Answers are grounded in this paper and cite their sources.
In this study, we used 327 cases of localized prostate cancer to determine the information content provided by 5 popular nomograms for predicting outcomes in localized prostate cancer. All study patients underwent radical prostatectomy. For each case and each nomogram, we calculated the estimated probability of outcome, and, from this probability, we calculated the information content as 1-S, where S is the entropy. With this definition, information content is minimized at 0 and maximized at 1. We found that the average information content ranged from 0.16 for the Partin tables to 0.44 for the recent Kattan nomogram for 10-year disease-free survival. Furthermore, the Kattan 10-year nomogram provided information content greater than 0.5 for 50% of study cases, so that among these 5 nomograms, we judged its performance the best. Nevertheless, because even this nomogram provided less than 0.5 information content for 50% of our cases, we believe that it can be improved and that additional measurements or markers observed on the biopsy tissues are likely to produce better nomograms.
No takes yet. Share an insight, caveat, or question.
Bismar et al. (2007) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: