李倩
中国医学科学院阜外医院 麻醉中心
OBJECTIVE:Several machine learning (ML) models have been used in perioperative red blood cell (RBC) transfusion risk for cardiac surgery with limited generalizability and no external validation. Hence, we sought to develop and comprehensively externally validate a ML model in a large dataset to estimate RBC transfusion in cardiac surgery with cardiopulmonary bypass (CPB).DESIGN:A retrospective analysis of a multicenter clinical trial (NCT03782350).PATIENTS:The study patients who underwent cardiac surgery with CPB came from four cardiac centers in China and Medical Information Mart for Intensive Cared (MIMIC-IV) dataset.MEASUREMENTS:Data from Fuwai Hospital were used to develop an individualized prediction model for RBC transfusion. The model was externally validated in the data from three other centers and MIMIC-IV dataset. Twelve models were constructed.MAIN RESULTS:A total of 11,201 eligible patients were included in the model development (2420 in Fuwai Hospital) and external validation (563 in the other three centers and 8218 in the MIMIC-IV dataset). A significant difference was observed between the Logistic Regression and CatboostClassifier (0.72 Vs. 0.74, P = 0.031) or RandomForestClassifier (0.72 Vs. 0.75 p = 0.012) in the external validation and MIMIV-IV datasets (age ≤ 70:0.63 Vs. 0.71, p < 0.001; age > 70:0.63 Vs. 0.70, 0.63 Vs. 0.71, p < 0.001). The CatboostClassifier and RandomForestClassifier model was comparable in development (0.83 Vs. 0.82, p = 0.419), external (0.74 Vs. 0.75, p = 0.268), and MIMIC-IV datasets (age ≤ 70: 0.71 Vs. 0.71, p = 0.574; age > 70: 0.70 Vs. 0.71, p = 0.981). Of note, they outperformed other ML models with excellent discrimination and calibration. The CatboostClassifier and RandomForestClassifier models achieved higher area under precision-recall curve and lower brier loss score in validation and MIMIC-IV datasets. Additionally, we confirmed that low preoperative hemoglobin, low body mass index, old age, and female sex increased the risk of RBC transfusion.CONCLUSIONS:In our study, enrolling a broad range of cardiovascular surgeries with CPB and utilizing a restrictive RBC transfusion strategy, robustly validates the generalizability of ML algorithms for predicting RBC transfusion risk. Notably, the CatboostClassifier and RandomForestClassifier exhibit strong external clinical applicability, underscoring their potential for widespread adoption. This study provides compelling evidence supporting the efficacy and practical value of ML-based approaches in enhancing transfusion risk prediction in clinical practice.
International journal of medical informatics 2024
Background:The potential myocardial protective effect of nicorandil (NICD) in patients undergoing percutaneous coronary intervention has been established. However, its efficacy in the context of cardiac surgery remains controversial. The present study aimed to evaluate the myocardial protective effect of perioperative NICD use in patients undergoing cardiac surgery with cardiopulmonary bypass (CPB).Methods:We retrospectively gathered data from patients undergoing cardiac bypass surgery between 12/2018 and 04/2021 in Fuwai Hospital. Subsequently, the patients were divided into two groups, NICD group and non-nicorandil (non-NICD) group. A 1, 3 propensity score matching (PSM) was conducted. The primary outcome was the incidence of myocardial injury. The secondary outcomes included the mechanical ventilation (MV) duration, intensive care unit (ICU) length of stay (LOS), hospital LOS, duration of chest drainage, the drainage volume, the total cost, the incidence of acute kidney injury (AKI), and the incidence of acute liver injury (ALI). Subsequently, we divided the entire population into two distinct subgroups based on their administration of NICD, and performed a comprehensive subgroup analysis.Results:A total of 2406 patients were ultimately included in the study. After PSM, 250 patients in NICD group and 750 patients in non-NICD group were included in the analysis. Perioperative NICD reduced the incidence of myocardial injury (47.2% versus 38.8%, P=0.025). Our subgroup analysis revealed that preoperative NICD administration not only provided myocardial protection benefits (45.7% vs 35.8%, OR 0.66, 95% CI [0.45-0.97], P=0.041), but also demonstrated statistically significant reduction in ALI, the ICU and hospital LOS, and the duration of chest drainage (all P<0.05).Conclusion:The perioperative NICD administration may confer myocardial protection in patients undergoing cardiac surgery with CPB. Furthermore, the preoperative utilization of NICD has the potential to mitigate the incidence of postoperative ALI, a reduction in the ICU and hospital LOS, and the duration of chest drainage.
Drug design, development and therapy 2024
INTRODUCTION:The present study aimed to explore the potential effect of ulinastatin on renal function and long-term survival in patients receiving cardiac surgery with cardiopulmonary bypass (CPB).METHODS:This prospective cohort study was conducted at Fuwai Hospital, Beijing, China. Ulinastatin was applied after induction anesthesia. The primary outcome was the rate of new-onset postoperative acute kidney injury (AKI). Moreover, a 10-year follow-up was conducted until January 2021.RESULTS:The rate of new-onset AKI was significantly lower in the ulinastatin group than in the control group (20.00 vs. 32.40%, p = 0.009). There was no significant difference in renal replacement therapy between the two groups (0.00 vs. 2.16%, p = 0.09). The postoperative plasma neutrophil gelatinase-associated lipocalin (pNGAL) and IL-6 levels were significantly lower in the ulinastatin group compared with the control group (pNGAL: p = 0.007; IL-6: p = 0.001). A significantly lower incidence of respiratory failure in the ulinastatin group compared with the control group (0.76 vs. 5.40%, p = 0.02). The nearly 10-year follow-up (median: 9.37, 95% confidence interval: 9.17-9.57) survival rates did not differ significantly between the two groups (p = 0.076).CONCLUSIONS:Ulinastatin significantly reduced postoperative AKI and respiratory failure in patients receiving cardiac surgery with CPB. However, ulinastatin did not reduce intensive care unit and hospital stays, mortality, and long-term survival rate.
Cardiorenal medicine 2023
OBJECTIVE:We aimed to develop and validate a predictive machine learning (ML) model for cardiac surgery associated with acute kidney injury (CSA-AKI) based on a multicenter randomized control trial (RCT) and a Medical Information Mart for Intensive Care-IV (MIMIC-IV) dataset.METHODS:This was a subanalysis from a completed RCT approved by the Ethics Committee of Fuwai Hospital in Beijing, China (NCT03782350). Data from Fuwai Hospital were randomly assigned, with 80% for the training dataset and 20% for the testing dataset. The data from three other centers were used for the external validation dataset. Furthermore, the MIMIC-IV dataset was also utilized to validate the performance of the predictive model. The area under the receiver operating characteristic curve (ROC-AUC), the precision-recall curve (PR-AUC), and the calibration brier score were applied to evaluate the performance of the traditional logistic regression (LR) and eleven ML algorithms. Additionally, the Shapley Additive Explanations (SHAP) interpreter was used to explain the potential risk factors for CSA-AKI.RESULT:A total of 6495 eligible patients undergoing cardiopulmonary bypass (CPB) were eventually included in this study, 2416 of whom were from Fuwai Hospital (Beijing), for model development, 562 from three other cardiac centers in China, and 3517 from the MIMICIV dataset, were used, respectively, for external validation. The CatBoostClassifier algorithms outperformed other models, with excellent discrimination and calibration performance for the development, as well as the MIMIC-IV, datasets. In addition, the CatBoostClassifier achieved ROC-AUCs of 0.85, 0.67, and 0.77 and brier scores of 0.14, 0.19, and 0.16 in the testing, external, and MIMIC-IV datasets, respectively. Moreover, the utmost important risk factor, the N-terminal brain sodium peptide (NT-proBNP), was confirmed by the LASSO method in the feature section process. Notably, the SHAP explainer identified that the preoperative blood urea nitrogen level, prothrombin time, serum creatinine level, total bilirubin level, and age were positively correlated with CSA-AKI; preoperative platelets level, systolic and diastolic blood pressure, albumin level, and body weight were negatively associated with CSA-AKI.CONCLUSIONS:The CatBoostClassifier algorithms outperformed other ML models in the discrimination and calibration of CSA-AKI prediction cardiac surgery with CPB, based on a multicenter RCT and MIMIC-IV dataset. Moreover, the preoperative NT-proBNP level was confirmed to be strongly related to CSA-AKI.
Journal of clinical medicine 2023
Purpose:To estimate the overall incidence, risk factors, and clinical outcomes of electroencephalographic (EEG) seizures for adults and children after mechanical circulatory support (MCS).Method and measurements:This systematic review and meta-analysis were carried out in accordance with the PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analysis) guidance document. MEDLINE EMBASE and CENTRAL were investigated for relevant studies. The related information was retrieved by two independent reviewers and all analyses were conducted by STATA (version 16.0; Stata Corporation, College Station, TX, United States).Result:Sixty studies including 36,191 adult and 55,475 pediatric patients with MCS were enrolled for evaluation. The study showed that the overall incidence of EEG seizures in adults was 2% (95%CI: 1-3%), in which 1% (95%CI: 1-2%) after cardiopulmonary bypass (CPB), and 3% (95%CI: 1-6%) after extracorporeal membrane oxygenation (ECMO). For pediatrics patients, the incidence of EEG seizures was 12% (95%CI: 11-14%), among which 12% (9-15%) after CPB and 13% (11-15%) after ECMO. The major risk factors of EEG seizures after MCS in adults were redo surgery (coefficient = 0.0436, p = 0.044), and COPD (coefficient = 0.0749, p = 0.069). In addition, the gestational week of CPB (coefficient = 0.0544, p = 0.080) and respiratory failure of ECMO (coefficient = -0.262, p = 0.019) were also indicated to be associated with EEG seizures in pediatrics.Conclusion:EEG seizures after MCS were more common in pediatrics than in adults. In addition, the incidence of EEG seizure after ECMO was higher than CPB both in adults and children. It is expected that appropriate measures should be taken to control modifiable risk factors, thus improving the prognosis and increasing the long-term survival rate of MCS patients.Systematic Review Registration:[https://www.crd.york.ac.uk/prospero], identifier [CRD42021287288].
Frontiers in cardiovascular medicine 2022
