Global, Regional, and National Epidemiology of Postpartum Hemorrhage (1990–2021): A Statistical Analysis of Incidence, Mortality, and DALYs

ABSTRACT

Background and Aims

Postpartum hemorrhage (PPH) is a leading global cause of maternal mortality and morbidity. This study utilizes global burden of disease (GBD) data to assess its long‐term trends in incidence, mortality, and disability‐adjusted life years (DALYs).

Methods

Using data from the GBD, we analyzed age‐standardized incidence, mortality, and DALY rates among reproductive‐aged women (15–49 years). Trend analysis was conducted via Joinpoint regression to estimate annual percentage changes (APCs), while future burden projections were modeled using Bayesian age‐period‐cohort (BAPC) analysis.

Results

From 1990 to 2021, the global age‐standardized incidence rate (ASIR) of PPH declined significantly from 998.26 (95% UI: 654.55–1,438.33) to 722.16 (95% UI: 482.99–1,022.71) per 100,000 population. The estimated annual percentage change (EAPC) was −0.83% (95% UI: −0.90% to −0.77%). Projections suggest this downward trend will continue, with ASIR expected to reach 321.17 (95% UI: 302.55–359.30) by 2035. In 2021, women aged 20–24 years continued to exhibit the highest DALYs, amounting to 749,063 years (95% UI, 629,108.37–894,035.97). From 1990 to 2021, the proportion of DALYs among women of reproductive age attributable to iron deficiency‐related PPH increased from 17.19% (95% UI, 8.65% to 21.48%) to 17.96% (95% UI, 8.67% to 23.95%).

Conclusion

A sustained decline in the incidence of PPH over the past thirty years—a trend projected to continue through 2035, reflecting the ongoing effectiveness of public health interventions. We also found a potential association with iron deficiency, alongside a consistently higher disease burden observed among women aged 20–24 years.

1. Introduction

Postpartum hemorrhage (PPH), defined as blood loss of at least 500 mL following vaginal delivery or 1,000 mL after cesarean section within 24 h of birth, stands as a predominant contributor to maternal mortality globally [1, 2]. The management of PPH, primarily stemming from uterine atony, placental disorders, or coagulopathy, involves uterotonics (e.g., oxytocin, methylergometrine), uterine massage, and transfusion support [3]. Annually, an estimated 14 million cases occur worldwide [4], resulting in approximately 70,000 deaths and constituting 27% of all maternal mortality [5]. Survivors often contend with significant long‐term sequelae, including anemia, postpartum depression, and post‐traumatic stress disorder [6]. This substantial burden is unevenly distributed, heavily influenced by socioeconomic disparities and variability in healthcare access and quality [7, 8].

A precise understanding of the epidemiological trends and evolving burden of PPH is therefore fundamental for guiding resource allocation and shaping effective public health strategies. Despite its recognized importance, systematic analyses of PPH's global and temporal patterns remain insufficient. The World Health Organization's Sustainable Development Goal 3.1 (SDG 3.1) aims to substantially reduce the global maternal mortality ratio [9, 10]. However, persistent data gaps, especially in high‐burden regions, hinder the accurate tracking of progress against this target. Consequently, comprehensive epidemiological evidence is urgently required.

The Global Burden of Disease Study (GBD) employs standardized methodologies to consistently evaluate the health loss attributable to major diseases, injuries, and risk factors, providing authoritative data for comparisons across time and geography [11]. The GBD provides estimates from 1990 onwards, covering 204 countries and territories, 371 diseases and injuries, and 3,499 related health outcomes [12].

To address the current evidence gap, this study leverages data from GBD 2021. We aim to analyze the incidence, mortality, and disability‐adjusted life years (DALYs) attributable to PPH among women aged 15–49 years from 1990 to 2021. Furthermore, we project the future burden to 2035. This analysis seeks to delineate the long‐term trends and regional variations in PPH burden, thereby providing critical evidence to inform targeted interventions and advance efforts toward achieving SDG 3.1.

2. Methods

2.1. Data Source

The GBD study systematically quantifies population health using a standardized framework that covers 371 diseases and injuries and 88 risk factors across 204 countries and territories. This surveillance system supports comparative analyses of health loss over time, across regions, and among demographic groups, providing policymakers with robust evidence for priority setting. To do so, GBD integrates over 15,000 heterogeneous data sources. These include vital registration records, surveys, and clinical records. The data are processed through advanced statistical models, such as Bayesian hierarchical spatial‐temporal modeling, to address data gaps and ensure consistency [13]. Uncertainty is quantified via Monte Carlo simulation, with results reported alongside 95% uncertainty intervals (UIs) that reflect estimate precision.

This study systematically examined the burden of PPH among reproductive‐aged women (15–49 years) from 1990 to 2021, analyzing data on case counts, incidence, mortality, and DALYs. All data were obtained from the GBD database (https://vizhub.healthdata.org/gbd-results), which includes estimates for 204 countries and territories. It downloaded on December 16, 2024, contained variables for year, age group (seven 5‐year categories from 15 to 19 to 45–49 years), and geographic region. Since the GBD database does not disaggregate estimates by race or ethnicity, related analyses were not performed. Reporting adhered to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines (https://www.goodreports.org/completed/97493) [14].

The PPH estimation methodology has been refined across GBD iterations to enhance data accuracy and comparability. In GBD 2017, the PPH case definition was standardized according to international obstetric guidelines. GBD 2019 introduced improved modeling of severe anemia as a sequela of PPH, incorporating updated disability weights. Several key methodological updates in GBD 2021 are relevant to this analysis: enhanced Cause of Death Ensemble modeling incorporating healthcare access and quality covariates; refined natural history models for PPH incidence to better account for underreporting; updated population weights for age‐standardization; expanded risk factor attribution for iron deficiency using improved exposure estimates; and enhanced spatial‐temporal modeling to address data sparsity in low‐resource settings. These refinements generally yield more conservative uncertainty estimates and better capture between‐country heterogeneity, while the overall temporal trends reported in this study remain consistent with previous GBD rounds, supporting the robustness of our findings. The GBD Collaborator Network provides a detailed appendix on GBD 2021 estimation approaches for maternal disorders, including PPH (http://ghdx.healthdata.org/gbd-2021).

2.2. Core Indicators and Calculation Methods

GBD study employs three primary indicators to assess disease burden, utilizing specific methodologies for their calculation. Mortality is quantified by counting the number of deaths directly or indirectly attributable to a particular disease, with adjustments made for underreporting and misdiagnosis through the application of Cause of Death Ensemble modeling [15]. Incidence is determined by recording the number of new cases and estimating undiagnosed cases using a disease natural history model. DALYs serve as a comprehensive metric that encompasses the effects of both premature mortality and disability. The calculation for DALYs is expressed as: DALYs = Years of Life Lost + Years Lived with Disability. Years of Life Lost is computed by determining the gap between the age at death and the standard life expectancy, based on the GBD global standard population model [16]. Years Lived with Disability is calculated by quantifying the loss of health status, which involves multiplying the prevalence rate by the disability weight. The Disability weight is established through extensive population surveys and the Delphi method involving experts [17].

