Development Of A Prediction Model Of Cardiovascular Event Death Risk In Hemodialysis Patients With TCM Syndromes

ObjectiveCardiovascular diseases have become one of the main causes of death in hemodialysis patients.At present,many studies have been conducted on the risk factors of cardiovascular diseases in hemodialysis patients,but effective prediction models have not yet been established.The purpose of this study is to develop a prediction model for the risk of death from cardiovascular diseases in hemodialysis centers in our hospital,and to provide evidence support for clinical decision-making benefits and risk assessment.MethodsThe study used the medical record system and the hemodialysis registration system of our hospital to find patients who were diagnosed with chronic kidney disease stage 5 and started hemodialysis in our dialysis center from January 1,2009 to December 31,2018.According to the inclusion and exclusion criteria,we collected baseline data such as TCM syndrome types,Western medicine indicators and dialysis time points of patients who met the requirements before entering dialysis,The patients were followed up until December 31,2019 to collect the end events and the time of end events.75%of the patients were randomly selected as the training dataset.Cox univariate regression analysis was used to screen for risk factors related to cardiovascular diseases.Cox multivariate regression analysis was used to screen for independent risk factors for modeling.Finally,a visual nomogram was developed,and the remaining 25%of the patients in the center were internally verified.We calculated the C-index,drew the ROC curve and calibration plot to evaluate the model.Results1.A total of 546 patients were included in this study,which were randomly divided into 410 cases in the training dataset and 136 cases in the validation dataset.99 patients(24.1%)died of cardiovascular events in the training dataset and a median follow-up time of 41.0[41.0]months;29 patients(21.3%)died of cardiovascular diseases in the verification set and the median time of follow-up for 38.0[48.3]months.2*Cox univariate regression analysis revealed age,smoking history,history of diabetes,systolic blood pressure,diastolic blood pressure,history of cardiovascular diseases,white blood cells,platelets,urea nitrogen,serum creatinine,serum albumin,serum calcium,phosphorus,and hypersensitive troponin,B-type natriuretic peptide,creatine kinase isoenzyme,glycated hemoglobin,C-reactive protein,spleen-kidney qi deficiency syndrome,spleen-kidney yang deficiency syndrome,damp-heat syndrome,and phlegm turbidity syndrome were statistically significant.The hazard ratio of platelet was 1,which was not related to the outcome events.Urea nitrogen,serum creatinine,phosphorus,serum albumin,spleen-kidney qi deficiency syndrome and damp-heat syndrome were protective factors,and the remaining were risk factors(P<0.1).Cox multivariate regression analysis screened systolic blood pressure,blood calcium,and spleen-kidney yang deficiency syndrome as independent.risk factors for death of cardiovascular events(P<0.05).3.Combining Cox multivariate regression analysis,literature research,and clinical experience,finally,five factors including age,systolic blood pressure,serum calcium,C-reactive protein,and spleen-kidney yang deficiency syndrome were used to construct a predictive model of cardiovascular death in hemodialysis patients and plotted nomogram.The C-inde.x of the model is 0.748,and the AUCs of the first,second,third,and fourth years are 0.694,0.725,0.777 and 0.802.The internal verification using the validation dataset yields a C-index of 0.768 and the AUCs of the first to fourth years are.0.716,0.778,0.779 and 0.744.Calibration plot were drawn in the training dataset and validation dataset,respectively,and both fit well.ConclusionThe nomogram established in this study had been verified the internal verification,which shows that the model has a good level of discrimination and calibration.It can better predict the risk of death due to cardiovascular diseases in hemodialysis patients,assist clinicians to make decisions,adjust treatment measures early for hemodialysis patients with higher risk of death,and develop personalized treatment plan to reduce the risk of death due to cardiovascular diseases.