Category: Obstetric Quality and Safety
Poster Session I
Delayed transfusion for massive hemorrhage during cesarean delivery can result in disseminated intravascular coagulopathy and hypoperfusion of vital organs, which may result in adverse outcomes such as hysterectomy or maternal death. Several studies have reported risk factors of massive transfusion. However, most of previous studies have focused on preoperative variables, although hemodynamic changes are more likely to represent the patients’ hypovolemic status needing massive transfusion. Therefore, the objective of this study was to develop real-time prediction of massive transfusion using continuous intraoperative hemodynamic parameters in pregnant women.
Study Design:
For the deep learning-based prediction model development and internal performance test, we obtained a dataset of 32,715 vital signs of non-pregnant subjects. For evaluating the model performance in pregnant subjects, this study included vital signs of 550 pregnant women during cesarean section. Massive transfusion was defined as the transfusion of 3 or more units of red blood cells in one hour. Intraoperative hemodynamic parameters extracted from non-invasive blood pressure, plethysmography, and hematocrit measured during cesarean section were used for construction of real-time prediction model of massive transfusion 10 minutes in advance.
Results:
Among 32,715 non-pregnant subjects and 550 pregnant women, 0.6% of patients and 1.3% of pregnant women received massive transfusion during non-obstetric surgery and cesarean section, respectively. Using only pre-operative features, massive transfusion prediction achieved an area under the receiver operating characteristic curve (AUROC) of 0.737 (95% CI= 0.578-0.895) in pregnant women. The performance of real-time prediction model constructed using both intraoperative hemodynamic parameters and preoperative variables was significantly improved (AUROC 0.931, 95% CI= 0.875-0.987) in pregnant women.
Conclusion:
The inclusion of intraoperative hemodynamic parameters significantly improved real-time prediction for massive transfusion and can allow early intervention for maternal safety.
Do Yun Kwon, BA
Interdisciplinary program of medical informatics, Seoul National University College of Medicine, Seoul, Korea
Seoul, Seoul-t'ukpyolsi, Republic of Korea
Young Mi Jung, MD
Seoul National University College of Medicine
Seoul, Seoul-t'ukpyolsi, Republic of Korea
Seung-Bo Lee, PhD
Department of Medical Informatics, Keimyung University School of Medicine
Daegu, South Korea, Republic of Korea
Taekyoung Kim, MD, PhD
Department of Anesthesiology and Pain Medicine, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul, Korea
Seoul, Seoul-t'ukpyolsi, Republic of Korea
Kwangsoo Kim, PhD
Transdisciplinary Department of Medicine and Advanced Technology, Seoul National University Hospital, Seoul, Korea
Seoul, Seoul-t'ukpyolsi, Republic of Korea
garam Lee, PhD
Department of Biostatistics, Epidemiology & Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, USA
Philadelphia, Pennsylvania, United States
Jihye Bae, MD (she/her/hers)
Department of Obstetrics and Gynecology, Seoul National University College of Medicine
Seoul, Seoul-t'ukpyolsi, Republic of Korea
Jeesun Lee, MD
Department of Obstetrics and Gynecology, Seoul National University College of Medicine
Seoul, Seoul-t'ukpyolsi, Republic of Korea
Chan-Wook Park, MD, PhD
Department of Obstetrics and Gynecology, Seoul National University College of Medicine
Seoul, Seoul-t'ukpyolsi, Republic of Korea
Joong Shin Park, MD,PhD
Seoul National University College of Medicine
Seoul, Seoul-t'ukpyolsi, Korea, Republic of
Jong Kwan Jun, MD,PhD
Seoul National University College of Medicine
Seoul, Seoul-t'ukpyolsi, Korea, Republic of
Dokyoon Kim, PhD
University of Pennsylvania
Philadelphia, Pennsylvania, United States
Hyung-Chul Lee, MD, PhD
Department of Anesthesiology and Pain Medicine, Seoul National University College of Medicine, Seoul, Korea
Seoul, Seoul-t'ukpyolsi, Republic of Korea