Correlation Analysis of Postpartum Hemorrhage with Labor Analgesia Based on Machine Learning

Labor analgesia, with a history of over 100 years, has been a significant aid in alleviating the pain experienced by parturients during labor and facilitating smoother deliveries. However, its potential adverse effects on both parturients and infants have been the subject of ongoing debate. This paper aims to investigate the correlation between the physiological indicators of parturients and the volume of bleeding following painless delivery. Mathematical and machine learning models were applied to predict early warning signs of postpartum blood loss. The data used for analysis was gathered from the Shenzhen Maternal & Child Healthcare Hospital. The study utilized the random forest machine learning algorithm to analyze the relevant factors contributing to postpartum hemorrhage in labor analgesia. The results of the comprehensive machine learning analysis identified intrapartum hemorrhage as the most significant factor influencing the volume of postpartum bleeding in parturients, followed by height, gravida, parity, oxytocin, and premature rupture of membranes. The random forest model utilizing the six most relevant factors achieved an AUC score of 0.874. These findings offer valuable insights that may guide clinical medical staff in the assessment and management of postpartum hemorrhage.