2.3. Disability Weight Standardisation

Disability weight standardization values including the European EQ‐5D and the US MEPS. These values are determined through the pairwise comparison method and the population equivalence method, which assign continuous values ranging from 0 (indicating perfect health) to 1 (indicating death) [18]. The validity of this method has been confirmed across different cultural contexts [19].

2.4. Socio‐Demographic Index

The socio‐demographic index (SDI) serves as a comprehensive metric for assessing the socio‐economic development status of a nation or region, encompassing various dimensions such as economic size and structure, educational attainment, living standards, and social welfare. The SDI is quantified on a scale ranging from 0 to 1, with higher values signifying elevated levels of socio‐economic development. Utilizing data from the GBD database, countries and regions are categorized into five distinct groups based on their SDI: low, low‐middle, middle, high‐middle, and high.

2.5. BAPC Model

The Bayesian age‐period‐cohort (BAPC) model is extensively employed within the fields of epidemiology and public health to examine temporal trends in disease burden and to forecast future variations. This model disaggregates time into three distinct dimensions: the age effect, the period effect, and the cohort effect, thereby facilitating a comprehensive understanding of shifts in disease risk.

2.6. Statistical Analysis

The disease burden was measured using age‐standardized incidence, mortality, and DALY rates per 100,000 population, presented with 95% UIs. Trends were analyzed by Joinpoint regression to identify significant inflection points, with annual percentage changes (APCs) and 95% confidence intervals (CIs) calculated for each segment. Long‐term trends were summarized by the estimated average annual percentage change (EAPC), derived from a log‐linear model [$\ln (\text{rate})=\text{α}+\text{β}\times \text{calendar year}$], where EAPC = $100\times (\text{exp}(\text{β})-1)$. A trend was considered increasing if both the EAPC and its lower 95% CI were > 0, and decreasing if both the EAPC and its upper 95% CI were < 0. Additionally, nonlinear smoothing curves were used to examine the association between each burden metric and the SDI, visualizing deviations from expected transition patterns across development levels.

We assessed socioeconomic inequalities in disease burden using two complementary measures. The Concentration Index (CI) quantifies relative inequality, while the Slope Index of Inequality (SII) captures absolute disparity across the socioeconomic gradient. The CI was calculated as $\mathrm{CI}=\frac{2}{\text{μ}}\text{Cov}(\mathrm{y},\mathrm{r})$, where y represents the health outcome, μis its mean, and r denotes the fractional socioeconomic rank (ranging from 0 for the most disadvantaged to 1 for the most advantaged). It reflects the disproportionate concentration of the health outcome, with values ranging from −1 to 1. In contrast, the SII was derived from a log‐binomial regression model, $\mathrm{g}(\mathrm{E}[\mathrm{y}])=\alpha +\text{β}r+X\text{γ}$, where r is the relative rank and X includes adjustment covariates. It represents the model‐predicted difference in the outcome between the extreme ends of the socioeconomic spectrum (r = 1 vs. r = 0). As the CI and SII operate on relative and absolute scales, respectively, their trends may diverge—for example, absolute disparities may decrease while relative inequalities persist. Interpreting both indices together therefore provides a more comprehensive understanding of inequality dynamics over time. In this study, a log‐binomial regression model was used, allowing direct estimation of risk differences on the absolute scale. SII estimation was conducted using the glm function in base R with a log link, supplemented by the margins package to derive predicted values and absolute contrasts.

Future disease burden was projected through 2035. We used a Bayesian age‐period‐cohort model within a hierarchical framework. This approach simultaneously accounts for age, period, and cohort effects. The model incorporated historical burden data, demographic projections, and random effects to produce stable estimates. Uncertainty was quantified as 95% prediction intervals derived from posterior predictive distributions.

All statistical analyses were performed using R software (version 4.4.2). The following R packages and versions were used:segmented (version 2.1–0) for joinpoint regression and APC calculation, BAPC (version 0.0.33) and INLA (version 24.02.09) for Bayesian age‐period‐cohort modeling and future burden projections; MASS (version 7.3–60) for robust linear regression in SII estimation; splines (base R) and mgcv (version 1.9–1) for spline smoothing; car (version 3.1–2) for heteroscedasticity testing; and base R stats package for general statistical modeling. A two‐sided p value < 0.05 was considered statistically significant.

3. Results

3.1. Global Burden Trends

3.1.1. Incidence

The global incidence of PPH among women of childbearing age declined from 1990 to 2021. In 1990, the incidence rate was 998.26 per 100,000 individuals (95% UI: 654.55–1,438.33). By 2021, this rate had decreased to 722.16 per 100,000 (95% UI: 482.99–1,022.71). The corresponding estimated annual percentage change was –0.83% (95% CI: –0.90% to –0.77%) (Figure 1A). This decline was not uniform over time. The steepest reduction occurred between 1990 and 1994, with an APC of –2.32% (95% CI: –2.45% to –2.19%). The slowest decrease was observed from 2008 to 2015, with an APC of –0.34% (95% CI: –0.40% to –0.27%) (Figure 1D). Concurrently, the age distribution of incident cases shifted over time. In 1990, women aged 20–24 years accounted for the highest number of cases, with 4.37 million cases representing 30.9% of the total. By 2021, the peak had shifted to the 25–29‐year age group, with 3.22 million cases comprising 28.2% of all cases (Figure 2A).

Figure 1.

Annual percent change (APC) and trends in global postpartum hemorrhage from 1990 to 2021. The EAPCs in ASIR (A), ASMR (B), and ASDR (C) due to postpartum hemorrhage from 1990 to 2019 by GBD region and by SDI. (D) Incidence rate. (E) Mortality rate. (F) DALYs rate.

Figure 2.

Age‐specific percentages of postpartum hemorrhage incidence cases, deaths, and disability‐adjusted life years in 1990 and 2021. (A) Incidence cases. (B) Deaths. (C) Disability‐adjusted life years.

3.1.2. Mortality

The mortality rate associated with PPH also declined over the past three decades. In 1990, the rate was 8.23 per 100,000 individuals (95% UI: 6.92–9.56). By 2021, it had fallen to 2.40 per 100,000 (95% UI: 1.97–2.92). This decline corresponded to an estimated annual percentage change of –4.05% (95% CI: –4.25% to –3.86%) (Figure 1B). The rate of decline varied across periods, with the most rapid reduction occurring between 2013 and 2014 (APC = –6.58%; 95% CI: –7.69% to –5.45%), while the slowest was observed from 1990 to 2000 (APC = –2.35%; 95% CI: –2.43% to –2.28%) (Figure 1E). Throughout this period, the highest burden of fatal PPH cases occurred among women aged 20–24 years. In 1990, this age group accounted for 25.2% of all reproductive‐age deaths, representing 28,407 deaths (95% UI: 23,514–33,153). By 2021, the proportion remained high at 22.8%, with 10,582 deaths (95% UI: 8,864–12,693) (Figure 2B).

