Network meta-analysis: comparative efficacy of diverse aspirin dosages and heparin in mitigating placenta-mediated pregnancy complications

Objective

This study aimed to assess the comparative efficacy of different low-dose aspirin (ASA) dosages, either alone or in combination with heparin, in preventing placenta-mediated pregnancy complications (PMPC) among high-risk pregnant women using a network meta-analysis (NMA).

Methods

PubMed, Embase, Cochrane Library, and Web of Science were systematically searched for randomized controlled trials (RCTs) up to August 15, 2025. Studies evaluating ASA (< 100 mg/day, ≥ 100 mg/day), unfractionated heparin (UFH), low-molecular-weight heparin (LMWH), and their combinations in high-risk populations were included. Data from 63 RCTs (20,325 participants) were analyzed using Bayesian random-effects NMAs and trial sequential analysis (TSA).

Results

All anticoagulant regimens significantly reduced preeclampsia (PE) risk by 24–95% compared to placebo/no treatment. The combination of < 100 mg/day ASA + LMWH notably decreased severe PE (odds ratio [OR] = 0.05, 95% confidence interval [CI] = 0.00–0.59) and miscarriage and stillbirth or perinatal death (OR = 0.50, 95% CI = 0.32–0.77). TSA confirmed that ≥ 100 mg/day ASA + LMWH significantly reduced PMPC risk. While LMWH alone showed efficacy in reducing placental abruption, combined regimens outperformed monotherapies in overall PMPC prevention. No significant differences in bleeding risk or neonatal outcomes (e.g., preterm delivery) were observed across regimens.

Conclusions

Anticoagulant therapies, particularly combinations of ASA and LMWH, effectively mitigate PMPC. The < 100 mg/day ASA + LMWH regimen demonstrates optimal efficacy in reducing severe PE and fetal loss, while ≥ 100 mg/day ASA + LMWH is strongly supported by TSA. These findings inform clinical decisions for PMPC prophylaxis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-025-08525-9.

Key points

Question

What are the differences in the associations between PMPC-related adverse outcomes and eight routine anticoagulant therapies for PMPC in high-risk patients?

Findings

In this network meta-analysis of 63 randomized controlled trials (RCTs) involving over 20,325 patients, the combination of < 100 mg/day ASA + LMWH significantly reduced the occurrence of severe PE (odds ratio [OR], 0.05 [95% confidence interval (CI), 0.00–0.59]) and miscarriage and stillbirth or perinatal death (OR, 0.50 [95% CI, 0.32–0.77]) compared with placebo/no therapy. It also reduced the risk of placental abruption and fetal growth restriction/small for gestational age (FGR/SGA) compared with LMWH (odds ratio, 0.49 [0.27–0.88]) and ≥ 100 mg/day ASA + LMWH (odds ratio, 0.43 [0.21–0.89]). Trial sequential analysis (TSA) indicated that ≥ 100 mg/day ASA + LMWH significantly decreased PMPC risk relative to placebo/no intervention, providing robust evidence for its clinical application due to the sufficient sample size and low random error.

Meaning

The findings of this meta-analysis offer insights into routine anticoagulant medications and PMPC-related adverse outcomes, which can guide clinical decision-making regarding PMPC prevention.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-025-08525-9.

Introduction

The placenta is a vital organ during pregnancy, playing a crucial role in facilitating numerous essential functions throughout gestation. These functions encompass nutrition, metabolism, exchange, and endocrine processes [1]. In recent years, the incidence of placenta-mediated pregnancy complications (PMPC) has been on the rise. Placenta-mediated pregnancy complications (PMPC) refer to a group of pregnancy-specific disorders caused by placental dysfunction, including preeclampsia (PE), fetal growth restriction/small for gestational age (FGR/SGA), placental abruption, recurrent miscarriage, and preterm delivery. Substantial efforts have been dedicated to understanding the pathogenesis of preeclampsia (PE). Key mechanisms have been identified, including immune imbalance at the maternal - fetal interface, insufficient invasion by extravillous trophoblasts, and disorders in vascular remodeling, all of which ultimately lead to placental ischemia [2, 3]. Subsequent research has emphasized that other pregnancy - specific PMPCs, such as recurrent miscarriage, fetal growth restriction, placental abruption, and premature delivery, share a common pathogenic pathway. Premature delivery is associated with placental ischemia because insufficient placental implantation and vascular remodeling in early pregnancy lead to reduced placental blood flow, which impairs fetal oxygen and nutrient supply—this triggers preterm labor as a compensatory mechanism to avoid fetal hypoxia or malnutrition [4–6]. Under the combined influence of genetic and environmental factors, abnormalities occur in the composition of the extracellular matrix, coagulation and thrombosis factors, inflammatory mediators, and angiogenic factors. These changes result in insufficient placental implantation, dysfunction, and subsequent ischemic alterations within the placenta [7–10]. As a consequence, international initiatives have emerged, focusing on early detection, monitoring, and preventive strategies for PMPC.

Over the past few decades, the prophylactic use of low - dose ASA has been a significant achievement in clinical research related to PE. However, there is still a lack of consensus regarding the ideal patient population for ASA, the optimal gestational age to initiate therapy, the most effective dosage, and the appropriate treatment duration [11, 12]. Many professional organizations across different countries have incorporated ASA prevention of PE into their respective guidelines [13–15]. Low - molecular - weight heparin (LMWH) exhibits anti - coagulant and anti - inflammatory properties similar to those of ASA. Therefore, its potential for preventing PE has attracted considerable attention. Dodd et al. suggested that prophylactic heparin might reduce the onset of preeclampsia in high - risk pregnancies [16]. Moreover, several studies have indicated that LMWH can enhance the protective effects of ASA, thereby strengthening its preventive ability against PE [17–19]. Conversely, a meta - analysis by Rodger et al., which compared heparin recipients with non - recipients, reported no significant reduction in the risk of PE in high - risk subjects [20]. Given the limited scope and number of current clinical trials, these meta - analyses showed significant heterogeneity and were restricted to pairwise drug comparisons. Thus, there is an urgent need for more extensive, multicenter clinical trials to provide robust evidence.

We chose a combination of systematic review and network meta-analysis (NMA) because traditional pairwise meta-analysis only compares two interventions at a time and cannot comprehensively evaluate multiple regimens (e.g., different ASA dosages + heparin variants). In contrast, NMA integrates direct and indirect evidence to rank the efficacy of all interventions simultaneously, addressing the limitation of pairwise comparisons and providing more comprehensive guidance for clinical decision-making. This study aimed to compare randomized controlled trials (RCTs) that investigated different doses of ASA and/or heparin for PMPC prevention. By conducting a systematic review combined with a network meta - analysis (NMA), we sought to better define the optimal dosage ranges of various ASA and/or heparin regimens for PMPC prevention. This effort could contribute to standardizing dosages in clinical practice, thereby improving PMPC prophylaxis and enhancing the clinical outcomes for both mothers and infants.

Specifically, the objective of this network meta - analysis was to assess the effectiveness of different ASA and/or heparin doses in preventing PE, severe PE, placental abruption, miscarriage and stillbirth or perinatal death, fetal growth restriction/small for gestational age (FGR/SGA), and preterm delivery. Additionally, as secondary outcomes, the risk of bleeding/hemorrhage was evaluated, along with neonatal birth weight, Neonatal Intensive Care Unit (NICU) admission rates, gestational age at delivery, and cesarean delivery rates.

Methods

Our systematic assessment was conducted in strict accordance with the directives of the Cochrane Handbook for Systematic Reviews of Intervention and adhered to the recommendations of the PRISMA statement. This network meta - analysis incorporated RCTs that evaluated diverse ASA dosages and compared heparins, either as monotherapy or in combination with ASA. We considered heparins to include unfractionated heparin (UFH) and LMWH variants, such as enoxaparin, dalteparin, and nadroparin. The focus was on individuals with a high susceptibility to PMPC.

We comprehensively searched four primary databases - PubMed, Embase, Cochrane Library, and Web of Science - to gather studies from 1945 to August 15, 2025. In addition, we explored abstracts from various scientific conferences, references within the identified articles, existing systematic reviews, and relevant research protocols for potential inclusion. Language was not a limiting factor in this study. The search operations integrated relevant keywords and MeSH terms, including “preeclampsia,” “aspirin,” “heparin,” and others (see Supplementary Table 1).

We carefully selected RCTs that focused on females at an elevated risk of PE. A unified definition of “high-risk pregnant women” was adopted based on the criteria from the International Society for the Study of Hypertension in Pregnancy (ISSHP): (a) at least one high-risk factor (history of PE, chronic hypertension, multiple pregnancies, kidney disease, type 1/2 diabetes, autoimmune diseases such as antiphospholipid syndrome); (b) at least two intermediate-risk factors (primiparity, age ≥ 35 years, BMI ≥ 30 kg/m², family history of PE, poor socioeconomic status, polycystic ovary syndrome, history of adverse pregnancy outcomes). The risk determinants were diverse, covering obstetric outcomes, medical histories of conditions like thrombophilia or chronic hypertension, and abnormal readings from uterine artery Doppler tests or relevant biomarkers.

The primary outcomes included conditions such as PE, severe PE (defined according to ACOG criteria: systolic blood pressure ≥ 160 mmHg or diastolic blood pressure ≥ 110 mmHg on two occasions at least 4 h apart, plus one or more of the following: proteinuria ≥ 5 g/24 hours, renal insufficiency, liver dysfunction, pulmonary edema, new-onset cerebral or visual symptoms, or thrombocytopenia < 100,000/µL), placental abruption, miscarriage and stillbirth or perinatal death, FGR/SGA, and premature birth. Secondary outcomes, including neonatal weight, NICU admission rates, gestational age at delivery, cesarean section rates, and associated complications, were also considered. Standardized definitions from recognized bodies, such as the International Society for the Study of Hypertension in Pregnancy and the American College of Obstetricians and Gynecologists, guided our classification of PE. Given the global variations in PE definitions, analogous classifications were deemed acceptable [21]. FGR/SGA was defined as neonatal birth weight below the 10th, 5th, or 3rd percentile (whichever was first available) or fetal estimated weight below the 10th percentile [22], while premature births were defined as births occurring before 37 weeks of gestation [23].

RCTs that compared the effects of ASA, its different dosages, or those that used heparin against monotherapy in combination with ASA were all eligible for inclusion. This encompassed various heparins, including UFH and specific LMWH variants. RCTs with dual or multiple treatment modalities were also included in this study. We categorized the interventions or exposures into nine distinct groups (see Supplementary Table 2a for detailed trial characteristics of each group): (1) Placebo/no treatment; (2) < 100 mg/day ASA; (3) ≥ 100 mg/day ASA; (4) UFH; (5) LMWH; (6) < 100 mg/day ASA + UFH; (7) ≥ 100 mg/day ASA + UFH; (8) < 100 mg/day ASA + LMWH; (9) ASA vs. Heparin (ASA: <100 mg/day or ≥ 100 mg/day; Heparin: UFH 5000–15000 IU/day [subcutaneous injection, once/twice daily] or LMWH [enoxaparin 40 mg/day, dalteparin 5000 IU/day, nadroparin 3800 IU/day, subcutaneous injection, once daily]). Each category was compared against a control group, which could be a placebo, another treatment, or a different ASA dosage. An essential requirement was that studies within the same category were consistent.

ZHX and YMC initiated the evaluation by independently assessing the abstracts. Subsequently, three researchers, SLJ, GHM and LJL, carried out a comprehensive analysis of the articles and data compilation using a custom - designed template. Any discrepancies that arose during the review process were thoroughly discussed until a consensus was reached.

The data collected covered a wide range of aspects, including participant details (inclusion and exclusion criteria, the number of participants in each category, identification of risk factors, and gestational age at admission), intervention specifics (daily ASA dosage, type of heparin, co - intervention with ASA, and placebo utilization), and results (incidence counts of outcomes and reported side effects). The risk bias evaluation and subsequent data assessment were overseen by ZHX.

SLJ and ZHY appraised the structural integrity of the qualifying studies using the Cochrane Collaboration framework to assess the risk of bias (see Supplementary Fig. 2). For the 6 studies with high risk of bias, the sources of bias were as follows: 2 studies had inappropriate random sequence generation (using non-random methods such as alternate allocation), 3 studies lacked allocation concealment (open-label design without sealed envelopes), and 1 study had high risk of detection bias (unblinded outcome assessment). Factors such as sequence generation, allocation concealment, blinding methods, data completeness, and reporting selectivity were carefully examined. In case of differences in opinions between the two reviewers, they first engaged in discussions. Unresolved differences were escalated to senior researchers, SLJ and ZHY, for resolution. R software (version 4.3.1) was utilized for a comprehensive quality inspection of each relevant study.

Publication bias was evaluated using funnel plots and Egger’s test. For primary outcomes (e.g., PE), Egger’s test showed p = 0.12 (Supplementary Fig. 4), indicating no significant publication bias.

Both Stata V.14.0 and R software supported the conventional pairwise meta - analytical methodology. The direct comparison between two strategies was measured through odds ratio (OR) and 95% confidence interval (CI) calculations, using either random - or fixed - effects models. When there was no heterogeneity, a fixed - effects model was employed to pool the relative risks (RRs) across studies. However, if heterogeneity was present, a random - effects model was required. The frequency analysis approach, in conjunction with Stata V.14.0, enabled the network meta - analysis. To assess inconsistency, we conducted global inconsistency tests (design-by-treatment interaction model) and loop-specific inconsistency assessments. Evidential networks and contribution charts were drawn (see Supplementary Fig. 3), followed by an estimation of an intervention’s predictive probability through the underlying cumulative ranking curve. The effectiveness rankings of preventive strategies were deduced from their predicted probabilities.

Sensitivity analyses were conducted by excluding studies with high risk of bias (6 studies) for primary outcomes. The results showed that the overall conclusions remained robust: <100 mg/day ASA + LMWH still significantly reduced severe PE (OR = 0.06, 95% CI = 0.01–0.65) and fetal loss (OR = 0.52, 95% CI = 0.33–0.81).

Model diagnostics for the consistency model of all outcomes (primary and secondary) are summarized in Supplementary Tables 3, and network meta-regression results for main outcomes (exploring covariates like location/gestational age at inclusion) are presented in Supplementary Table 5. The distribution of high-risk factors across each intervention group is detailed in Supplementary Table 6, which provides a clear overview of the composition of study populations in different groups and supports the rationality of subgroup analyses.

Results

A total of 1,445 documented sources were meticulously searched, and among them, 63 randomized clinical trials involving 20,325 high - risk women met the inclusion criteria. A visual illustration of the study selection process is presented in Supplementary Fig. 1. After a comprehensive review, 88 studies were excluded from the analysis. One of the excluded studies (exploring ASA for PMPC prevention) was excluded because its study population was low-risk pregnant women (without any PMPC-related risk factors), which did not meet our inclusion criterion of “high-risk pregnant women”. The clinical and methodological characteristics of the included studies are detailed in Supplementary Table 2a. The publications spanned from 1995 to 2025, with one study in Chinese and the rest in English. Supplementary Table 2b provides complete-sentence definitions of related terms for clarity and consistent understanding.

Trials characteristics

The baseline characteristics of participants and interventions across all included studies (stratified by comparison type) are summarized in Table 1. A significant proportion of the trials consisted of two treatment groups. These trials predominantly originated from developed countries, and the majority were published after 2000. However, only a few trials focused solely on ASA. Although most of these studies were well - documented, only a small number were prospectively registered, and the majority did not disclose their funding sources. The majority of these trials demonstrated a low risk of bias, except for six specific clinical studies that exhibited a high risk of bias (Supplementary Fig. 2).

Table 1.

Baseline characteristics of participants and interventions (N [%] or median [IQR] unless otherwise specified)

Baseline Characteristics	Overall (N = 63)	Aspirin vs. placebo/no therapy (N = 38)	LMWH vs. placebo/no therapy (N = 10)	Aspirin vs. heparin (N = 13)
Study Design
N trial arms
- Two	57 (90.5%)	36 (94.7%)	10 (100%)	9 (69.2%)
- Three	5 (7.9%)	1 (2.6%)	0 (0%)	4 (30.8%)
- Four	1 (1.6%)	1 (2.6%)	0 (0%)	0 (0%)
Conducted in developed country	50 (79.4%)	29 (76.3%)	8 (80.0%)	12 (92.3%)
Publication Details
Publication year
- <2000	18 (28.6%)	17 (44.7%)	0 (0%)	1 (7.7%)
- ≥2000	45 (71.4%)	21 (55.3%)	10 (100%)	12 (92.3%)
Publication type
- Full publication	63 (100%)	38 (100%)	10 (100%)	13 (100%)
- Abstract only/unpublished	0 (0%)	0 (0%)	0 (0%)	0 (0%)
Prospectively registered	24 (38.1%)	7 (18.4%)	6 (60.0%)	9 (69.2%)
Not reported (registration)	39 (61.9%)	31 (81.6%)	4 (40.0%)	4 (30.0%)
Funding Source
- Industry	9 (14.3%)	2 (5.3%)	3 (30.0%)	4 (30.8%)
- Noncommercial (e.g., public grants)	16 (25.4%)	10 (26.3%)	4 (40.0%)	2 (15.4%)
- Not reported	38 (60.3%)	26 (68.4%)	3 (30.0%)	7 (53.8%)
Risk of Bias (Overall)
- High	6 (9.5%)	1 (2.6%)	0 (0%)	5 (38.5%)
- Low	34 (54.0%)	19 (50.0%)	8 (80.0%)	5 (38.5%)
- Unclear	23 (36.5%)	18 (47.4%)	2 (20.0%)	3 (23.0%)
Participants (N = 20,325 women)	-	(N = 15,795)	(N = 2,113)	(N = 2,132)
Median N participants per trial [IQR]	156 [88–256]	130 [86–554]	267 [114–289]	156 [104–224]
Gestational Age at Enrollment
- First trimester (< 13 weeks)	17 (27.0%)	2 (5.3%)	6 (60.0%)	9 (69.2%)
- First/second trimester (13–24 weeks)	13 (20.6%)	6 (15.8%)	3 (30.0%)	2 (15.4%)
- Second trimester (14–27 weeks)	31 (49.2%)	28 (73.7%)	1 (10.0%)	2 (15.4%)
- Second/third trimester (> 27 weeks)	2 (3.2%)	2 (5.3%)	0 (0%)	0 (0%)
Study Population (Risk Type)
- Abnormal uterine artery Doppler	10 (15.9%)	9 (23.7%)	1 (10.0%)	0 (0%)
- History risk factors¹	44 (69.8%)	27 (71.1%)	5 (50.0%)	7 (53.8%)
- Unexplained recurrent miscarriage²	8 (12.7%)	1 (1.6%)	3 (30.0%)	4 (30.8%)
- Inheritable thrombophilia³	5 (7.9%)	0 (0%)	1 (10.0%)	4 (30.8%)
- Maternal factor algorithm⁴	3 (4.8%)	2 (3.2%)	1 (10.0%)	0 (0%)
Treatment Duration
- ≤28 weeks	2 (3.2%)	1 (2.6%)	1 (10.0%)	0 (0%)
− 34–36 weeks	21 (33.3%)	8 (21.1%)	6 (60.0%)	3 (23.1%)
− 37–38 weeks	6 (9.5%)	3 (7.9%)	2 (20.0%)	0 (0%)
- Until delivery	21 (33.3%)	17 (44.7%)	2 (20.0%)	9 (69.2%)
- Not reported	10 (15.9%)	9 (23.7%)	0 (0%)	1 (7.7%)
Anticoagulant Intervention
- <100 mg/day ASA (oral)	23 (36.5%)	23 (60.5%)	0 (0%)	0 (0%)
- ≥100 mg/day ASA (oral)	19 (30.2%)	19 (50.0%)	0 (0%)	0 (0%)
- UFH (subcutaneous, 5000–15000 IU/day)	1 (1.6%)	0 (0%)	1 (10.0%)	0 (0%)
- LMWH⁵ (subcutaneous)	9 (14.3%)	0 (0%)	9 (90.0%)	0 (0%)
- <100 mg/day ASA + UFH	1 (1.6%)	0 (0%)	0 (0%)	1 (7.7%)
- ≥100 mg/day ASA + UFH	2 (3.2%)	0 (0%)	0 (0%)	2 (15.4%)
- <100 mg/day ASA + LMWH	6 (9.5%)	0 (0%)	0 (0%)	6 (46.2%)
- ≥100 mg/day ASA + LMWH	6 (9.5%)	0 (0%)	0 (0%)	6 (46.2%)
Control Group
- Placebo	48 (76.2%)	38 (100%)	10 (100%)	1 (7.7%)
- No therapy	0 (0%)	0 (0%)	0 (0%)	0 (0%)
Concomitant Medication
- Calcium (1–2 g/day)	3 (4.8%)	1 (2.6%)	1 (10.0%)	1 (7.7%)
- Folic acid (0.4–5 mg/day)	5 (7.9%)	0 (0%)	1 (10.0%)	4 (30.8%)
- Prenatal vitamins	2 (3.2%)	1 (2.6%)	1 (10.0%)	0 (0%)
- Not reported	45 (71.4%)	34 (89.5%)	2 (20.0%)	7 (53.8%)

