An example of how patient selection can influence survival in glioblastoma multiforme.

Drawing on experience in evaluating postsurgical treatment of glioblastoma multiforme, it is concluded that results of clinical trials programs can be greatly affected by the status of patients who are selected for study. Investigations report results only on patients considered "evaluable," a portion of the total which may be less than 20%. Improvement in patient survival should not be conceived to be due to aggressive treatment unless patient selection has been accounted for.