3.1.3. DALY

The DALY rate associated with PPH among women of childbearing age declined significantly from 1990 to 2021. In 1990, the rate was 506.23 per 100,000 population (95% UI: 426.65–585.78). By 2021, it had decreased to 151.93 per 100,000 (95% UI: 126.10–183.93), corresponding to an estimated annual percentage change of –3.97% (95% CI: –4.17% to –3.78%) (Figure 1C). The pace of decline varied over time. The most pronounced reduction occurred between 2010 and 2019, with an APC of –4.89% (95% CI: –4.98% to –4.79%). The slowest decline was observed from 1990 to 2000, with an APC of –2.22% (95% CI: –2.30% to –2.13%) (Figure 1F). Throughout the study period, women aged 20–24 years consistently bore the highest DALY burden. In 1990, this age group accounted for 27.9% of all DALYs among reproductive‐aged women, representing 1.97 million years (95% UI: 1.64–2.28 million). In 2021, they still accounted for 25.6% of DALYs, representing 749,063 years (95% UI: 629,108–894,036) (Figure 2C).

3.2. Regional Burden Trends

3.2.1. Incidence

Between 1990 and 2021, the age‑standardized incidence rate (ASIR) of PPH showed an overall decline across all regions (Figure 3A). In 1990, the highest ASIR was observed in the low SDI region (2,357.15 per 100,000; 95% UI: 1,532.05–3,408.19), with Central Sub‑Saharan Africa reporting the highest regional rate (2,983.68 per 100,000). By 2021, the low SDI region remained the highest‑burden category, though the ASIR had decreased to 1,487.20 per 100,000 (95% UI: 970.21–2,152.57). The steepest decline in PPH incidence occurred in South Asia, with an EAPC of –2.70% (95% CI: –2.86% to –2.54%). In contrast, tropical Latin America experienced an upward trend, with an EAPC of 1.37% (95% CI: 0.72%–2.02%). Eastern Europe showed a marginal increase, with an EAPC of 0.20% (95% CI: –0.24%–0.65%) (Table 1). At the SDI level, the ASIR was inversely correlated with socioeconomic development (r = –0.7227, p < 0.01) (Figure 4A). The most pronounced reduction was observed in low‐SDI areas, with an EAPC of –2.07% (95% CI: –2.15% to –1.99%) (Figure 1A).

Figure 3.

Temporal trend of ASIR (A), ASMR (B), and ASDR (C) of postpartum hemorrhage from 1990 to 2021 across the globe and 5 SDI regions.

Table 1.

Incidence of postpartum hemorrhage in fertile female between 1990 and 2021 at the global and regional level.

Location	Rate per 100,000(95%UI)
	1990		2021		1990‐2021
	Incident cases	Age‐Std incidence rate	Incident cases	Age‐Std Incidence Rate	EAPC
Global	14123712 (10771873.17, 18750120.07)	998.26 (654.55, 1438.33)	13914839 (10854607.28, 17706034.51)	722.16 (482.99, 1022.71)	−0.83 (−0.90, −0.77)
SDI
Low SDI	2750281 (2134367.68, 3600586.50)	2357.15 (1532.05, 3408.19)	4222939 (3244664.98, 5506998.29)	1487.20 (970.21, 2152.57)	−1.52 (−1.65, −1.40)
Low‐middle SDI	3910057 (3009472.32, 5114853.82)	1364.16 (891.25, 1986.16)	3713976 (2832851.38, 4870811.18)	712.50 (467.54, 1031.73)	−2.07 (−2.15, −1.99)
Middle SDI	4271749 (3191209.13, 5719101.58)	879.47 (570.86, 1282.94)	3504993 (2721208.83, 4507789.33)	580.54 (393.84, 812.17)	−0.85 (−1.06, −0.64)
High‐middle SDI	2084106 (1514564.87, 2811273.56)	705.51 (458.71, 1031.45)	1518217 (1175797.71, 1925673.23)	528.50 (364.17, 720.57)	−0.31 (−0.65, 0.04)
High SDI	1094500 (838674.40, 1417235.42)	483.20 (326.29, 683.63)	943670 (747247.04, 1182512.23)	389.12 (277.32, 520.20)	−0.76 (−0.83, −0.69)
Region
Andean Latin America	195467 (148711.08, 256002.26)	1944.80 (1266.03, 2800.16)	215233 (167175.77, 271099.96)	1215.92 (841.50, 1665.90)	−1.35 (−1.42, −1.28)
Australasia	38971 (31116.05, 48035.06)	722.56 (523.60, 947.57)	47213 (35927.32, 61813.91)	638.55 (425.78, 905.68)	−0.22 (−0.51, 0.07)
Caribbean	84213 (64152.55, 111972.97)	837.05 (539.77, 1229.92)	70242 (54134.45, 90460.24)	589.04 (379.70, 859.71)	−1.11 (−1.16, −1.06)
Central Asia	141044 (105355.35, 193936.36)	748.09 (481.25, 1122.28)	129780 (98308.08, 173583.71)	536.37 (346.25, 805.81)	−0.68 (−0.87, −0.50)
Central Europe	308114 (219613.72, 410709.79)	1074.50 (691.30, 1552.85)	168566 (133876.72, 211910.79)	734.48 (520.57, 985.13)	−0.87 (−1.21, −0.54)
Central Latin America	661861 (502700.79, 879690.05)	1467.65 (955.20, 2115.54)	584123 (457239.15, 747493.20)	862.69 (578.83, 1214.75)	−1.43 (−1.66, −1.19)
Central Sub‐Saharan Africa	380812 (292211.13, 499114.88)	2983.68 (1907.30, 4342.53)	591209 (451037.81, 768245.08)	1785.58 (1130.00, 2590.41)	−1.51 (−1.72, −1.30)
East Asia	2365062 (1669851.71, 3302814.33)	629.740 (401.20, 941.34)	1451881 (1064249.10, 1873463.95)	473.92 (319.68, 636.58)	−0.05 (−0.70, 0.60)
Eastern Europe	417511 (297833.06, 569065.34)	806.85 (520.77, 1199.35)	271072 (203005.67, 363229.49)	636.60 (406.89, 925.92)	0.20 (−0.24, 0.65)
Eastern Sub‐Saharan Africa	1169910 (897182.36, 1547708.45)	2525.49 (1631.19, 3675.56)	1740319 (1316294.09, 2318518.45)	1540.10 (995.62, 2249.17)	−1.60 (−1.70, −1.50)
High‐income Asia Pacific	76368 (56207.05, 102562.66)	178.60 (118.44, 256.00)	57560 (41703.13, 77478.00)	157.61 (103.54, 227.73)	−0.26 (−0.40, −0.13)
High‐income North America	349999 (265890.35, 457052.21)	469.21 (307.12, 675.10)	341127 (280279.65, 412178.33)	405.70 (305.42, 518.59)	−0.89 (−1.24, −0.54)
North Africa and Middle East	1058867 (795306.47, 1428355.78)	1265.27 (819.91, 1850.71)	968955 (738563.93, 1290111.43)	608.52 (399.34, 886.67)	−1.99 (−2.13, −1.84)
Oceania	21570 (16060.75, 28661.19)	1320.40 (848.40, 1935.38)	35832 (27031.63, 47259.83)	998.67 (642.73, 1470.82)	−0.90 (−0.96, −0.84)
South Asia	3385357 (2520743.56, 4443869.00)	1263.28 (812.41, 1845.30)	2882972 (2131048.70, 3862440.64)	565.08 (367.63, 822.35)	−2.70 (−2.86, −2.54)
Southeast Asia	1367199 (1044914.45, 1775460.76)	1079.93 (701.04, 1549.55)	1275671 (986922.38, 1623501.96)	701.06 (474.38, 976.92)	−1.12 (−1.28, −0.97)
Southern Latin America	143807 (104938.49, 191479.47)	1143.96 (740.74, 1655.26)	131040 (103445.91, 169891.10)	759.81 (521.67, 1061.84)	−1.22 (−1.29, −1.15)
Southern Sub‐Saharan Africa	103274 (81416.03, 134064.59)	736.96 (478.65, 1088.83)	118394 (91395.57, 154260.07)	532.55 (343.51, 780.88)	−0.52 (−0.73, −0.31)
Tropical Latin America	124920 (102115.43, 150004.74)	291.45 (205.57, 393.28)	216782 (170390.11, 276624.51)	374.66 (257.64, 536.76)	1.37 (0.72, 2.02)
Western Europe	505008 (383042.22, 657311.41)	524.67 (350.95, 739.86)	417504 (326099.62, 530190.32)	460.81 (317.42, 634.88)	−0.46 (−0.53, −0.39)
Western Sub‐Saharan Africa	1224379 (949535.47, 1602642.39)	2725.01 (1752.45, 3924.58)	2199362 (1690799.66, 2838692.32)	1803.57 (1159.01, 2612.98)	−1.23 (−1.34, −1.13)

