Survival after in-hospital cardiopulmonary arrest of noncritically ill patients: A prospective study

Background: The rising healthcare costs and the ethical and economic implications of cardiopulmonary resuscitation (CPR) have generated interest in defining criteria to predict the appropriateness of CPR in specific patients. Age has been proposed as one such a criterion. Methods: As part of a quality assurance program, all instances of CPR (code-500) at our VA Medical Center were prospectively studied over a period of 45 months. Only events in noncritical care hospital areas were included in this analysis. The CPR data were prospectively collected, and follow-up of initial survivors was continued until the end of the study period or until a patient died. Results: Of a total of 422 code-500 events, 387 (92 percent) met our study definition of cardiorespiratory arrest, and 255 of these occurred in a noncritical care area and were included in the study. Our immediate survival was 52 percent (n=132), survival after intensive care unit (ICU) stay was 22 percent (n=55), survival to hospital discharge was 11 percent (n=28), and 4 percent of the patients (n=10) were alive at the end of follow-up (mean, 22 months). None of the patients discharged alive had a significant new neurologic deficit, and all but one returned to their preadmission environment. The post-CPR hospital charges for each of the surviving patients was estimated at $63,000. Age, the admitting diagnosis, and main comorbidity did not predict long-term survival. The post-CPR Apache II score correlated with a patient surviving the ICU stay, but did not correlate with long-term survival either. Conclusions: Age alone is not a valid criterion to decide whether a patient is a suitable candidate for CPR, and the principal diagnosis and main comorbidity at the time of admission do not appear to predict long-term survival either. Whether in-hospital CPR in noncritical care areas is cost-effective is an issue that society at large must eventually decide.