Trying to predict which patients will have progressive disease after certain treatments has been one of the main problems of cancer care. Using retrospective analyses of any database to find prognostic indicators is difficult at any time, but it is especially difficult if the number of patients is small because multivariate analysis of all other factors cannot be used effectively. To find predictive variables in a database of 258 patients, which is what the study by Lidsky et al1 tried to accomplish, is practically impossible. Thus, the finding that there were no significant predictors of response just underlines the fact that their study1 did not include enough patients to get this information.
Kemeny MM. Having Trouble Making Predictions: Comment on “Predicting Disease Progression After Regional Therapy for In-Transit Melanoma”. JAMA Surg. 2013;148(6):499. doi:10.1001/jamasurg.2013.706
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: