Purpose Acute kidney injury (AKI) is a frequent complication of severe acute pancreatitis (AP) and carries a very poor prognosis. The present study aimed to construct a model capable of accurately identifying those patients at high risk of harboring occult acute kidney injury (AKI) characteristics. Patients and Methods We retrospectively recruited a total of 424 consecutive patients at the Gezhouba central hospital of Sinopharm and Xianning central hospital between January 1, 2016, and October 30, 2021. ML-assisted models were developed from candidate clinical features using two-step estimation methods. The receiver operating characteristic curve (ROC), decision curve analysis (DCA), and clinical impact curve (CIC) were performed to evaluate the robustness and clinical practicability of each model. Results Finally, a total of 30 candidate variables were included, and the AKI prediction model was established by an ML-based algorithm. The areas under the ROC curve (AUCs) of the random forest classifier (RFC) model, support vector machine (SVM), eXtreme gradient boosting (XGBoost), artificial neural network (ANN), and decision tree (DT) ranged from 0.725 (95% CI 0.223–1.227) to 0.902 (95% CI 0.400–1.403). Among them, RFC obtained the optimal prediction efficiency via adding inflammatory factors, which are serum creatinine (Scr), C-reactive protein (CRP), platelet-to-lymphocyte ratio (PLR), neutrophil-to-lymphocyte ratio (NLR), neutrophil-to-albumin ratio (NAR), and CysC, respectively. Conclusion We successfully developed ML-based prediction models for AKI, particularly the RFC, which can improve the prediction of AKI in patients with AP. The practicality of prediction and early detection may be greatly beneficial to risk stratification and management decisions.

Background: The N-terminal pro B type natriuretic peptide (NT-proBNP) is important for prognosis of heart failure in patients with chronic kidney disease (CKD). However, the NT-proBNP level is easily affected by renal insufficiency, which limits its clinical use.Methods: This study included 396 patients with CKD. Plasma levels of NT-proBNP and cystatin C (CysC) were measured during hospitalization. The echocardiographic parameters were also detected. Patients were divided into the heart failure group and control group according to the European Society of Cardiology Guideline on Chronic Heart Failure 2021. Multiple modeling analysis of the values of NT-proBNP and CysC, including NT-proBNP/Cyscn and NT-proBNP/nCysC was performed. The receiver operating characteristic (ROC) curve, combined with the cardiac function, was used to determine the formula with the best diagnostic efficiency. Then, the sensitivity and specificity of new predictors for cardiac insufficiency in CKD patients were calculated. Pearson correlation analysis was used to analyze the relationship between new predictors and the NT-proBNP level. The clinical data of CKD patients from another local hospital were used to validate the new predictors and the cut-off values.Results: An elevated NT-proBNP/CysC1.53 ratio was an independent risk factor for cardiac dysfunction in CKD and the best predictor derived from multiple modeling analysis. There was no correlation between the NT-proBNP/CysC1.53 ratio and the NT-proBNP level (r = 0.376, p = 6.909). The area under the ROC curve for the NT-proBNP/CysC1.53 ratio was 0.815 (95% confidence interval: 0.772–0.858), and for a cut-off point of 847.964, this ratio had a sensitivity of 78.24%, and a specificity of 69.44%. When applied to the data of CKD patients from another local hospital, the NT-proBNP to CysC1.53 ratio had a sensitivity of 70.27% and a specificity of 67.74%.Conclusion: The NT-proBNP to CysC1.53 ratio was superior to NT-proBNP alone for predicting cardiac dysfunction in patients with CKD.

AimsBoth the Clauss and prothrombin time-derived fibrinogen (PT-Fib) methods have widely been used in fibrinogen (Fib) evaluation. We aimed to improve the interchangeability between these two methods.MethodsThirty fresh plasmas of low, normal and high Fib concentration were mixed to prepare the pooled human plasma. The Fib concentration was assessed by the Clauss method. The PT-Fib values were further recalibrated by the assigned plasma pools and the precision, linearity and reference intervals were verified according to the guidelines of American Society for Clinical and Laboratory Standards (CLSI) EP15-A and EP6-A documents. Finally, the recalibrated PT-Fib method was compared with Clauss method by the simultaneous Fib test of total 5259 coagulation samples.ResultsThe results indicated that the recalibrated PT-Fib method can detect the Fib concentration with clinically acceptable third-order linearity in the range of 1.27–8.00 g/L. Only one result out of 39 healthy people tested using the recalibrated PT-Fib method did not fall within the reference range defined by the manufacturer. We also found more than 99% of results between these two methods were interchangeable in the range of 1.51–8.00 g/L. The disagreement between these two methods was found only in patients with certain underlying conditions.ConclusionsAfter recalibration, the consistency between the PT-Fib and Clauss methods was enhanced and the interchangeability was improved. The application of the recalibrated PT-Fib method provided accurate and reliable results with the manufacturer-provided reagents and improved detection speed and cost-effectiveness.