Causes of and risk factors for postpartum haemorrhage: a systematic review and meta-analysis

Summary

Background

An understanding of the causes of postpartum haemorrhage is needed to provide appropriate treatment and services. Knowledge of the risk factors for postpartum haemorrhage can help address modifiable risk factors. We did a systematic review and meta-analysis to identify and quantify the various causes and risk factors for postpartum haemorrhage.

Methods

In this systematic review and meta-analysis, we did a systematic literature search in MEDLINE, Embase, Web of Science, Cochrane Library, and Google Scholar for cohort studies of postpartum haemorrhage from Jan 1, 1960, to Nov 30, 2024 without language restrictions. At least two authors independently undertook study selection, data extraction, and quality assessment. Population-based cohort studies available in English were eligible. Rates of postpartum haemorrhage causes as well as crude and adjusted odds ratios (ORs) for risk factors were pooled using a random-effects model. Risk factors were classified as having weak, moderate, or strong association based on the pooled ORs: weak (OR >1 to 1·5), moderate (OR >1·5 to 2), and strong (OR >2). This study is registered with PROSPERO, CRD42023479686.

Findings

We synthesised data from 327 studies, including 847 413 451 women with no restriction on age, race, or ethnicity. Most studies were of high methodological quality. The pooled rates of the five commonly reported causes of postpartum haemorrhage were uterine atony (70·6% [95% CI 63·9–77·3]; n=834 707 women, 14 studies), genital tract trauma (16·9% [9·3–24·6]; n=18 449 women, six studies), retained placenta (16·4% [12·3–20·5]; n=235 021 women, nine studies), abnormal placentation (3·9% [0·1–7·6]; n=29 638 women, two studies), and coagulopathy (2·7% [0·8–4·5]; n=236 261, nine studies). The pooled rate of women with multiple postpartum haemorrhage causes was 7·8% (95% CI 4·7–10·8; n=666, two studies). Risk factors with a strong association with postpartum haemorrhage included anaemia, previous postpartum haemorrhage, caesarean birth, female genital mutilation, sepsis, no antenatal care, multiple pregnancy, placenta praevia, assisted reproductive technology use, macrosomia with a birthweight of more than 4500 g, and shoulder dystocia. Risk factors with moderate association with postpartum haemorrhage included BMI ≥30 kg/m², COVID-19 infection, gestational diabetes, polyhydramnios, pre-eclampsia, and antepartum haemorrhage. Risk factors with weak association with postpartum haemorrhage included Black and Asian ethnicity, BMI 25–29·9 kg/m², asthma, thrombocytopenia, uterine fibroids, antidepressant use, induction of labour, instrumental birth, and premature rupture of membranes.

Interpretation

The finding that uterine atony is the commonest cause of postpartum haemorrhage supports the WHO recommendation for all women giving birth to be given prophylactic uterotonics. Knowledge of risk factors with a strong association with postpartum haemorrhage can help to identify women at high risk of postpartum haemorrhage who could benefit from enhanced prophylaxis and treatment. The importance of multiple concurrent causes of postpartum haemorrhage supports the use of treatment bundles.

Funding

Gates Foundation.