Yongkui Ren

Patients with acute heart failure (AHF) complicated by acute cardiorenal syndrome (CRS) demonstrate substantial mortality risk. Conventional blood urea nitrogen (BUN) assessments using single-point measurements inadequately capture the temporal dynamics of renal dysfunction and disease progression. This study investigated the prognostic utility of BUN trajectory patterns for predicting in-hospital and 28-day mortality in AHF patients with CRS. We conducted a retrospective cohort study using the Medical Information Mart for Intensive Care IV (MIMIC-IV) database as the derivation cohort and the eICU Collaborative Research Database (eICU-CRD) as the validation cohort. BUN measurements were obtained at baseline (< 24 h), 72 h, and 144 h across both databases. Longitudinal BUN trajectory patterns were identified using latent class growth modeling (LCGM). Multivariable logistic and Cox proportional hazards regression analyses evaluated the association between trajectory groups and in-hospital and 28-day all-cause mortality, respectively. Subgroup and interaction analyses assessed the robustness of trajectory-based predictions across patient characteristics. Among 2931 patients across both cohorts, LCGM identified three distinct BUN trajectory patterns: persistent-low, persistent-moderate, and persistent-high trajectory groups. Restricted cubic spline analysis revealed a linear relationship between baseline BUN levels and both mortality outcomes (P > 0.05 for non-linearity), though with inconsistent predictive performance. Compared with the persistent-high trajectory group, the persistent-low trajectory group demonstrated significantly reduced in-hospital mortality (derivation cohort: OR 0.47, 95% CI 0.27–0.85; validation cohort: OR 0.59, 95% CI 0.38–0.93) and 28-day mortality (derivation cohort: HR 0.56, 95% CI 0.34–0.94; validation cohort: HR 0.66, 95% CI 0.45–0.85). Subgroup analyses indicated enhanced predictive performance of the persistent-low trajectory group in non-diabetic patients across both cohorts. No significant interactions were observed across subgroups, including different acute kidney injury stages. BUN trajectory patterns provide independent prognostic information for short-term outcomes in AHF patients with CRS. Persistent-low BUN trajectory are associated with significantly improved survival compared to persistent-high trajectory.

This study attempts to compare the predictive effects of several prediction models on obstructive coronary artery disease (OCAD) in young patients (30–50 years old), with a view to providing a new evaluation tool for the prediction of premature coronary artery disease (PCAD). A total of 532 hospitalized patients aged 30–50 were included in the study.All of them underwent coronary computed tomography angiography (CCTA) for suspected symptoms of coronary heart disease.Coronary artery calcium score (CACS) combined with traditional risk factors and pre-test probability models are the prediction models to be compared in this study.The PTP model was selected from the upgraded Diamond-Forrester model (UDFM) and the Duke clinical score (DCS). All patients included in the study were aged 30–50 years. Among them, women accounted for 24.4%, and 355 patients (66.7%) had a CACS of 0. OCAD was diagnosed in 43 patients (8.1%). The CACS combined with traditional risk factors to predict the OCAD area under the curve of receiver operating characteristic (ROC) (AUC = 0.794,p < 0.001) was greater than the PTP models (AUCUDFM=0.6977,p < 0.001;AUCDCS=0.6214,p < 0.001). By calculating the net reclassification index (NRI) and the integrated discrimination index (IDI), the ability to predict the risk of OCAD using the CACS combined with traditional risk factors was improved compared with the PTP models (NRI&IDI > 0,p < 0.05). The predictive value of CACS combined with traditional risk factors for OCAD in young patients is better than the PTP models.

Background: The addition of coronary artery calcium score (CACS) to prediction models has been verified to improve performance. Machine learning (ML) algorithms become important medical tools in an era of precision medicine, However, combined utility by CACS and ML algorithms in hypertensive patients to forecast obstructive coronary artery disease (CAD) on coronary computed tomography angiography (CCTA) is rare. Methods: This retrospective study was composed of 1,273 individuals with hypertension and without a history of CAD, who underwent dual-source computed tomography evaluation. We applied five ML algorithms, coupled with clinical factors, imaging parameters, and CACS to construct predictive models. Moreover, 80% individuals were randomly taken as a training set on which 5-fold cross-validation was done and the remaining 20% were regarded as a validation set. Results: 16.7% (212 out of 1,273) of hypertensive patients had obstructive CAD. Extreme Gradient Boosting (XGBoost) posted the biggest area under the receiver operator characteristic curve (AUC) of 0.83 in five ML algorithms. Continuous net reclassification improvement (NRI) was 0.55 (95% CI (0.39–0.71), p < 0.001), and integrated discrimination improvement (IDI) was 0.04 (95% CI (0.01–0. 07), p = 0.0048) when the XGBoost model was compared with traditional Models. In the subgroup analysis stratified by hypertension levels, XGBoost still had excellent performance. Conclusion: The ML model incorporating clinical features and CACS may accurately forecast the presence of obstructive CAD on CCTA among hypertensive patients. XGBoost is superior to other ML algorithms.