We appreciate Lewis re-emphasizing the important issue of patient selection and its potential role in explaining the wide variance in survival statistics found in the medical literature. It is important to note that other studies have documented rates of successful cardiopulmonary resuscitation similar to those found in our study.1 George et al,2 in 1989, found cardiopulmonary resuscitation survival rates of 55% and survival to hospital discharge rates of 23.3% in a community hospital setting.2 Other studies conducted in different settings with different patient populations have found survival rates to be significantly lower.
We agree with Lewis that one explanation for this variance may be that physicians in some institutions are better able to select patients with poor prognoses for "do not resuscitate" status. Institution-specific variations in comorbidities may also be responsible for a portion of the observed differences in survival. Our study found that conventional clinical definitions