Editorial Comment

Abstract

The authors indeed describe an elegant system for essentially automatically updating their web-based statistical prediction models. Most other authors, when they update models, simply do this periodically, as we have done 1 Stephenson A.J. Scardino P.T. Eastham J.A. et al. Preoperative nomogram predicting the 10-year probability of prostate cancer recurrence after radical prostatectomy. J Natl Cancer Inst. 2006; 98: 715-717 Crossref PubMed Scopus (513) Google Scholar , 2 Kattan M.W. Vickers A.J. Yu C. et al. Preoperative and postoperative nomograms incorporating surgeon experience for localized prostate cancer. Cancer. 2009; 115: 1005-1010 Crossref PubMed Scopus (64) Google Scholar with the original preoperative nomogram. 3 Kattan M.W. Eastham J.A. Stapleton A.M.F. Wheeler T.M. Scardino P.T. A preoperative nomogram for disease recurrence following radical prostatectomy for prostate cancer. J Natl Cancer Inst. 1998; 90: 766-771 Crossref PubMed Scopus (1099) Google Scholar The net result seems practically the same, because, as the authors point out, the rapid updates can have negligible effects themselves and must be accumulated before a noteworthy difference occurs. The periodically published updates, which are more typical, would seem to be easier for others to externally validate as opposed to a constantly moving target. However, for use within a single major cancer center, this may be less of a concern. Applying more weight to the more recently treated patients seems to make good sense and probably goes beyond adjusting for the year of surgery, which is more common, although optimizing the weights seems challenging. But the quality of the personalized reports being provided is truly fantastic.