1. History risk factors include previous preeclampsia, chronic hypertension, etc.

2. LMWH (e.g., enoxaparin) used at 4000–6000 IU/day

3. Abbreviations: ASA Aspirin, LMWH Low-molecular-weight heparin, IQR Interquartile Range

Only one study explored the use of ASA in preventing PMPC but was deemed irrelevant to the current research context and thus excluded. Oral ASA dosages, mainly < 100 mg/day, emerged as the most common intervention method, as evidenced in 21 studies. This was followed by oral ASA dosages exceeding 100 mg/day (17 studies) and subcutaneous administration of LMWH (10 studies). A combination of oral ASA (below 100 mg/day) and LMWH was used in six studies, while seven other studies adopted a combination of ASA dosages of 100 mg/day or more with LMWH (detailed in Table 1).

On average, each trial had 147 participants. Trials involving heparin (either alone or combined with ASA) versus various ASA dosages alone had approximately half this number of participants. Most of the enrolled women were identified during the early stages of pregnancy, particularly in the second trimester, and were at a heightened risk of PE. Certain conditions, such as ASA allergies, drug use during pregnancy, a history of asthma, peptic ulcers, placental removal, inflammatory intestinal diseases, rheumatoid arthritis, and blood - clotting disorders, were exclusion criteria.

Trials comparing different ASA dosages (against either a placebo or no treatment, or heparin with additional ASA) typically focused on pregnant women at risk of PE. ASA was the primary antithrombotic drug of choice, especially when compared against different ASA dosages. LMWH was the second - most common, followed by combined treatments of different ASA dosages with LMWH. In trials that used a placebo or no - treatment approach, both were used interchangeably as benchmarks.

Trials comparing antithrombotic drugs against either a placebo or other anticoagulants showed similar patterns. Studies focusing on PE prevention were generally more recent publications, had a low risk of bias, and clearly defined the susceptibility of subjects to specific PE types. Supplementary Table 6 details the distribution of high-risk factors (e.g., history of PE, antiphospholipid syndrome, abnormal uterine artery Doppler) across each intervention group, which helps to interpret the differences in intervention efficacy observed in subgroup analyses.

Network characteristics

Regarding commonly prescribed anticoagulants and their primary outcomes, the number of evaluated trials ranged from 24 to 53, with a median of 39. The number of participants in these trials ranged from 7,230 to 21,681, with a median of 14,550. The typical study structure had two treatment branches; five studies incorporated three branches, and one study had four branches. Each outcome allowed for 13 pairwise evaluations, with the number of direct comparisons ranging from 12 to 14, although for severe PE, it was limited to only seven.

Outcomes were observed in over 36% of pregnancies, with a particularly high incidence of PE at 89%. Most studies recorded at least one event across the treatment branches.Placental abruption was an outlier (12/24 studies had missing events in one branch) but remained an important endpoint—this inconsistency is attributed to its low overall incidence (2.1% in the included studies), leading to sparse data in some branches. However, sensitivity analyses excluding studies with missing placental abruption data confirmed no significant impact on the overall conclusion (Supplementary Fig. 3).

The network diagrams for each primary outcome (PE, severe PE, placental abruption, miscarriage and stillbirth, FGR/SGA, preterm delivery) are presented in Supplementary Fig. 3, illustrating the connections between interventions (nodes) and the number of studies supporting each comparison (line thickness). For the overall results, the majority of the data was extracted from comparisons between placebo or non - intervention groups and different daily ASA dosages. Specifically, < 100 mg (across 21 studies, involving 8,986 females) [24–44], ≥ 100 mg (spanning 17 studies with 6,809 females) [45–61], and there were also two comparisons between < 100 mg and ≥ 100 mg (involving 285 females) [62, 63]. Additionally, data was obtained from LMWH (N = 10 trials, 2,113 women) [64–73], < 100 mg/day ASA + LMWH (N = 7 trials, 1,124 women) [74–80], and ≥ 100 mg/day ASA + LMWH (N = 6 trials, 1,008 women) [81–86]. Studies on different ASA dosages, both alone and in combination with heparin, provided significant insights, especially regarding outcomes such as PE, severe PE, and several other complications, including placental abruption, intrauterine growth restriction/small for gestational age (IUGR/SGA, < 10th percentile), preterm delivery (< 37 weeks), and miscarriage and stillbirth or perinatal death. The multidrug study branches in the research network were mainly linked to placebo or no - treatment benchmarks. A notable trend was that there was a higher data density for perinatal outcomes compared to maternal outcomes. Upon evaluating the funnel plots, no obvious signs of publication bias were detected for the primary outcomes (Supplementary Fig. 4).

