Abstract
AIM: Investigate how the results of predictive models of preoperative MRI for breast cancer change based on available data. MATERIALS & METHODS: A total of 1919 insured women aged ≥18 with stage 0-III breast cancer diagnosed 2002-2009. Four models were compared using nested multivariable logistic, backwards stepwise regression; model fit was assessed via area under the curve (AUC), R(2). RESULTS: MRI recipients (n = 245) were more recently diagnosed, younger, less comorbid, with higher stage disease. Significant variables included: Model 1/Claims (AUC = 0.76, R(2) = 0.10): year, age, location, income; Model 2/Cancer Registry (AUC = 0.78, R(2) = 0.12): stage, breast density, imaging indication; Model 3/Medical Record (AUC = 0.80, R(2) = 0.13): radiologic recommendations; Model 4/Risk Factor Survey (AUC = 0.81, R(2) = 0.14): procedure count. CONCLUSION: Clinical variables accounted for little of the observed variability compared with claims data.