Mathematical Model to Predict Metabolic Acidosis at Birth in Pregnancies with Absent or Reversed End-Diastolic Velocity - Revista Brasileira de Ginecologia e Obstetrícia

Mathematical Model to Predict Metabolic Acidosis at Birth in Pregnancies with Absent or Reversed End-Diastolic Velocity |

Purpose: to develop a mathematical model based on the fetal heart rate (FHR) monitoring to predict metabolic acidosis at birth (base excess < -10 mEq/L), in pregnancies with absent or reversed end-diastolic velocity (AREDV) in the umbilical arteries. Methods: the last FHR tracing of 127 AREDV cases was studied by visual analysis. The analyzed parameters included: gestational age, interval between AREDV diagnosis and delivery, FHR variability, FHR accelerations, decelerations, and sinusoidal- like pattern. Multivariate logistic regression was applied to find the best mathematical model to predict acidosis. Results: metabolic acidosis at birth occurred in 51 cases (40.2%). The model included the parameters: interval between AREDV diagnosis and delivery (X1), gestational age in weeks (X2), FHR variability <5 bpm (X3), and FHR variability 5-9 bpm (X4). To each variability parameter was assigned the value of 1 when present or 0 when absent. The z value is: z = 2.2348 + (-0.0117 X1) + (-0.09 X2) + (1.9552 X3)+ (-0.4474 X4). By applying the expression p=e z/(1 + e z), the probability is estimated. Conclusion: the mathematical model allowed us to estimate the probability of metabolic acidosis at birth, in pregnancies with AREDV, studying FHR-monitoring parameters.