a WHO systematic analysis. Lancet Glob Health. 2025;13(4):e626–34. doi: 10.1016/S2214-109X(24)00560-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Social and health system factors associated with maternal mortality in Eastern and Western China: Population health estimates using provincial-level data

Xiaojing Zeng

Dongjian Yang

Shiyang Li

Xiaolin Hua

Yanlin Wang

Jun Zhang

Zhiwei Liu

Roles

Abstract

Background

Methods and findings

Conclusions

Author summary

Why was this study done?

What did the researchers do and find?

What do these findings mean?

Introduction

Methods

Data source

Definition of variables

Statistical analysis

Results

Trends in maternal mortality in Eastern and Western China from 2004 to 2020

Fig 1. Trends in maternal mortality in Eastern and Western China during 2004-2020.

Table 1. Characteristics of social and health-system factors in Eastern and Western China during 2004-2012 and 2013–2020.

Fig 2. Cause-specific maternal mortality rate (left panel) and fractions (right panel) in China, 2004–2020.

Fig 3. Cause-specific maternal mortality rate for hemorrhage, coexisting medical diseases, and HDP in Eastern and Western China, 2004–2020.

Factors associated with total and cause-specific maternal mortality in Eastern and Western China

Sensitivity analyses

Discussion

Supporting information

Abbreviation:

Data Availability

Funding Statement

References

Decision Letter 0

Heather Van Epps

Roles

Decision Letter 1

Alison Farrell

Roles

Author response to Decision Letter 2

Decision Letter 2

Alison Farrell

Roles

Author response to Decision Letter 3

Decision Letter 3

Alison Farrell

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases