Transfusion of blood and blood products for the management of postpartum haemorrhage.

Abstract

Rationale: Postpartum haemorrhage (PPH) is commonly defined as blood loss of 500 mL or greater within 24 hours after birth. Intravenous transfusions of whole blood, red blood cells (RBC), or other blood components collected from a donor may be administered to manage PPH. Key questions remain regarding optimal timing for initiating blood and blood product transfusion in managing PPH and whether the use of fractionated blood products, either as replacement for or in addition to whole blood transfusion, could improve maternal outcomes. No systematic review has examined appropriate transfusion strategies for managing PPH.

Objectives: To assess the benefits and harms of transfusion of whole blood or other blood products for preventing morbidity and mortality among women with PPH.

Search methods: We searched CENTRAL, MEDLINE, Embase, and two trials registers, together with reference checking, citation searching, and contact with study authors to identify studies for inclusion in the review. The latest search was 18 July 2024.

Eligibility criteria: We considered randomised controlled trials (RCTs), cluster-randomised trials, or controlled non-randomised studies of interventions (NRSI) evaluating the efficacy and safety of blood transfusion for managing PPH, regardless of the mode of birth.

Outcomes: Our critical outcomes were maternal death, severe maternal morbidity, and adverse effects.

Risk of bias: We assessed risk of bias in included studies using the Cochrane RoB 2 tool and the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool.

Synthesis methods: We synthesised results for each outcome within each comparison using meta-analysis where possible, and used GRADE to assess the certainty of evidence for each outcome.

Included studies: We included 12 studies with 17,868 participants. We excluded five NRSIs from outcome analyses due to critical risk of bias related to confounding.

Synthesis of results: One threshold for initiating transfusion versus another threshold for initiating transfusion None of the studies assessed this comparison. One- to two-unit RBCs versus no transfusion Among women with moderate blood loss, low-certainty evidence from one NRSI found that transfusing 1 to 2 units of RBCs to treat PPH may increase severe maternal morbidity - composite excluding intensive care unit (ICU) admission (risk ratio (RR) 7.00, 95% confidence interval (CI) 2.75 to 17.80; 2130 women) and severe maternal morbidity - ICU admission (RR 2.12, 95% CI 1.20 to 3.75; 2130 women), though we have substantial concerns about the potential bias due to confounding as the volume of blood lost was not controlled for in the analysis. The study did not report maternal death or adverse effects. Packed RBCs versus whole blood versus combination of blood products One NRSI assessed this comparison but had critical risk of bias and was subsequently excluded from analysis. Fresh frozen plasma (FFP)/RBCs with fibrinogen concentrate versus FFP/RBCs alone One NRSI assessed this comparison but had critical risk of bias and was subsequently excluded from analysis. Fibrinogen concentrate versus placebo or no fibrinogen concentrate The evidence is very uncertain about the effect of fibrinogen concentrate on maternal death (0 events; 2 studies, 674 women; very low-certainty evidence). Fibrinogen concentrate probably results in little to no difference in severe maternal morbidity - ICU admission (RR 1.09,0 95% CI 0.80 to 1.49; 2 studies, 485 women; moderate-certainty evidence). The evidence is very uncertain about the effect of fibrinogen concentrate on severe maternal morbidity - arterial embolisation (1 study, 430 women; very low-certainty evidence). One RCT (430 women) and one NRSI (730 women) reported severe maternal morbidity - hysterectomy, each of which reported different directions of effect with low-certainty evidence. Fibrinogen concentrate may result in little to no difference in adverse effect - thromboembolic events (RR 0.19, 95% CI 0.01 to 3.95; 2 studies, 674 women; low-certainty evidence). The evidence is very uncertain about the effects of fibrinogen concentrate on additional adverse effects, such as shivering or fever (1 study, 244 women; very low-certainty evidence). Cryoprecipitate versus no cryoprecipitate The evidence is very uncertain about the effect of cryoprecipitate on maternal death. One RCT (0 deaths; 180 women; very low-certainty evidence) and one NRSI (0 deaths; 157 women; very low-certainty evidence) reported this outcomes. The evidence is also very uncertain about the effects of cryoprecipitate on severe maternal morbidity - ICU admission, severe maternal morbidity - any organ failure, severe maternal morbidity - laparotomy, or severe maternal morbidity - uterine artery embolisation (1 study, 180 women; very low-certainty evidence). One RCT (180 women; very low-certainty evidence) and one NRSI (157 women; very low-certainty evidence) reported severe maternal morbidity - hysterectomy and the evidence is very uncertain. The evidence is also very uncertain about the effects of cryoprecipitate on adverse effects, such as thromboembolic events or transfusion-related reactions (1 study, 180 women; very low-certainty evidence). Massive transfusion protocol versus no massive transfusion protocol Two NRSIs assessed this comparison but had critical risk of bias and were subsequently excluded from analysis.

Authors' conclusions: Overall, available evidence for the effects of blood and blood product transfusion on priority maternal outcomes is largely uncertain. Low-certainty evidence suggests that 1 to 2 units of RBC transfusion may increase the risk of severe maternal morbidity; however, we urge caution when interpreting this finding as the effect estimates are at serious risk of bias due to possible confounding. We are unable to comment on the effects of larger blood transfusion amounts on severe maternal morbidity.

Funding: This review received no dedicated funding.

Registration: This protocol for this Cochrane review is registered with PROSPERO. Available from: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42024599608.