Comparing different venous thromboembolism risk assessment machine learning models in Chinese patients.

OBJECTIVE: Venous thromboembolism (VTE) is a fatal complication and the most common preventable cause of death in hospitals. The risk-to-benefit ratio of thromboprophylaxis depends on the performance of the risk assessment model. A linear model, the Padua model, is recommended for medical inpatients in the United States but is not suitable for Chinese inpatients due to differences in race and disease spectrum. Currently, machine learning (ML) methods show advantages in modeling complex data patterns and have been applied to clinical data analysis. This study aimed to build VTE risk assessment ML models among Chinese inpatients and compare the predictive validity of the ML models with that of the Padua model. METHODS: We used 376 patients, including 188 patients with VTE, to build a model and then evaluate the predictive validity of the model in a consecutive clinical dataset from Peking Union Medical College Hospital. Nine widely used ML methods were trained on the model derivation set and then compared with the Padua model. RESULTS: Among the nine ML methods, random forest (RF), boosting-based methods, and logistic regression achieved a higher specificity, Youden index, positive predictive value, and area under the receiver operating characteristic curve than the Padua model on both the test and clinical validation sets. However, their sensitivities were inferior to that of the Padua model. Combined with the receiver operating characteristic curve, RF, as the best performing model, maintained high specificity with relatively better sensitivity and captured VTE patients' patterns more precisely. CONCLUSIONS: Advances in ML technology provide powerful tools for medical data analysis, and choosing models conforming to the disease pattern would achieve good performance. Popular ML models do not surpass the Padua model on all indicators of validity, and the drawback of low sensitivity should be improved upon in the future.

Glycated albumin may be a choice, but not an alternative marker of glycated hemoglobin for glycemic control assessment in diabetic patients undergoing maintenance hemodialysis.

BACKGROUND: It has been suggested that glycated hemoglobin (HbA1c) underestimate the actual glycemic control levels in maintenance hemodialysis (MHD) patients, because of anemia and the using of erythropoietin (EPO); it was recommended that glycated albumin (GA) should be an alternative marker. Therefore, the assessment performances of glycemic control were compared between GA and HbA1c in this research by referring to mean plasma glucose (MPG) in diabetes mellitus (DM) patients undergoing MHD or not. METHODS: MPG was calculated according to the data registered at enrollment and follow-up 2 months later and corresponding HbA1c, albumin (ALB), GA, etc. were measured in 280 cases. A case-control study for comparing GA and HbA1c was done among the groups of MHD patients with DM (n=88) and without DM (NDM; n=90), and non-MHD ones with DM (n=102) using MPG for an actual glycemic control standard. RESULTS: In these 3 groups, only for DM patients' (whether undergoing MHD or not), GA and HbA1c correlated with MPG significantly (P < 0.01). Through linear regression analysis, it could be found that the regression curves of GA almost coincided in MHD and non-MHD patients with DM, because the intercepts (2.418 vs. 2.329) and slopes (0.053 vs. 0.057) were very close to each other. On the contrary, regression curves of HbA1c did not coincide in the two groups, because variance of the slopes (0.036 vs. 0.052) were relatively large. Through comparing receiver operating characteristic (ROC) areas under the curve (AUC), it could be understood that the assessment performances of GA and HbA1c in MHD patients were lower than those in non-MHD ones, and assessment performance of HbA1c in MHD patients was better than GA (P < 0.05). In addition, the effects of Hb and EPO dose on HbA1c, or that of ALB on GA were unobvious in our study. CONCLUSIONS: Actual glycemic control level in MHD patients with DM may be underestimated by HbA1c, and it could be avoided by GA; however, glycemic evaluating performance of HbA1c may be still better than that of GA. Therefore, HbA1c should not be replaced completely although GA can be used as a choice to monitor glycemic level.

[Quantitative assessment of healthy People's cardiovascular aging and analysis of the relevant law: study in three cities in Northern China].

OBJECTIVE: To quantitatively assess the progress of cardiovascular aging in the healthy people in North China by non-invasive ways, and to analyze the law of cardiovascular aging. METHODS: 855 healthy people randomly sampled from 3 cities in northern China: Shenyang, Dalian, and Beijing, divided into 4 age groups: < or = 44, 45 - 59, 60 - 74, and > or = 75, underwent interview of the life style and dietary structure, measurement of blood pressure, heart and carotid ultrasonography, blood biochemistry, blood routine, urine routine, and detection of micro-inflammation indexes such as C-reactive protein, fibrinogen (FIB), thrombomodulin, and interleukin-6. Totally 70 items were surveyed. Then the cardiovascular aging assessment equation was constructed and cardiovascular biological aging score (CBAS) was calculated to analyze the law of cardiovascular aging. RESULTS: Seven items including pulse pressure, carotid artery intimal-medial thickness, carotid artery end-diastolic velocity, mitral valve anulus anterior wall A, mitral valve anulus lateral wall E, FIB, total cholesterol were correlated with chronological age significantly (r = 0.63, 0.54, -0.41, 0.45, -0.42, 0.43, and 0.46, all P < 0.01). The CBAS was positively correlated with chronological age (r = 0.85, P < 0.05) with the most rapid change rate in the group aged 45 - 59. CONCLUSION: CAS can be used as an objective index to quantitatively assess the cardiovascular aging. With an acceleration of cardiovascular aging rate after the age 45, the speed of cardiovascular aging in healthy people may be inconstant. The age 45 - 59 may be the key threshold for cardiovascular aging.