Establishment of a nomogram for personalized prediction model of heart valve calcification in elderly patients with chronic renal failure on hemodialysis

Objective To establish a nomogram of effectively predicting the occurrence of cardiac valve calcification(CVC) in elderly patients with chronic renal failure(CRF) on hemodialysis.Methods Retrospective analysis was performed for clinical data of elderly patients undergoing hemodialysis from December 2015 to December 2018,using single factor and Logistic regression for analyzing the independent risk factors of CVC and establishing the correlation Nomogram prediction model.Results Dialysis age ≥ 24 months(OR=7.488,95%CI 2.524~22.215),diabetes mellitus(OR=6.256,95%CI 2.624~14.919),pAlb <200 mg/L(OR=5.623,95%CI 1.965~16.091),calcium & phosphorus product ≥ 55 mg²/dL²(OR=3.954,95%CI 1.679~9.311),iPTH ≥ 100 ng/L(OR=6.883,95%CI 2.478~19.118) and C-reactive protein(CRP) ≥ 5 mg/L(OR=3.422,95%CI 1.424~8.221) were independent risk factors for CVC in elderly CRF hemodialysis patients(P<0.05).Based upon the above six independent risk factors,a nomogram model indicated that the predicted value was basically consistent with the actual value and C-index index was 0.878(95%CI 0.852~0.904),indicating that this nomogram model had an excellent predictive capability.Conclusions Dialysis age ≥ 24 months,diabetes mellitus,pAlb <200 mg/L,calcium & phosphorus product ≥ 55 mg²/dL²,iPTH ≥ 100 ng/L and CRP ≥ 5 mg/L are independent risk factors for CVC in elderly CRF hemodialysis patients.The nomogram model can effectively predict the probability of CVC in elderly CRF hemodialysis patients.