A Nomogram for Predicting BK Virus Activation in Kidney Transplantation Recipients Using Clinical Risk Factors

Abstract

BK virus is a common opportunistic viral infection that could cause BK virus-associated nephropathy in renal transplant recipients. Thus, we retrospectively analyzed clinical and laboratory data associated with a higher risk of BK virus activation from 195 renal transplant recipients by the multivariate logistic regression analysis and performed the external validation. Results showed that patients with BK virus active infection were associated with a deceased donor, had lower direct bilirubin levels, a higher proportion of albumin in serum protein electrophoresis, and lower red blood cells and neutrophil counts. The multivariate logistic regression analyses revealed that the living donor, direct bilirubin, and neutrophil counts were significantly associated with BK virus activation. The logistic regression model displayed a modest discriminability with the area under the receiver operating characteristic curve of 0.689 (95% CI: 0.607–0.771; P < 0.01) and also demonstrated a good performance in the external validation dataset (the area under the receiver operating characteristic curve was 0.699, 95% CI: 0.5899–0.8081). The novel predictive nomogram achieved a good prediction of BK virus activation in kidney transplant recipients.