Note: ^aEAPC is expressed as 95% confidence interval.

Abbreviations: EAPC, estimated annual percentage change; SDI, Sociodemographic Index; UI, uncertainty interval.

Figure 4.

Association between incidence, mortality, and disability‐adjusted life years rates of postpartum hemorrhage and regionalSociodemographic Index (SDI), 1990–2021. (A) Incidence rate. (B) Mortality rate. (C) DALYs rates.

3.2.2. Mortality

Between 1990 and 2021, the age‑standardized mortality rate (ASMR) for PPH exhibited a sustained decline across all regions (Figure 3B). In 1990, the highest ASMR was recorded in low SDI regions (30.90 per 100,000; 95% UI: 25.20–36.84), with Eastern Sub‑Saharan Africa representing the highest‑burden geographic area (30.63 per 100,000). By 2021, although the overall burden had decreased, low SDI regions still reported the highest ASMR (9.94 per 100,000; 95% UI: 7.85–12.44), with Western Sub‑Saharan Africa showing the highest regional rate (13.61 per 100,000). At the SDI level, ASMR was strongly and inversely correlated with socioeconomic development (r = –0.9407, p < 0.01) (Figure 4B). The greatest reduction occurred in medium‑high SDI regions (EAPC = –7.37%; 95% CI: –7.62% to –7.12%), whereas low SDI regions experienced the smallest decline (EAPC = –3.55%; 95% CI: –3.81% to –3.30%) (Figure 1B). Geographically, the most pronounced decrease was observed in East Asia (EAPC = –8.98%; 95% CI: –9.35% to –8.60%) (Table 2).

Table 2.

Deaths of postpartum hemorrhage in fertile female between 1990 and 2021 at the global and regional level.

Location	Rate per 100,000(95%UI)
	1990		2021		1990–2021
	Deaths	Age‐Std mortality rate	Deaths	Age‐Std mortality rate	aEAPC
Global	112778 (100262.52, 125894.77)	8.23 (6.92, 9.56)	46345 (38790.83, 55740.44)	2.40 (1.97, 2.92)	−4.05 (−4.25, −3.86)
SDI
Low SDI	33392 (28239.93, 38264.78)	30.90 (25.20, 36.84)	26538 (21402.37, 32443.45)	9.94 (7.85, 12.44)	−3.55 (−3.81, −3.30)
Low‐middle SDI	51324 (45047.94, 57616.73)	18.22 (14.94, 21.80)	14383 (12118.86, 17236.27)	2.80 (2.25, 3.47)	−6.13 (−6.46, −5.81)
Middle SDI	24480 (21145.37, 28397.96)	5.35 (4.40, 6.39)	4992 (4217.1, 6072.61)	0.82 (0.65, 1.04)	−6.14 (−6.34, −5.93)
High‐middle SDI	3053 (2449.97, 3649.84)	1.07 (0.84, 1.32)	306 (254.67, 376.96)	0.10 (0.08, 0.13)	−7.37 (−7.62, −7.12)
High SDI	466 (351.66, 605.27)	0.20 (0.15, 0.27)	83 (67.74, 103.66)	0.03 (0.03, 0.05)	−5.03 (−5.30, −4.75)
Region
Andean Latin America	1013 (852.20, 1206.32)	11.37 (8.70, 14.49)	362 (259.76, 493.71)	2.05 (1.33, 3.02)	−6.02 (−6.33, −5.72)
Australasia	3 (2.12, 3.72)	0.05 (0.03, 0.08)	1 (0.5, 0.9)	0.01 (0.01, 0.02)	−4.96 (−5.59, −4.32)
Caribbean	425 (318.94, 553.64)	4.69 (3.12, 6.79)	451 (285.98, 653.74)	3.74 (2.12, 6.05)	−0.07 (−0.46, 0.33)
Central Asia	326 (287.82, 363.58)	1.87 (1.57, 2.19)	77 (61.5, 94.46)	0.31 (0.23, 0.40)	−5.77 (−6.00, −5.54)
Central Europe	109 (93.17, 128.12)	0.36 (0.28, 0.45)	9 (7.22, 10.29)	0.04 (0.03, 0.04)	−6.85 (−7.15, −6.55)
Central Latin America	1447 (1302.35, 1595.26)	3.50 (2.94, 4.10)	415 (338.12, 510.83)	0.61 (0.46, 0.79)	−5.55 (−5.80, −5.31)
Central Sub‐Saharan Africa	3060 (2154.51, 4058.9)	27.03 (17.22, 39.46)	3290 (2285.78, 4589.75)	11.03 (6.71, 16.93)	−2.15 (−2.71, −1.59)
East Asia	5346 (3624.02, 7349.54)	1.57 (1.06, 2.17)	298 (212.21, 410.47)	0.09 (0.07, 0.13)	−8.98 (−9.35, −8.60)
Eastern Europe	222 (185.86, 269.88)	0.39 (0.29, 0.53)	20 (15.73, 24.93)	0.04 (0.03, 0.06)	−6.71 (−7.16, −6.25)
Eastern Sub‐Saharan Africa	11515 (9682.27, 13432.97)	30.63 (23.97, 37.86)	6632 (5300.32, 8138.94)	6.74 (5.14, 8.64)	−4.90 (−5.20, −4.60)
High‐income Asia Pacific	63 (53.27, 76.1)	0.14 (0.11, 0.19)	9 (6.97, 11.33)	0.03 (0.02, 0.03)	−5.53 (−5.68, −5.38)
High‐income North America	61 (47.95, 78.24)	0.08 (0.05, 0.12)	30 (22.76, 38.16)	0.04 (0.02, 0.05)	−1.95 (−2.28, −1.62)
North Africa and Middle East	6605 (5342.19, 7846.99)	8.74 (6.83, 10.80)	2794 (2005.3, 3843.08)	1.75 (1.19, 2.47)	−4.83 (−4.94, −4.71)
Oceania	215 (107.84, 328.81)	14.08 (6.60, 23.50)	283 (186.83, 398.94)	8.01 (4.65, 12.48)	−1.62 (−1.77, −1.48)
South Asia	54686 (46674.75, 63339.09)	20.22 (15.51, 25.42)	11533 (9365.13, 14453)	2.26 (1.67, 3.03)	−7.41 (−7.87, −6.94)
Southeast Asia	14269 (12085.57, 16600.8)	11.98 (9.48, 14.68)	3583 (2881.7, 4673.08)	1.95 (1.46, 2.59)	−5.97 (−6.13, −5.80)
Southern Latin America	96 (77.84, 118.76)	0.79 (0.53, 1.12)	26 (19.63, 34.7)	0.15 (0.10, 0.23)	−4.38 (−4.73, −4.03)
Southern Sub‐Saharan Africa	761 (584.37, 993.17)	6.02 (4.24, 8.39)	355 (250.92, 497.48)	1.62 (1.03, 2.43)	−2.70 (−3.79, −1.60)
Tropical Latin America	986 (836.48, 1145.03)	2.54 (1.80, 3.40)	221 (179.73, 264.58)	0.37 (0.25, 0.52)	−5.32 (−5.77, −4.87)
Western Europe	67 (59.47, 75.52)	0.07 (0.06, 0.08)	15 (13.34, 17.24)	0.02 (0.01, 0.02)	−3.87 (−4.21, −3.52)
Western Sub‐Saharan Africa	11502 (8870.1, 14400.35)	27.81 (20.67, 36.38)	15942 (12022.93, 21318.69)	13.61 (9.93, 18.51)	−1.80 (−2.05, −1.56)

