ABSTRACT
In the past few years, many nomograms and algorithms have been published predicting stage and outcome of clinically localized prostate cancer treated with radical prostatectomy or radiation therapy1-8. External validation of published nomograms is essential, because otherwise it remains unclear whether the predictive accuracy reported in the original study can be expected elsewhere. Different prostate-specific antigen (PSA) assays and differences in Gleason grading, surgical experience, number of surgeons and pathologists involved are just a few potential confounders that might disturb the performance of a nomogram when applied to a different cohort of patients. Therefore, differences related to the geographic region, country or continent that may impact upon each of the steps related to diagnosing prostate cancer may affect the performance of a nomogram that incorporates such diagnostic variables. In theory, in a stable nomogram, such differences should result in minimal predictive accuracy variations, as a truly generalizable nomogram should adjust properly for any such differences.