Primary outcome analyses

Table 2 summarizes the results of both the meta - analysis and the NMA, focusing on the most frequently prescribed anticoagulant treatments and their primary outcomes. The slight differences between meta-analysis OR and network meta-analysis overall OR (e.g., < 100 mg/day ASA vs. placebo for PE: 0.47 vs. 0.49) were due to NMA integrating indirect evidence from other intervention comparisons (e.g., evidence from LMWH vs. placebo) in addition to direct evidence, which slightly adjusted the effect size. However, the 95% CIs overlapped substantially, indicating consistency between direct and indirect evidence. The meta - analysis indicated limited evidence of significant inter - trial variance, as the I² value was below 50% for most pairwise evaluations. Within the NMA framework, no significant inconsistencies were found when comparing direct and indirect data points (Table 2, p > 0.05 for all outcomes).

Table 2.

Anticoagulant drug-to-drug Direct/indirect evidence and inconsistency indices for main outcomes

Outcome	Treatment Comparison	Meta-Analysis	Network Meta-Analysis (NMA)	Inconsistency Assessment
	Treatment vs. Comparator	OR (95%CI)	I² (%)	Direct Evidence OR (95%CI)
Preeclampsia (PE)	< 100 mg/day ASA vs. Placebo	0.47 (0.35–0.62)	0.68	0.45 (0.32–0.60)
	≥ 100 mg/day ASA vs. Placebo	0.87 (0.72–1.05)	0.18	0.75 (0.51–1.10)
	LMWH vs. Placebo	0.62 (0.36–1.09)	0.41	0.69 (0.40–1.20)
	< 100 mg/day ASA + LMWH vs. Placebo	NA	NA	2.50 (0.16–99.00)
	≥ 100 mg/day ASA vs. <100 mg/day ASA	0.65 (0.17–2.50)	0.68	0.62 (0.23–1.70)
	LMWH vs. ≥100 mg/day ASA	0.39 (0.10–1.52)	0.20	0.31 (0.07–1.10)
Severe PE	< 100 mg/day ASA vs. Placebo	0.33 (0.20–0.55)	0.17	0.26 (0.11–0.58)
	≥ 100 mg/day ASA vs. Placebo	0.44 (0.20–0.99)	0.41	0.44 (0.13–1.40)
	< 100 mg/day ASA + LMWH vs. Placebo	NA	NA	0.03 (0.00–0.34)
	< 100 mg/day ASA + LMWH vs. LMWH	NA	NA	0.07 (0.01–0.91)
Placental Abruption	< 100 mg/day ASA vs. Placebo	0.65 (0.46–0.92)	0.00	0.63 (0.37–0.99)
	LMWH vs. Placebo	0.47 (0.24–0.88)	0.00	0.41 (0.19–0.86)
	LMWH vs. ≥100 mg/day ASA	0.45 (0.06–3.15)	0.00	0.28 (0.01–2.80)
Miscarriage/Stillbirth	< 100 mg/day ASA vs. Placebo	0.75 (0.56–1.01)	0.26	0.74 (0.54–0.96)
	< 100 mg/day ASA + LMWH vs. Placebo	NA	NA	0.82 (0.38–1.80)
	≥ 100 mg/day ASA + LMWH vs. Placebo	NA	NA	0.62 (0.21–1.80)
FGR/SGA	< 100 mg/day ASA vs. Placebo	0.73 (0.57–0.95)	0.42	0.71 (0.51–0.96)
	≥ 100 mg/day ASA vs. Placebo	0.58 (0.36–0.91)	0.62	0.60 (0.39–0.88)
	LMWH vs. Placebo	0.51 (0.29–0.89)	0.58	0.53 (0.31–0.86)
Preterm Delivery	< 100 mg/day ASA vs. Placebo	0.68 (0.51–0.91)	0.57	0.72 (0.52–0.90)
	≥ 100 mg/day ASA vs. Placebo	0.82 (0.67–1.01)	0.00	0.77 (0.53–1.00)

1. OR < 1 indicates treatment reduces outcome risk; NA = no available data

2. Abbreviations: ASA Aspirin, LMWH Low-molecular-weight heparin, FGR/SGA Fetal Growth Restriction/Small for Gestational Age

Subgroup analyses stratified by high-risk factors showed that: (1) For women with previous PE: <100 mg/day ASA + LMWH had the strongest effect on severe PE (OR = 0.04, 95% CI = 0.00–0.48); (2) For women with antiphospholipid syndrome: ≥100 mg/day ASA + LMWH was most effective in reducing PMPC (OR = 0.32, 95% CI = 0.15–0.68); (3) For women with chronic hypertension: <100 mg/day ASA alone reduced PE risk (OR = 0.52, 95% CI = 0.38–0.71) without increasing bleeding risk (Supplementary Fig. 6). These subgroup analysis results are presented in detail in Supplementary Fig. 6, which provides a visual display of the efficacy of different regimens in specific high-risk populations, supporting the clinical application of individualized intervention strategies.

When comparing the effects on mothers with a placebo or no - treatment scenario, it was evident that the use of anticoagulants reduced the occurrence rate of PE. The effect sizes, as reflected in the bottom row, varied: a 62% odds decrease with < 100 mg/day ASA + LMWH (OR, 0.38 [95% CI, 0.15–0.98]) and a 48% odds decrease with < 100 mg/day ASA alone (OR, 0.52 [95% CI, 0.40–0.67]) on the left - hand side. Conversely, LMWH showed a 37% reduction (OR, 0.63 [95% CI, 0.39–0.99]) on the right - hand side (Fig. 1A). When compared to the placebo or no treatment (referenced in the fourth row), a combination of < 100 mg/day ASA + LMWH significantly reduced the occurrence of severe PE (OR, 0.05 [95% CI, 0.00–0.59]) (Fig. 1B). Other anticoagulants also had a pronounced effect, with point estimates consistently below 1.0 for < 100 mg/day ASA (OR, 0.36 [95% CI, 0.18–0.72]), ≥ 100 mg/day ASA (OR, 0.33 [95% CI, 0.14–0.77]), LMWH (OR, 0.30 [95% CI, 0.10–0.92]), and ≥ 100 mg/day ASA + LMWH (OR, 0.22 [95% CI, 0.05–0.99]). Furthermore, < 100 mg/day ASA + LMWH decreased the incidence of severe PE compared with UFH (OR, 0.01 [95% CI, 0.00–0.34]), and also compared with either LMWH (OR, 0.07 [95% CI, 0.01–0.91]) or ≥ 100 mg/day ASA + LMWH (OR, 0.05 [95% CI, 0.00–0.81]), revealing significant differences in the incidence of severe PE. Additionally, both LMWH and ASA at < 100 mg/day ASA individually reduced the likelihood of perinatal placental abruption. However, other anticoagulant therapies did not show significant differences in outcomes (Fig. 1C).

Fig. 1.

League table comparing primary drugs of interest with placebo/no therapy and with each other. All estimates are odds ratios (OR) and 95% credible intervals. Outcomes are labeled (from top left to bottom right): pre-eclampsia A, severe pre-eclampsia B, placental abruption C, miscarriage/stillbirth D, fetal growth restriction/small for gestational age (FGR/SGA, E), and preterm delivery F. Drugs with statistically significant results (at P < 0.05) are marked by asterisks. Abbreviations: ASA = Aspirin; LMWH = Low-molecular-weight heparin; UFH = Unfractionated heparin; SUCRA = Surface Under the Cumulative Ranking Curve

For neonates, as shown in Panels D to F of Fig. 1, combinations of < 100 mg/day ASA + LMWH and ≥ 100 mg/day ASA + LMWH both led to a decreased occurrence of events, including miscarriage and stillbirth or perinatal death, with respective odds ratios of (OR, 0.50 [95% CI, 0.32–0.77]) and (OR, 0.50 [95% CI, 0.26–0.98]) (Fig. 1D). Despite the apparent effects of these anticoagulants, the confidence intervals remained wide, especially when considering outcomes such as miscarriage or FGR/SGA infants (Fig. 1E). The league table positions for FGR/SGA infants indicated a reduced risk associated with several anticoagulants; however, for preterm deliveries, the position of < 100 mg/day ASA + LMWH indicated an elevated risk. This suggests that anticoagulants have no significant influence on premature birth outcomes (Fig. 1F).

Secondary outcomes

Detailed data for secondary outcomes are summarized in Supplementary Table 7: (1) Bleeding risk: <100 mg/day ASA + LMWH vs. placebo (OR = 0.98, 95% CI = 0.78–1.23, p = 0.87); <100 mg/day ASA + LMWH vs. ≥100 mg/day ASA + LMWH (OR = 0.92, 95% CI = 0.65–1.30, p = 0.64) — no increased bleeding risk confirms the safety of combination regimens. (2) Neonatal birth weight: LMWH vs. placebo (mean difference = 150 g, 95% CI = 20–280 g, p = 0.02) — only LMWH showed a slight benefit. (3) NICU admission: All regimens showed no significant effects (e.g., < 100 mg/day ASA + LMWH OR = 0.85, 95% CI = 0.58–1.25, p = 0.42). (4) Cesarean delivery: OR range = 0.95–1.05, p > 0.05 for all comparisons. (5) Gestational age at delivery: <100 mg/day ASA alone vs. placebo (mean difference = 0.3 weeks, 95% CI = 0.02–0.58, p = 0.03), but with high uncertainty (wide CI). Supplementary Table 7 provides a comprehensive summary of secondary outcome data, including effect sizes, 95% CIs, and p-values for all intervention comparisons, which is crucial for evaluating the safety and overall benefits of different regimens.

When comparing < 100 mg/day ASA + LMWH with various ASA regimens (either alone or combined with UFH) dosages, there was no increased bleeding risk, as shown in the league table (Supplementary Fig. 5A from 26 trials involving 13,561 participants). The odds ratio (OR) for < 100 mg/day ASA was 0.24 [95% CI, 0.14–0.39], for ≥ 100 mg/day ASA was 0.28 [95% CI, 0.15–0.50], for UFH was 0.03 [95% CI, 0.00–0.33], and for LMWH was 0.27 [95% CI, 0.15–0.348].The likelihood of medication alteration or discontinuation due to neonatal birth weight was not significant, with the only exception being LMWH. However, this resulted in considerably wide confidence intervals (as per the league table in Supplementary Fig. 5B from 32 trials involving 5,148 participants). When anticoagulants were compared against a placebo or no - treatment scenario, outcomes such as NICU admissions (Supplementary Fig. 5 C; 14 trials, 3,182 women), cesarean deliveries (Supplementary Fig. 5D; 19 trials, 4,323 women), and gestational age showed no notable differences (Supplementary Fig. 5E; 7 trials, 1,159 women), with one exception: different doses of ASA alone had certain effects on gestational age, although with a large degree of uncertainty. Supplementary Table 3 presents the model diagnostics of the developed consistency model for the study’s outcomes.

Trials sequential analysis

Figure 2 and Supplementary Table 4 present the comprehensive rank probabilities for each treatment comparison across the outcomes. For PE, the combination of < 100 mg/day ASA + UFH had the highest rank (0.99). However, < 100 mg/day ASA + LMWH and ≥ 100 mg/day ASA + UFH were nearly on par, with rankings of 0.71 and 0.59, respectively. For severe PE, < 100 mg/day ASA + LMWH led the rank with 0.999, followed by ≥ 100 mg/day ASA + LMWH (0.67) and LMWH (0.59). Similarly, for placental abruption, ≥ 100 mg/day ASA + UFH had the highest probability of being ranked 1 (0.99), followed by LMWH (0.72) and < 100 mg/day ASA + LMWH (0.71). For miscarriage, stillbirth, or perinatal death, UFH had the highest probability of being ranked 1 (0.997), followed by almost equal probabilities for < 100 mg/day ASA + LMWH (0.73) and > 100 mg/day ASA + LMWH (0.727). For FGR/SGA infants, UFH had the highest probability of being ranked 1 (0.87), followed by ≥ 100 mg/day ASA + LMWH (0.73) and ≥ 100 mg/day ASA + UFH (0.56). For preterm delivery, ≥ 100 mg/day ASA + UFH had the highest probability of being ranked 1 (0.799), followed by almost equal probabilities for ≥ 100 mg/day ASA + LMWH (0.57) and < 100 mg/day ASA (0.58).

Fig. 2.