Note: ^aEAPC is expressed as 95% confidence interval.

Abbreviations: EAPC, estimated annual percentage change; SDI, Sociodemographic Index; UI, uncertainty interval.

3.2.3. DALY

The age‑standardized DALY rate (ASDR) for PPH declined substantially across all regions from 1990 to 2021 (Figure 3C). In 1990, the highest burden was observed in low SDI regions (ASDR = 1,820.26 per 100,000; 95% UI: 1,490.52–2,162.73), with Eastern Sub‑Saharan Africa recording the highest regional rate (1,706.66 per 100,000). By 2021, low SDI regions continued to bear the highest burden, though the ASDR had decreased to 595.53 per 100,000 (95% UI: 473.96–740.50); Western Sub‑Saharan Africa showed the highest regional rate in that year (819.63 per 100,000). A strong inverse correlation was found between SDI and ASDR (r = –0.9437, p < 0.01) (Figure 4C). The greatest reduction occurred in medium‑high SDI regions (EAPC = –6.21%; 95% CI: –6.36% to –6.05%), while low SDI regions experienced the smallest decline (EAPC = –3.51%; 95% CI: –3.77% to –3.26%) (Figure 1C). Geographically, East Asia showed the steepest decrease (EAPC = –7.89%; 95% CI: –8.20% to –7.59%) (Table 3).

Table 3.

DALY of postpartum hemorrhage in fertile female between 1990 and 2021 at the global and regional level.

