Objective: New-onset diabetes after kidney transplantation (NODAT) has adverse clinical and economic implications. A risk score for NODAT could help identify research subjects for intervention studies.
Research design and methods: We conducted a single-center retrospective cohort study using pretransplant clinical and laboratory measurements to construct a risk score for NODAT. NODAT was defined by hemoglobin A(1c) (HbA(1c)) ≥6.5%, fasting serum glucose ≥126 mg/dL, or prescribed therapy for diabetes within 1 year posttransplant. Three multivariate logistic regression models were constructed: 1) standard model, with both continuous and discrete variables; 2) dichotomous model, with continuous variables dichotomized at clinically relevant cut points; and 3) summary score defined as the sum of the points accrued using the terms from the dichotomous model.
Results: A total of 316 subjects had seven pretransplant variables with P < 0.10 in univariate logistic regression analyses (age, planned corticosteroid therapy posttransplant, prescription for gout medicine, BMI, fasting glucose and triglycerides, and family history of type 2 diabetes) that were selected for multivariate models. Areas under receiver operating curves for all three models were similar (0.72, 0.71, and 0.70). A simple risk score calculated as the sum of points from the seven variables performed as well as the other two models in identifying risk of NODAT.
Conclusions: A risk score computed from seven simple pretransplant variables can identify risk of NODAT.