To the Editor.
—The article by Ms O'Brien and colleagues1 on nursing home patient preferences regarding cardiopulmonary resuscitation (CPR) perplexed me. Not only is the baseline consent inadequate, but the study avoids the critical issues: When is CPR good medical care and therefore indicated? When not? When should it be offered?First, success rates for out-of-hospital resuscitation are dismal, except in exceptional circumstances. Murphy et al2 reported that of 244 elderly patients resuscitated out of hospital, including 40 nursing home residents, there were only 2 survivors and none when the arrest was not witnessed. Of these 2 patients, 1 did not have a cardiac arrest and 1 had a cardiac arrest after arrival in the emergency department, but before being evaluated. A study from Seattle, Wash,3 reported a 3% survival to hospital discharge for cardiac arrest among nursing home residents compared with a 36% survival to discharge.