Location	Rate per 100,000 (95%UI)
	1990		2021		1990–2021
	DALY	Age‐Std DALY rate	DALY	Age‐Std DALY rate	aEAPC
Global	7036063 (6268324.86, 7830171.27)	506.23 (426.65, 585.78)	2924613 (2467490.52, 3488363.43)	151.93 (126.1, 183.93)	−3.97 (−4.17, 3.78)
SDI
Low SDI	2029254 (1720436.51, 2320090.87)	1820.26 (1490.52, 2162.73)	1637423 (1333518.33, 1977395.25)	595.53 (473.96, 740.50)	−3.51 (−3.77, −3.26)
Low‐middle SDI	3231389 (2816234.27, 3624259.7)	1124.09 (922.70, 1342.64)	917800 (778974.46, 1094640.83)	176.91 (143.79, 218.40)	−6.07 (−6.40, −5.75)
Middle SDI	1538383 (1338169.23, 1783327.2)	328.77 (271.00, 391.76)	328608 (282255.06, 390397.23)	54.67 (44.23, 67.82)	−5.89 (−6.06, −5.70)
High‐middle SDI	199122 (163290.38, 237343.65)	69.06 (55.00, 84.22)	28248 (22946.64, 34786.81)	9.85 (7.57, 12.65)	−6.21 (−6.36, −6.05)
High SDI	34077 (26054.06, 43017.26)	14.90 (11.11, 19.45)	9967 (7685.78, 12721.5)	4.10 (3.00, 5.50)	−3.70 (−3.91, −3.49)
Region
Andean Latin America	60703 (51073.88, 72225.68)	662.05 (512.65, 838.72)	22149 (16429.49, 29996.74)	125.33 (83.82, 180.47)	−5.84 (−6.13, −5.55)
Australasia	363 (265.86, 486.18)	6.73 (4.61, 9.48)	266 (160.48, 398.87)	3.61 (1.95, 5.96)	−1.79 (−2.09, −1.49)
Caribbean	25836 (19506.13, 33112.01)	278.85 (189.06, 399.70)	27156 (17373.77, 39352.54)	225.69 (130.33, 360.88)	−0.02 (−0.42, 0.38)
Central Asia	21047 (18663.03, 23539.89)	118.04 (99.62, 138.14)	5898 (4683.24, 7108.43)	23.82 (17.87, 30.97)	−5.10 (−5.33, −4.86)
Central Europe	8739 (7335.98, 10355.48)	29.11 (22.63, 37.08)	1662 (1172.68, 2224.21)	7.08 (4.72, 10.11)	−4.14 (−4.57, −3.71)
Central Latin America	91156 (81200.19, 100496.3)	214.30 (180.96, 250.70)	28016 (23038.82, 33751.37)	41.15 (32.06, 52.33)	−5.22 (−5.45, −5.00)
Central Sub‐Saharan Africa	183315 (129631.56, 241516.64)	1558.00 (1010.24, 2246.34)	194750 (136865.17, 265068.02)	631.75 (392.23, 951.71)	−2.20 (−2.75, −1.65)
East Asia	341778 (238442.45, 463870.30)	98.04 (67.63, 134.31)	26222 (20064.96, 34001.8)	8.37 (6.14, 11.28)	−7.89 (−8.20, −7.59)
Eastern Europe	16260 (13728.13, 19413.83)	29.44 (22.14, 38.45)	2943 (2156.90, 3957.05)	6.66 (4.40, 9.77)	−4.37 (−4.60, −4.14)
Eastern Sub−Saharan Africa	670694 (566566.88, 776977.84)	1706.66 (1348.22, 2097.36)	401339 (325053.30, 488700.9)	391.90 (302.50, 498.35)	−4.76 (−5.05, −4.47)
High‐income Asia Pacific	4173 (3514.27, 4978.74)	9.55 (7.17, 12.40)	842 (655.44, 1086.56)	2.31 (1.650, 3.20)	−4.45 (−4.62, −4.28)
High‐income North America	5569 (4376.43, 6937.3)	7.45 (5.18, 10.38)	3506 (2730.58, 4596.04)	4.17 (2.93, 5.82)	−1.58 (−1.75, −1.40)
North Africa and Middle East	405630 (329690.46, 481795.08)	520.28 (410.49, 637.77)	176552 (128733.59, 237795.55)	110.82 (77.35, 154.10)	−4.64 (−4.74, −4.54)
Oceania	13184 (6648.40, 19963.18)	839.06 (406.29, 1386.55)	17617 (11704.19, 24538.21)	493.60 (291.44, 759.98)	−1.52 (−1.67, −1.38)
South Asia	3501909 (2979225.78, 4060091.95)	1276.48 (98.437, 1598.99)	765320 (625361.22, 947090.51)	149.53 (112.28, 197.50)	−7.28 (−7.73, −6.82)
Southeast Asia	865539 (731422.06, 1004049.82)	709.21 (562.63, 866.61)	218711 (176620.22, 282359.15)	119.46 (90.84, 156.78)	−5.84 (−6.00, −5.68)
Southern Latin America	6327 (5128.10, 7736.98)	51.24 (35.55, 71.26)	2265 (1741.75, 3004.65)	13.02 (8.56, 19.58)	−3.69 (−3.95, −3.42)
Southern Sub‐Saharan Africa	45900 (35337.98, 59192.94)	352.27 (248.86, 489.73)	21678 (15606.66, 29923.31)	98.31 (63.92, 145.39)	−2.61 (−3.69, −1.52)
Tropical Latin America	58534 (49736.29, 68012.27)	147.72 (105.12, 197.26)	15085 (12564.98, 17919.35)	25.35 (17.96, 34.88)	−4.73 (−5.14, −4.33)
Western Europe	6470 (5125.65, 7873.11)	6.70 (5.11, 8.72)	2867 (1973.81, 3930.82)	3.15 (2.09, 4.62)	−2.11 (−2.27, −1.94)
Western Sub‐Saharan Africa	702939 (548196.81, 883885.62)	1640.80 (1223.37, 2137.93)	989768 (752170.46, 1316476.89)	819.63 (600.06, 1103.20)	−1.74 (−1.99, −1.50)

Note: ^aEAPC is expressed as 95% confidence interval.

Abbreviations: EAPC, estimated annual percentage change; SDI, Sociodemographic Index; UI, uncertainty interval.

3.3. National Burden Trends

3.3.1. Incidence

The incidence of PPH declined from 998.26 per 100,000 individuals (95% UI: 654.55–1,438.33) in 1990 to 722.16 per 100,000 (95% UI: 482.99–1,022.71) in 2021. During this period, the location of the highest national incidence shifted. In 1990, Niger recorded the highest rate worldwide, with 3,593.17 per 100,000 (95% UI: 2,301.80–5,251.31). By 2021, Nigeria reported the highest rate, with 3,081.84 per 100,000 (95% UI: 1,913.61–4,531.77). Notably, Brazil showed the greatest rise in incidence, with an EAPC of 1.48% (95% CI: 0.81%–2.15%). In 1990, the global incidence exceeded that of 83 countries and was lower than 121 countries. By 2021, it surpassed the rates in 120 countries but remained below those in 84 countries (Figure 5A, Table S1).

Figure 5.

Incidence, mortality, and disability‐adjusted life years rates of postpartum hemorrhage across 204 countries and territories. (A) Incidence rate. (B) Mortality rate. (C) DALYs rates.

3.3.2. Mortality

The global mortality rate of PPH declined from 8.23 (95% UI: 6.92–9.56) to 2.40 (95% UI: 1.97–2.92) per 100,000 individuals between 1990 and 2021. In 1990, this global rate exceeded the rates in 119 countries and was lower than those in 65 countries. By 2021, it was higher than in 151 countries and lower than in 53 (Figure 5B, Table S2). Sierra Leone reported the highest national mortality rate in both 1990 (78.13 per 100,000; 95% UI: 49.01–113.74) and 2021 (36.66 per 100,000; 95% UI: 20.84–56.91). Zimbabwe experienced the greatest relative increase in mortality over the period, with an EAPC of 0.40% (95% CI: –0.70% to 1.51%).

3.3.3. DALY

The global DALY rate for PPH declined from 506.23 per 100,000 individuals (95% UI: 426.65–585.78) in 1990 to 151.93 per 100,000 (95% UI: 126.10–183.93) in 2021. In 1990, the global rate was higher than that in 140 of 204 countries and lower than in 64. By 2021, it exceeded the rates in 153 countries and was below those in 51 (Figure 5C, Table S3). Throughout this period, Sierra Leone recorded the highest national DALY rate, with 4,551.70 (95% UI: 2,880.12–6,588.20) per 100,000 in 1990 and 2,160.78 (95% UI: 1,237.91–3,337.46) per 100,000 in 2021. Zimbabwe showed the greatest increase in DALYs over the three decades, with an EAPC of 0.45% (95% CI: –0.63% to 1.54%).

3.4. Risk Factors

Iron deficiency is identified in the GBD 2021 study as the specific risk factor for maternal disorders. Between 1990 and 2021, the proportion of global DALYs from PPH attributable to iron deficiency among women of reproductive age showed an upward trend, increasing from 17.19% (95% UI: 8.48–22.06%) in 1990% to 17.96% (95% UI: 8.45–23.93%) in 2021. This trend varied by socioeconomic development level. In low‐ and middle‐SDI regions, the attributable proportion rose from 17.06% (95% UI: 8.65–21.48%) to 18.15% (95% UI: 8.67–23.95%). Within low‐SDI regions specifically, it increased from 16.97% (95% UI: 8.55–21.67%) to 17.87% (95% UI: 8.46–23.75%). Conversely, in high‐SDI regions, the proportion declined from 18.59% (95% UI: 8.20–26.37%) to 16.72% (95% UI: 7.00–24.49%) (Figure 6A).

Figure 6.

