Prophylactic Carbetocin Versus Oxytocin and Blood Loss After Vaginal Delivery: Agent Selection and Confounding by Indication in a Retrospective Cohort Study.
Carbetocin was associated with higher blood loss before adjustment, but this difference was not significant after adjustment, indicating potential confounding factors.
Where it sits
this study against the rest of the carbetocin corpusSummary and findings
This observational study examined the association between carbetocin and blood loss after vaginal delivery in 559 women. Carbetocin was administered at a dose of 100 μg intravenously to 474 women, while 85 women received oxytocin at 10 IU by infusion. The study found that adjusted blood loss did not differ significantly between the two groups, although a modest increase with carbetocin could not be excluded.
Abstract
<i>Background and Objectives</i>: Carbetocin is increasingly used for postpartum hemorrhage (PPH) prophylaxis, but its selection in routine practice may bias comparisons with oxytocin. This observational study examined the association between the agent received and blood loss after vaginal delivery. <i>Materials and Methods</i>: This retrospective cohort included 559 women delivering vaginally in 2022-2023 who received carbetocin 100 μg intravenously (<i>n</i> = 474) or oxytocin 10 IU by infusion (<i>n</i> = 85). Gravimetric blood loss was analyzed using linear (primary), robust and median (co-primary) regression, and propensity score methods. <i>Results</i>: Carbetocin recipients more often underwent induction (63.3% vs. 43.5%), vacuum extraction (26.8% vs. 15.3%), and epidural analgesia (76.2% vs. 29.4%). Blood loss was higher with carbetocin before adjustment (223.7 vs. 180.0 mL, <i>p</i> = 0.003) but not after adjustment (β 12.72 mL, 95% CI -50.19 to 75.64, <i>p</i> = 0.691; robust <i>p</i> = 0.479; median <i>p</i> = 0.354). Propensity score results conflicted: inverse probability weighting showed higher blood loss with carbetocin (47.84 mL, <i>p</i> = 0.002), whereas matching and trimmed weighting did not. PPH occurred in 2.7% vs. 0% (<i>p</i> = 0.234). <i>Conclusions</i>: Carbetocin was preferentially used in higher-risk labors, indicating confounding by indication. Adjusted blood loss did not differ significantly, although a modest increase could not be excluded. The study was severely underpowered for PPH.
Background
This paper addresses the use of carbetocin for postpartum hemorrhage (PPH) prophylaxis and its comparison with oxytocin. Prior knowledge indicates that carbetocin is increasingly used, but the selection bias in routine practice may affect outcomes. Understanding the association between the agent received and blood loss is crucial for clinical decision-making.
Methods
This retrospective cohort study included 559 women who delivered vaginally in 2022-2023, with 474 receiving carbetocin and 85 receiving oxytocin. The primary outcome was gravimetric blood loss analyzed using linear regression, with co-primary outcomes assessed via robust and median regression, and propensity score methods.
Results
The primary endpoint showed that blood loss was higher with carbetocin before adjustment (223.7 vs. 180.0 mL, p=0.003). After adjustment, the difference was not statistically significant (β 12.72 mL, 95% CI -50.19 to 75.64, p=0.691). Propensity score analysis revealed conflicting results regarding blood loss.
Interpretation
The findings suggest that while carbetocin recipients experienced higher blood loss before adjustment, this difference was not clinically significant after adjustment. The study's limitations include confounding by indication and a small sample size for PPH outcomes, which may affect the reliability of the conclusions drawn.
Key findings
- Blood loss was higher with carbetocin before adjustment (223.7 vs. 180.0 mL, p=0.003).
- After adjustment, blood loss did not differ significantly (β 12.72 mL, 95% CI -50.19 to 75.64, p=0.691).
- Inverse probability weighting showed higher blood loss with carbetocin (47.84 mL, p=0.002), while matching and trimmed weighting did not.
- PPH occurred in 2.7% vs. 0% (p=0.234).
Limitations
- Severely underpowered for PPH outcomes.
- Confounding by indication due to preferential use in higher-risk labors.
- Small sample size for oxytocin group (n=85).
- Retrospective design may introduce bias.