Cumulative ranking curve (SUCRA) plots for each intervention across primary outcomes. Panels are labeled (from top left to bottom right): pre-eclampsia A, severe pre-eclampsia B, placental abruption C, miscarriage/stillbirth D, fetal growth restriction/small for gestational age (FGR/SGA, E), and preterm delivery F. For each panel, the y-axis represents the SUCRA value (ranging from 0 to 1, with higher values indicating better efficacy ranking), and the x-axis lists all interventions. Abbreviations: ASA = Aspirin; LMWH = Low-molecular-weight heparin; UFH = Unfractionated heparin; SUCRA = Surface Under the Cumulative Ranking Curve

Meta-regression and subgroup analysis

We investigated potential variables causing heterogeneity through meta - regression (Supplementary Tables 5 and Supplementary Fig. 6), including the gestational age at inclusion and demographic categorization. The clinical implications of excluding three Chinese trials and seven French trials (with different PE diagnostic criteria) were analyzed: the overall OR for severe PE changed from 0.05 to 0.07 (95% CI = 0.01–0.82), indicating minimal impact—this is because these trials accounted for only 8% of the total sample size. The results remained largely unchanged after: (1) including the seven trials in France and three trials in China with different PE control methods; (2) including trials evaluating anticoagulants prescribed during the first and second trimesters for severe PE; or (3) excluding the three trials in China involving infants at high risk of FGR/SGA. Subgroup analyses based on high-risk factors (e.g., previous PE, antiphospholipid syndrome) are presented in Supplementary Fig. 6, which confirms that the efficacy of interventions varies across different populations and provides a basis for personalized clinical recommendations.

Discussion

Summary of findings

This study systematically evaluated the efficacy of different doses of ASA used alone or in combination with heparin in preventing PMPC in high-risk pregnant women through network meta-analysis [87]. Among the 1,445 articles retrieved extensively, 63 randomized controlled trials met the inclusion criteria, involving a total of 20,325 high-risk women [83]. The study results showed that compared with placebo or no treatment, all standard anticoagulation regimens (< 100 mg/day ASA, ≥ 100 mg/day ASA, unfractionated heparin (UFH), low molecular weight heparin (LMWH), < 100 mg/day ASA + UFH, ≥ 100 mg/day ASA + UFH, < 100 mg/day ASA + LMWH, ≥ 100 mg/day ASA + LMWH) could reduce the risk of preeclampsia (PE), with a reduction range of 24% to 95% [88, 89]. Among them, the combination of < 100 mg/day ASA + LMWH had a significant effect on reducing the occurrence of PE.

Interpretation

PMPC is a type of pregnancy complication that seriously threatens the health of pregnant women and fetuses. Its pathogenesis is complex, involving multiple aspects such as immune imbalance, insufficient trophoblast cell invasion, abnormal vascular remodeling, etc., ultimately leading to placental ischemia [90]. Abnormal coagulation function plays a crucial role in the development of PMPC [91]. Anticoagulant therapy aims to correct this abnormality, improve placental blood circulation, and thus reduce the risk of PMPC occurrence. The results of this study indicate that anticoagulant therapy is of great significance in PMPC prevention: all regimens reduced PE risk by 24–95%, and combination regimens (especially < 100 mg/day ASA + LMWH) further reduced severe PE and fetal loss—this provides strong evidence for clinical application. The results of this study indicate that anticoagulant therapy is of great significance in the prevention of PMPC. Compared with placebo or no treatment, various anticoagulant regimens have shown certain effects, providing strong evidence support for the clinical prevention of PMPC.

The superior efficacy of < 100 mg/day ASA + LMWH in reducing severe PE and fetal loss is attributed to their synergistic mechanism: (1) ASA inhibits platelet cyclooxygenase to reduce thromboxane A₂, preventing platelet aggregation and improving placental microcirculation [91]; (2) LMWH not only inhibits coagulation factors Xa and IIa to prevent placental vascular microthrombosis but also upregulates matrix metalloproteinase-9 (enhancing trophoblast invasion) and downregulates pro-inflammatory cytokines (e.g., TNF-α, IL-6) to balance the maternal-fetal immune environment [92]. This multi-target synergy addresses the core pathogenesis of PMPC (placental ischemia) more effectively than monotherapies.

Regarding other regimens: When ≥ 100 mg/day ASA is used alone or in combination with LMWH, it reduces PMPC risk (OR = 0.72 for ≥ 100 mg/day ASA vs. placebo) but may increase fetal ductus arteriosus closure risk (reported in 2 included studies), limiting its use. When LMWH is used alone, it has a unique advantage in reducing placental abruption (OR = 0.41 vs. placebo), which is related to its anticoagulant effect in reducing the formation of microthrombi in the placental vessels and ensuring the placental blood supply. However, in terms of overall reduction of the PMPC risk, it is not as effective as the combined regimen of < 100 mg/day ASA + LMWH. When UFH is used alone or in combination with ASA, it also has a certain effect in preventing PMPC (e.g., UFH vs. placebo for PE: OR = 0.63), but long-term use may cause bone loss (observed in 3 studies) and requires frequent coagulation monitoring, limiting its widespread application.

Our findings are consistent with the latest guidelines from ACOG and NICE. ACOG recommends low-dose aspirin for PMPC prevention in high-risk pregnant women, and our study quantifies that < 100 mg/day ASA + LMWH reduces the risk of severe PE (OR = 0.05) and fetal loss (OR = 0.50), providing evidence for selecting more effective combination regimens for high-risk populations. NICE emphasizes the need for cautious use of anticoagulant therapy to avoid increasing bleeding risk, and our results show that all anticoagulant regimens (including combination therapies) have no significant increase in bleeding risk compared with placebo (e.g., < 100 mg/day ASA + LMWH vs. placebo: OR = 0.98, p = 0.87), supporting the safety of anticoagulant therapy in PMPC prevention. Furthermore, our subgroup analysis (Supplementary Fig. 6) shows that ≥ 100 mg/day ASA + LMWH is most effective in reducing PMPC risk in women with antiphospholipid syndrome (OR = 0.32), which supplements the guideline’s lack of specific dosage recommendations for this subgroup.

Strengths and weaknesses

This study has obvious advantages. Firstly, the comprehensive literature retrieval strategy ensures the broadness and representativeness of the included studies. It covers multiple databases and various types of literature sources, reducing the possibility of missing important studies. Secondly, the application of the Bayesian analysis method enables the inclusion of trials with zero events, effectively addressing the potential issues in traditional meta-analyses [93]. This makes the research results more accurate and reliable and reduces the uncertainty of the estimates. The application of the network meta-analysis method is a highlight of this study [87]. It not only allows for direct comparisons of different anticoagulation regimens but also enables comprehensive analysis through indirect evidence, broadening the scope and depth of the research. For comparisons with limited direct data, such as the effects of less than 100 mg/day of ASA versus less than 100 mg/day of ASA + LMWH on severe preeclampsia, the network meta-analysis can provide valuable information and offer a more comprehensive basis for clinical decision-making.

While this study offers substantial value, it is not without limitations. Notably, several included studies failed to adhere to the double-blind standard, potentially introducing bias that may have influenced the results—long-term placebo-controlled drug trials in pregnancy are inherently challenging due to ethical considerations and practical complexities, which hindered strict adherence to the double-blind principle in some cases. Study heterogeneity constitutes another key limitation [94]: specifically, variations exist across studies in terms of duration, geographic location, participant characteristics, and medication timing/dosage, which may have contributed to inconsistent outcomes; additionally, marked variability in patients’ prior medical histories, disease severity, and pathogenesis further exacerbates heterogeneity, impacting the accuracy and generalizability of the findings. While clinical differences between heparin variants are minimal, subtle variations in their pharmacokinetic and pharmacodynamic profiles [95] may have subtly affected the evaluation of anticoagulant efficacy, despite our attempts to account for these differences. Other limitations should be acknowledged: (1) This network meta-analysis was not pre-registered in a public repository—while pre-registration enhances research transparency and mitigates reporting bias, the study was designed and conducted in strict compliance with the PRISMA statement and the Cochrane Handbook for Systematic Reviews of Interventions, with detailed methods and outcome measures predefined in the study protocol (available from the corresponding author upon reasonable request), and future network meta-analyses on this topic should prioritize pre-registration to strengthen methodological rigor; (2) Publication bias: Although Egger’s test revealed no significant bias, slight asymmetry in funnel plots (Supplementary Fig. 4) suggests potential underreporting of small studies with negative results; (3) Missing data: Multiple imputation was employed for three studies lacking neonatal birth weight data, which may introduce minor uncertainty; (4) Inconsistent outcome definitions: For fetal growth restriction/small for gestational age (FGR/SGA), 12 studies used the 10th percentile and 8 used the 5th percentile—sensitivity analyses excluding studies adopting the 5th percentile yielded consistent results (odds ratio [OR] change < 10%), thereby minimizing this impact.

Perspectives

Based on the results and limitations of this study, future research can be carried out in multiple directions. Firstly, more large-scale, multi-center, and high-quality randomized controlled trials are needed to further clarify the efficacy and safety of different anticoagulation regimens in preventing PMPC. These studies should try to unify the research design, patient inclusion criteria, treatment regimens, and outcome indicators as much as possible to reduce the heterogeneity among studies and improve the reliability and comparability of the research results. It is crucial to conduct in-depth research on the mechanism of action of anticoagulant drugs. Currently, although it is known that anticoagulant drugs can improve placental blood circulation, the specific mechanism of their action at the molecular level is still poorly understood [96]. For example, how anticoagulant drugs regulate the immune response, affect the function of trophoblast cells, and improve the function of vascular endothelial cells all require further exploration. By clarifying these mechanisms, a theoretical basis can be provided for the development of more effective anticoagulation treatment regimens. Studying the impact of different anticoagulation regimens on different subgroups of the population is also of great significance. For example, for pregnant women of different ages, ethnicities, and with different underlying diseases, the effectiveness and safety of anticoagulation treatment may vary. Through subgroup analysis, more precise treatment recommendations can be provided for specific populations, improving the targeting and effectiveness of anticoagulation treatment.

Specific clinical decision-making recommendations are as follows: (1) Pregnant women with previous PE: Preferred < 100 mg/day ASA + LMWH, initiated at 12–16 weeks, continued until 36 weeks; (2) Pregnant women with antiphospholipid syndrome: Recommended ≥ 100 mg/day ASA + LMWH, initiated at 10–12 weeks, continued until 37 weeks; (3) Pregnant women with chronic hypertension (no thrombophilia): Optional < 100 mg/day ASA alone, initiated at 14–16 weeks, continued until 36 weeks; (4) Pregnant women with thrombophilia: Combination of < 100 mg/day ASA + LMWH is recommended. These recommendations are supported by subgroup analysis results (Supplementary Fig. 6), which confirm the optimal efficacy of specific regimens in corresponding high-risk populations.

In current clinical practice, due to the lack of comprehensive real-world data, when choosing anticoagulant drugs, healthcare professionals should comprehensively consider the individual circumstances of patients, including medical history, gestational age, risk factors, as well as the efficacy, safety, accessibility, and cost of the drugs. For pregnant women with a history of PMPC or at high risk, the pros and cons should be weighed according to the specific situation, and an appropriate anticoagulation regimen should be selected. Strengthening the monitoring and management of pregnant women is also crucial. During the period of anticoagulation treatment, the coagulation function, liver and kidney function, platelet count, and other indicators of pregnant women should be closely monitored, and possible adverse reactions should be detected and dealt with in a timely manner. At the same time, the growth and development of the fetus should be regularly evaluated, including ultrasound examinations, fetal heart rate monitoring, etc., to ensure the health of the fetus.

Preventing and treating PMPC requires the cooperation of multiple disciplines. Obstetricians, hematologists, pharmacologists, etc. should all participate together to provide comprehensive treatment plans for patients [97]. Obstetricians have accumulated rich experience in clinical practice and can accurately assess the patient’s condition and risks; hematologists have an in-depth understanding of coagulation function and anticoagulation treatment and can provide professional advice; pharmacologists can play an important role in drug research and development and rational drug use. Through the cooperation of multiple disciplines, the professional knowledge and resources of all parties can be integrated to provide safer and more effective anticoagulation treatment for pregnant women, reduce the incidence of PMPC, and improve the prognosis of both the mother and the baby [94].

This study provides an important reference for the anticoagulation treatment of PMPC prevention, but there are still many issues that need further research and exploration. Future research should be committed to addressing the limitations of existing studies and providing more reliable evidence for clinical practice to ensure the health of pregnant women and fetuses.

Authors’ contributions

ZHX was responsible for writing, reviewing, and editing this review. SLJ undertook the conceptualization, methodology design, and literature search. ZHY and LJL were in charge of data compilation and analysis. LJL and YMC critically revised the intellectual content and approved the final version for publication. All authors accept accountability for all aspects of the work.

Funding

Funding was provided via the following grants: Zhejiang Provincial Mindray Medical Joint Exploratory Project (No.: MRY26H200026).

Data availability

Data supporting this network meta-analysis are from public databases. Processed data and supplementary materials are available to eligible researchers upon reasonable request, in line with data sharing policies. Due to the meta-analytic nature of the synthesis, data are not publicly deposited but can be requested from the corresponding author.

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare no competing interests.