RESUMO
Background: Although evidence-based medicine proposes personalized care that considers the best evidence, it still fails to address personal treatment in many real clinical scenarios where the complexity of the situation makes none of the available evidence applicable. "Medicine-based evidence" (MBE), in which big data and machine learning techniques are embraced to derive treatment responses from appropriately matched patients in real-world clinical practice, was proposed. However, many challenges remain in translating this conceptual framework into practice. Objective: This study aimed to technically translate the MBE conceptual framework into practice and evaluate its performance in providing general decision support services for outcomes after congenital heart disease (CHD) surgery. Methods: Data from 4774 CHD surgeries were collected. A total of 66 indicators and all diagnoses were extracted from each echocardiographic report using natural language processing technology. Combined with some basic clinical and surgical information, the distances between each patient were measured by a series of calculation formulas. Inspired by structure-mapping theory, the fusion of distances between different dimensions can be modulated by clinical experts. In addition to supporting direct analogical reasoning, a machine learning model can be constructed based on similar patients to provide personalized prediction. A user-operable patient similarity network (PSN) of CHD called CHDmap was proposed and developed to provide general decision support services based on the MBE approach. Results: Using 256 CHD cases, CHDmap was evaluated on 2 different types of postoperative prognostic prediction tasks: a binary classification task to predict postoperative complications and a multiple classification task to predict mechanical ventilation duration. A simple poll of the k-most similar patients provided by the PSN can achieve better prediction results than the average performance of 3 clinicians. Constructing logistic regression models for prediction using similar patients obtained from the PSN can further improve the performance of the 2 tasks (best area under the receiver operating characteristic curve=0.810 and 0.926, respectively). With the support of CHDmap, clinicians substantially improved their predictive capabilities. Conclusions: Without individual optimization, CHDmap demonstrates competitive performance compared to clinical experts. In addition, CHDmap has the advantage of enabling clinicians to use their superior cognitive abilities in conjunction with it to make decisions that are sometimes even superior to those made using artificial intelligence models. The MBE approach can be embraced in clinical practice, and its full potential can be realized.
RESUMO
OBJECTIVES: To investigate the risk factors of postoperative neuro-developmental abnormalities in neonates with critical congenital heart disease (CCHD). METHODS: Clinical data of 50 neonates with CCHD admitted in the Cardiac Intensive Care Unit, The Children's Hospital, Zhejiang University School of Medicine from November 2020 to December 2021 were retrospectively analyzed. Neurological assessment was performed with cranial ultrasonography, CT/MRI, video electroencephalogram and clinical symptoms before and after surgical treatment for all patients, and neurodevelopmental abnormalities were documented. Binary logistic stepwise regression was used to analyze risk factors of postoperative new-onset neurodysplasia in children with CCHD, and the predictive value of the risk factors on postoperative neurodevelopmental abnormalities were evaluated using the receiver operating characteristic (ROC) curve. RESULTS: Neurodevelopmental abnormalities were detected in 22 cases (44.0%) and not detected in 28 cases (56.0%) before surgery. There were no significant differences in gender, birth weight, age at admission, gestational age, preoperative SpO2 level, prematurity, cyanotic congenital heart disease, and ventilator support between the two groups (all P>0.05). After surgery, there were 22 cases (44.0%) with new-onset neurological abnormalities and 28 cases (56.0%) without new-onset abnormalities. Multivariate logistic regression analysis showed that postoperative 24 h peak lactic acid (OR=1.537, 95%CI: 1.170-2.018, P<0.01) and postoperative length of ICU stay (OR=1.172, 95%CI:1.031-1.333, P<0.05) were independent risk factors for postoperative new-onset neurodevelopmental abnormalities. The area under ROC curve (AUC) of the postoperative 24 h peak lactic acid for predicting the new-onset neurological abnormalities after operation was 0.829, with cut-off value of 4.95 mmol/L. The diagnostic sensitivity and specificity were 90.0% and 64.3%, respectively. The AUC of postoperative length of ICU stay for predicting the new-onset neurological abnormalities after operation was 0.712, with cut-off value of 18.0 d. The diagnostic sensitivity and specificity were 50.0% and 96.4%, respectively. The AUC of the combination of the two indicators was 0.917, the diagnostic sensitivity and specificity were 95.5% and 64.3%, respectively. CONCLUSIONS: The incidence of neurodysplasia in neonatal CCHD is high, and new neurological abnormalities may occur after surgery. The postoperative 24 h peak lactic acid and postoperative length of ICU stay are risk factors for new-onset neurodysplasia after surgery. The combination of the two indicators has good predictive value for neurodevelopmental outcomes after surgery in CCHD infants.