Substituted Judgment and the Decision to Withhold Life Support

Excerpt

To the Editors: The results of the provocative study by Seckler and colleagues (1) lack generalizability because of two major flaws in the study design. The investigators' choice of study setting, an urban geriatrics teaching clinic, introduces pre-selection bias into the study, and their use of a single-interviewer method creates an uncontrolled confounder (2).

In contrast to patients in previous studies (3), most of those (67%) in the study by Seckler and colleagues answered "yes" when asked whether they wanted to be resuscitated if they were no longer able to recognize family and friends and if they had a sudden