The temporal trend of the proportion of age‐standardized DALY rate attributable to postpartum hemorrhage due to iron deficiency across the globe and 5 SDI regions from 1990 to 2021. (A) Health inequality regression curves. (B) Concentration curves. (C) The age‐standardized DALY rate of postpartum hemorrhage from 1990 to 2021 across the world.

3.5. Cross‐Country Inequality Analysis

The cross‐country inequality analysis revealed that the PPH burden was disproportionately concentrated in lower socioeconomic groups. We employed two complementary measures to assess this disparity. The CI measures relative inequality. In 1990, the CI was −0.63 (95% CI: −0.68 to −0.58), indicating significant pro‐poor inequality. By 2021, the CI decreased to −0.70 (95% CI: −0.75 to −0.65). This more negative value indicates that relative inequality has worsened, meaning the remaining burden has become even more concentrated among disadvantaged populations relative to the declining global average (Figure 6B). The SII measures absolute inequality. In 1990, the SII was −1,669.61 (95% CI: −1,779.06 to −1,560.15), reflecting a large absolute gap in DALY rates across socioeconomic groups. By 2021, the SII decreased substantially to −357.90 (95% CI: −396.44 to −319.37), indicating a marked reduction in absolute disparity. However, the persistently negative value confirms that the absolute burden remains significantly higher in low‐SDI populations, underscoring the continued need for targeted interventions (Figure 6C).

3.6. Projections 2022–2035

Using a BAPC model, we projected the ASIR, ASMR, and ASDR of PPH among women aged 15–49 years from 2022 to 2035. The projections indicate a continued decline in all three burden metrics over this period. By 2035, the global ASIR is expected to decrease to 466.60 per 100,000 population (95% UI: 417.66–515.55). Similarly, the ASMR is projected to fall to 1.23 per 100,000 (95% UI: 1.03–1.43), and the ASDR to 76.60 per 100,000 (95% UI: 64.09–89.10) (Figure 7).

Figure 7.

The temporal trends of ASIR (A), ASMR (B), and ASDR (C) of postpartum hemorrhage between 1990 and 2021 and their projections through 2035 across the globe.

4. Discussion

This study demonstrates a significant global decline in the burden of PPH among women of reproductive age from 1990 to 2021. The age‐standardized incidence, mortality, and DALY rates decreased substantially during this period. The estimated annual percentage changes were –0.83%, –4.05%, and –3.97%, respectively. This trend aligns with overall WHO reports on maternal health, reflecting worldwide improvements in care. However, disparities persist across regions. The burden remains alarmingly high in low‐SDI settings, particularly in sub‐Saharan Africa. In 2021, the age‐standardized mortality rate in Western Sub‐Saharan Africa was 13.61 per 100,000 population. This rate was approximately 400 times higher than that in high‐SDI regions, where the rate was 0.034 per 100,000. This inequality is consistent with reports that 87% of maternal deaths occur in low‐income countries [20, 21]. It also underscores how the unequal distribution of medical resources contributes to disparities in PPH outcomes [22]. These findings call for urgent policy attention. Notably, East Asia achieved the steepest decline in mortality, with an EAPC of –8.98%. This improvement was likely driven by the scale‐up of community‐based distribution of uterotonics, including oxytocin and misoprostol [23, 24]. These findings demonstrate that targeted interventions can substantially improve PPH outcomes. In contrast, tropical Latin America experienced a rising incidence (EAPC = 1.37%), which may be linked to increasing cesarean delivery rates [25, 26]. These findings highlight the need to carefully weigh the risks and benefits of obstetric interventions to protect maternal and neonatal health.

The sustained decline in the incidence of PPH is attributed to the progressive strengthening of global health systems and the widespread implementation of targeted interventions. Key contributors include the expanded coverage of uterotonics, which has been associated with a 60% reduction in severe PPH, related cesarean sections, and associated maternal mortality [27]. Additionally, systematic training of community health workers and effective task‐shifting strategies have enhanced the emergency response capacity of primary care facilities, particularly in South Asia [28]. However, progress has been uneven. In the Caribbean, for example, the ASMR declined only minimally (EAPC = –0.069%), a stagnation that may be linked to persistent racial inequities and substantial wealth gaps that constrain improvements in healthcare access [29, 30]. The COVID‐19 pandemic has further highlighted how public health emergencies can disrupt maternal health services [31], intensifying risks for vulnerable pregnant populations.

Literature reports indicate that iron deficiency is an independent risk factor for postpartum hemorrhage, and severe anemia during pregnancy significantly increases the risk of PPH [32]. Globally, the proportion of PPH burden attributable to iron deficiency has gradually declined, reflecting public health progress. However, this proportion has increased in low‐SDI regions, where persistently high anemia rates during pregnancy (exceeding 40%), low iron supplementation coverage (below 50%), and cultural dietary practices have collectively impeded improvements [33, 34, 35]. Consequently, iron deficiency now constitutes a growing share of the PPH burden in these settings. From a mechanistic perspective, maternal iron deficiency anemia may increase PPH risk through multiple pathways. Anemia can induce compensatory placental hypertrophy and uteroplacental vascular remodeling, while elevated nitric oxide and progesterone levels may impair myometrial sensitivity to oxytocin, disrupting third‐stage uterine contractions. Prior oxytocin use for labor induction may further exacerbate this effect. These mechanisms may explain the increased susceptibility of iron‐deficient women to atonic PPH [36]. From a socioeconomic perspective, iron deficiency disproportionately affects women in resource‐limited settings, where dietary diversity is limited, antenatal iron supplementation remains inadequate, and access to skilled birth attendance and emergency obstetric care is often constrained. Addressing this persistent disparity requires targeted nutritional interventions and health system strengthening. Digital pregnancy management platforms offer potential to enhance monitoring and intervention efficiency [37]. In resource‐limited settings, scalable low‐cost tools such as non‐pneumatic anti‐shock garments can help stabilize patients during emergencies, buying critical time for referral and improving survival outcomes [38].

The age group of 20–24 years bears the highest incidence and mortality burden of PPH, accounting for 28.2% of global cases in 2021. This pattern may be explained by the elevated fertility rate and the high proportion of first births in this demographic. In sub‑Saharan Africa, adolescent pregnancy rates are notably above the global average [39], influenced by factors such as religious norms, early marriage, limited education, and poverty. Low antenatal care coverage further exacerbates PPH risk in these settings. To address this, policymakers and community leaders should strengthen comprehensive sexuality education, support girls to stay in school, and implement community‐based awareness programs. Effective adolescent pregnancy prevention initiatives should include peer‑involvement mechanisms and engagement of key stakeholders. Evidence suggests that establishing adolescent‐friendly health services in schools and clinics can help reduce adolescent pregnancy rates. These services, when combined with youth capacity‐building initiatives, may further enhance their effectiveness. Such a multi‑level approach can help build an integrated prevention system spanning knowledge, social support, and healthcare access [40].

The contrasting trends between the SII and CI provide a nuanced perspective on changing health inequities. The SII improved substantially, indicating that the absolute gap in PPH burden between the highest and lowest socioeconomic groups has narrowed over time. In contrast, the CI became more negative, reflecting a growing relative concentration of the remaining burden among the most disadvantaged populations. This divergence suggests that while absolute disparities have decreased, relative inequality has intensified. In other words, despite overall progress in reducing the global PPH burden, the proportional share borne by lower‐resource countries has increased. These findings underscore the need for targeted, equity‐oriented interventions. Priority should be given to expanding access to quality obstetric care in underserved regions, addressing barriers such as healthcare infrastructure and skilled birth attendance. Furthermore, addressing social determinants, including poverty, education, and nutrition, remains essential for achieving sustainable reductions in maternal health disparities. The persistent inequality in PPH burden highlights that global progress has not been uniformly distributed, reinforcing the importance of continued efforts to promote health equity.

The substantial burden of PPH in low‐SDI regions is shaped by intersecting socioeconomic, educational, and infrastructural factors. Economically, vast disparities in health expenditure persist across regions. For example, per capita health spending in sub‐Saharan Africa is less than $100, compared with over $5,000 in high‐income countries. This gap severely limits access to essential equipment and medications, contributing to preventable deaths. Cross‐country analyses reveal that the drivers of medical investment differ across regions. High‐income countries typically allocate resources through public budgets that are closely aligned with health needs. In contrast, upper‐middle‐income settings rely more on urbanization and industrial employment, with investment levels constrained by overall economic capacity [41, 42]. This disparity reinforces inequitable healthcare environments and contributes to unequal health outcomes. Educational disparities compound these challenges. In countries such as Nigeria, where only about 40% of women complete secondary education, limited health literacy and awareness reduce antenatal care uptake. Evidence suggests that higher educational attainment, particularly among male partners, is positively associated with improved utilization of maternal health services [40], and broader social inequalities in education are linked to worse health outcomes for women [43]. Geographic and infrastructural barriers further entrench spatial inequities. Skilled birth attendance remains critically low in many settings—for example, 5.0% in Equatorial Guinea and 9.8% in Ethiopia—compared with 40.1% in Kenya [44]. Inadequate midwifery coverage delays care and increases risks during delivery. Addressing these multilevel gaps requires integrated strategies that combine spatial, human‑resource, and economic interventions. Promising examples include Rwanda's drone‑based blood delivery system, which reduces transport times to under 45 min [45], and solar‑powered maternity units that improve service availability [46]. Strengthening the health workforce is also essential; in Sierra Leone, for instance, fewer than 500 midwives serve the population, with over 40% concentrated in urban areas [47]. Implementing tiered training programs aligned with WHO standards—such as a 72‑hour simulation‑based curriculum for community midwives—could help build capacity. Finally, expanding community health insurance through legislative and financial support can improve financial access to care and reduce the burden of maternal mortality [48].

Based on the findings of this study, several recommendations are proposed. First, priority should be given to regions with lower socio‐demographic indicators and a high incidence of PPH, particularly sub‑Saharan Africa and South Asia. In these settings, the promotion of PPH prevention kits—containing essential supplies such as oxytocin, non‑pneumatic anti‑shock garments, and rapid screening tools—is recommended [49, 50]. Second, low‑income countries should be supported in developing electronic health registration systems and adopting standardized PPH diagnostic protocols [51]. Additionally, training and mobilizing community health workers can strengthen local capacity for PPH prevention and management. Efforts to promote adolescent sexual health education are also encouraged to reduce unplanned pregnancies [39]. Together, improving antenatal care access, enhancing emergency obstetric readiness, and monitoring policy implementation can contribute to safer postpartum outcomes.

This study has several limitations. First, although the GBD database provides standardized and comparable estimates, its reliance on modelled data may obscure local realities and introduce bias in regional interpretations. Second, data quality varies across sources, with reporting from low‑income countries often being incomplete or less accurate [52, 53], which challenges precise burden assessment. Furthermore, clinical underestimation of blood loss during delivery may bias morbidity statistics and affect the reliability of findings [54]. Future research should incorporate cross‑validation with local survey data (e.g., demographic health surveys) to improve accuracy. The BAPC projections are based on linear trends and do not account for potential disruptions from public health emergencies such as pandemics or conflicts. The GBD framework identifies iron deficiency as a key modifiable risk factor for maternal hemorrhage. However, it does not include other well‐established clinical and obstetric risk factors. These include high parity, prior PPH, placental abnormalities, prolonged labor, and coagulation disorders. This omission limits the comprehensiveness of risk attribution. GBD 2021 data extend only through 2021 and cannot capture the effects of recently scaled interventions, which may affect the timeliness of policy insights. Future studies would benefit from multi‑center clinical data and dynamic modelling approaches to support more agile and context‑adapted maternal health strategies. Finally, the projection models used in this study, while robust for estimating long‐term trends, did not account for potential public health emergencies such as pandemics or conflicts. This exclusion may affect forecast accuracy in several ways. First, models based on historical data assume that future patterns will follow past trends, potentially underestimating sudden disruptions. Second, emergencies can severely strain health systems, diverting resources away from routine maternal care and temporarily increasing PPH‐related mortality. Third, such events may exacerbate underlying risk factors, including iron deficiency prevalence, through supply chain interruptions and reduced access to nutrition programs. Conversely, forecasts may overestimate future burden if they fail to capture accelerated improvements following emergency responses. Therefore, our projections should be interpreted as estimates under stable conditions, with actual outcomes potentially deviating during unforeseen global health crises.

5. Conclusion

Despite the global decline in PPH incidence, persistent regional inequalities continue to hinder progress toward SDG 3.1. Achieving the projected reduction in maternal mortality to 1.23 per 100,000 by 2035 will require targeted, data‐driven strategies. These strategies must prioritize high‐burden, low‐resource regions. They should also address key determinants of PPH burden, including the persistent role of iron deficiency as a modifiable risk factor and the high burden among women aged 20–24 years. Intensified global cooperation and equitable resource mobilization remain critical for strengthening health systems and implementing targeted interventions, particularly in regions such as sub‐Saharan Africa.

Author Contributions

Jidong Huang: conceptualization, formal analysis, methodology, writing – original draft, writing – review and editing. Xiujing Lu: formal analysis, visualization; writing – original draft, writing – review and editing. Yachang Zeng: funding acquisition, investigation, project administration, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Transparency Statement

The lead author Yachang Zeng affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

Supporting information

Table S1: Incidence of postpartum hemorrhage in fertile female between 1990 and 2021 at the national level. Table S2: Deaths of postpartum hemorrhage in fertile female between 1990 and 2021 at the national level. Table S3: DALY of postpartum hemorrhage in fertile female between 1990 and 2021 at the national level.

Huang J., Lu X., and Zeng Y., “Global, Regional, and National Epidemiology of Postpartum Hemorrhage (1990–2021): a Statistical Analysis of Incidence, Mortality, and DALYs,” Health Science Reports 9 (2026): e72163. 10.1002/hsr2.72163.

Data Availability Statement

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