(931) Preoperative predictors and risk stratification of blood transfusion at time of cesarean delivery

To evaluate preoperative factors and develop a risk stratification tool to predict the need for blood transfusion (BT) at time of cesarean delivery (CD.)

Study Design:

This was a retrospective cohort study of 3,268 CDs at one university medical center between 2018-2022. Cases were excluded for placenta accreta spectrum (n=146), nonviable fetus (n=43), or delivery outside of L&D (n=46). Variables significant in the univariate analysis were analyzed with multiple logistic regression. Regression coefficients, adjusted odds ratios, and 95% confidence intervals were computed for each variable. A three-part scoring system was derived by assigning points based on the magnitude of the regression coefficients.

Results:
There were 114 BTs out of 3,033 CDs (3.8%.) In the multiple logistic regression model, factors significantly associated with BT included placenta previa (aOR 5.4 [95%CI 2.2-13.2]), history of BT (aOR 4.2 [95%CI 2.6-6.7]), preoperative hemoglobin < 11 (aOR 3.1 [95%CI 2.0-4.7]), suspected intraamniotic infection (aOR 2.4 [95%CI 1.2-5.1]), dilation  ≥ 6cm (aOR 2.4 [95%CI 1.5-3.9]), fibroids (aOR 2.2 [95% CI 1.3-3.7]), vaginal bleeding (aOR 2 [95%CI 1.1-3.8]), multifetal gestation (aOR 1.8 [95%CI 1-3.9.]) ROC curve produced an AUC of 0.74, p< 0.0001. A risk score created from weighted regression coefficients was categorized into risk groups: low risk (< 5% risk, 0-2 points), medium risk (5-9.99% risk, 3 points), and high risk ( >10% risk, ≥4 points.) ROC curve of the risk score produced an AUC of 0.74, p< 0.0001, sensitivity 56%, specificity 79%. The actual rate and predicted probabilities of transfusions in the low risk group were 2.1% and 2.1%, in the medium risk group 5.4% and 5.8%, and in the high risk group 15% and 15.1%.

Conclusion:

Multiple preoperative risk factors are associated with intraoperative BT. This tool can be used to reduce unnecessary preparation of crossmatched blood and improve blood utilization during CD. Recommend crossmatched blood be available in the operating room for high or medium risk patients. A type and screen may be permissible for patients categorized as low risk.