Prediction tools for various intraoperative bleeding events remain scarce. We aim to develop machine learning-based models and identify the most important predictors by real-world data from electronic medical records (EMRs).
An established database of surgical inpatients in Shanghai was utilized for analysis. A total of 51,173 inpatients were assessed for eligibility. 48,543 inpatients were obtained in the dataset and patients were divided into haemorrhage (N = 9728) and without-haemorrhage (N = 38,815) groups according to their bleeding during the procedure. Candidate predictors were selected from 27 variables, including sex (N = 48,543), age (N = 48,543), BMI (N = 48,543), renal disease (N = 26), heart disease (N = 1309), hypertension (N = 9579), diabetes (N = 4165), coagulopathy (N = 47), and other features. The models were constructed by 7 machine learning algorithms, i.e., light gradient boosting (LGB), extreme gradient boosting (XGB), cathepsin B (CatB), Ada-boosting of decision tree (AdaB), logistic regression (LR), long short-term memory (LSTM), and multilayer perception (MLP). An area under the receiver operating characteristic curve (AUC) was used to evaluate the model performance.
The mean age of the inpatients was 53 ± 17 years, and 57.5% were male. LGB showed the best predictive performance for intraoperative bleeding combining multiple indicators (AUC = 0.933, sensitivity = 0.87, specificity = 0.85, accuracy = 0.87) compared with XGB, CatB, AdaB, LR, MLP and LSTM. The three most important predictors identified by LGB were operative time, D-dimer (DD), and age.
We proposed LGB as the best Gradient Boosting Decision Tree (GBDT) algorithm for the evaluation of intraoperative bleeding. It is considered a simple and useful tool for predicting intraoperative bleeding in clinical settings. Operative time, DD, and age should receive attention.

Background: Cesarean scar pregnancy (CSP) involves a rare form of placental attachment that often leads to life-threatening conditions. The best treatment for CSP has been debated for decades. We aimed to evaluate the different treatments for CSP and analyzed the risk factors for intraoperative hemorrhage. Methods: CSP patients treated at the Affiliated Hospital of Jiangnan University were reviewed retrospectively from January 2014 to 2020. CSP was classified into three types based on the location and shape of gestational tissue, blood flow features, and thickness of the myometrium at the incision site. The clinical characteristics, types, approaches of treatment, and clinical outcomes of CSP were analyzed. Results: A total of 55 patients were included in this study, 29 (52.7%) of whom underwent transvaginal curettage after uterine artery embolization (UAE) and 22 (40%) of whom underwent transabdominal ultrasound-guided hysteroscopic curettage (USHC) in type I and II. Four patients (7.3%) classified as type III underwent laparoscopic cesarean scar resection (LCSR). Intraoperative blood loss, blood transfusion rate, and scar diverticulum were significantly higher in type II than in type I (P < 0.05). Even though USHC showed no differences in intraoperative blood loss, length of stay, and scar diverticulum compared with curettage after UAE (P > 0.05), superiority was found in surgical time and hospitalization cost (P < 0.05). Furthermore, the type of CSP (OR = 10.53, 95% CI: 1.69-65.57; P = 0.012) and diameter of the gestational sac (OR = 25.76, 95% CI: 2.67-248.20; P = 0.005) were found to be risk factors for intraoperative hemorrhage. Conclusions: Transabdominal ultrasound-guided hysteroscopic curettage is an effective and relatively safe treatment option for patients with CSP. Type of CSP and diameter of the gestational sac were found to be associated with excessive intraoperative hemorrhage.

BACKGROUND: One of the primary intraoperative challenges during burn surgery is to adequately excise the burn while avoiding massive hemorrhage. This has become increasingly important, as we see more burn patients that are older and with more medical comorbidities. While adequate excision down to healthy tissues for deep burns is essential for skin graft to take, it also leads to active bleeding that can be a challenge to control. Good hemostasis is imperative as a hematoma is the most common cause of graft loss. Several new products have become available to help control intraoperative hemorrhage. A new hemostatic dressing, NuStat®, is available and approved by FDA in United States.
METHODS: A single institution prospective randomized control trial was completed at Regional Burn Center of the University of South Alabama comparing NuStat® with the institutional historic standard of care. Twenty such patients were included in our study. A cost analysis was also completed as part of the study retrospectively.
RESULTS: For dressings used to treat the burn site, blood loss on the side treated with NuStat® was on average less (27g/100cm2) than the side treated with our historic standard of care (31g/100cm2), though it was not statistically significant (p=0.81). Similarly, on the donor site, blood loss on the side treated with NuStat® was on average less (14g/100cm2) than the side treated with our historic standard of care (15g/100cm2), but it was also not statistically significant (p=0.92). Average total blood loss from both excision and donor sites was also less with NuStat® (10g/100cm2) compared to the historic standard of care (12g/100cm2), but it was also not significant (p=0.77). There was no difference in the number of cycles required to achieve hemostasis for either the burn (1.15 NuStat® vs. 1.1 for historic standard of care, p=0.70) or the donor site (1 vs. 1, p=1.0). When comparing the cost of NuStat® versus the historic standard of care, the actual costs incurred for the wounds was less for the portion treated with NuStat® ($148.43) when compared to the historic standard of care ($186.45) (p<0.001).
CONCLUSIONS: Based on these findings, NuStat® hemostatic action should be comparable to the historic standard of care, and these newer hemostatic agents evaluated further in burn surgery and bleeding during other procedures such as trauma surgery.