In-Hospital CPR: When to Stop
Abstract & Commentary
Synopsis: Survival to hospital discharge following in-hospital cardiac arrest and CPR could be predicted with 99% sensitivity using a clinical decision aid that incorporated whether the arrest was witnessed, whether the initial cardiac rhythm was ventricular tachycardia or ventricular fibrillation, and whether a pulse was regained within the first 10 minutes.
Source: van Walraven C, et al. JAMA. 2001;285:1602-1606.
Van walraven and colleagues sought to validate a previously derived clinical decision aid to reliably predict patients who would have a poor outcome from in-hospital cardiac arrest. They used data from a large registry of in-hospital resuscitation attempts at a community-teaching hospital to determine whether the answers to 3 questions could be used to predict survival to hospital discharge:
1. Was the arrest witnessed?
2. Was the initial cardiac rhythm either ventricular tachycardia or ventricular fibrillation?
3. Was a pulse regained during the first 10 minutes of chest compressions?
If the answer to any one of the above questions was "yes," the patient was classified as having a reasonable likelihood of survival following resuscitation. If none of the questions could be answered in the affirmative, the patient was classified by the decision aid as having no chance for survival.
Data used were from 2181 cardiac resuscitation attempts (in 1884 patients) at the 550-bed hospital from 1987 through 1996. In 15.1% of resuscitations (327/2181), the patient survived to hospital discharge. For 99.1% of these successful resuscitations (324/327), the decision aid would have placed the patient in the favorable prognostic group (95% confidence interval, 97.1-99.8%). Only 3 of 269 patients (1.1%) who were predicted by the decision aid to have no chance of survival did survive to hospital discharge (negative predictive value, 98.9%), and none of these 3 individuals was able to live independently following discharge. van Walraven et al; conclude that this decision aid can be used to help physicians to identify patients who have an extremely small likelihood of benefiting from continued resuscitative efforts.
Comment by David J. Pierson, MD, FACP, FCCP
A number of factors have been shown to be associated with poor outcomes from attempted resuscitation from in-hospital cardiopulmonary arrest. These include such pre-arrest patient characteristics as hypotension, renal failure, metastatic cancer, pneumonia, and low functional status. Initial cardiac rhythms other than ventricular tachycardia or ventricular fibrillation are associated with unsuccessful resuscitation attempts. In addition, the longer the duration of attempted cardiopulmonary resuscitation, the less the likelihood of patient survival. Previous decision aids based on these and other factors have proven either too cumbersome for quick clinical application or have not been validated in relevant patient populations. This study appears to overcome these problems.
In their discussion, van Walraven et al point out that clinical decision aids should meet several strict methodological standards. These include a clinically important and easily determined outcome (in this case, survival and the ability to live independently) and the requirement that the aid itself be clinically sensible, using components that have been associated with survival in other studies. The latter is met by the present study, since whether an arrest is witnessed, whether ventricular tachycardia or ventricular fibrillation is the initial rhythm, and the duration of resuscitative efforts have all been independently associated with survival in previous studies. In addition, the aid should be validated in a well-defined, appropriately described population.
Use of the simple 3-factor decision aid described in this article could help clinicians to decide when resuscitative efforts should be discontinued after in-hospital cardiac arrest, thus preventing prolonged but ultimately futile resuscitations and the unfruitful use of critical care resources. As van Walraven et al point out, it could also be helpful in discussions with patients about resuscitation in the event of cardiac arrest: For patients who wished not to be subjected to invasive life support without a reasonable likelihood of benefit, the decision aid could be used to assure them that their wishes would be honored.
Dr. Pierson is Professor of Medicine, University of Washington, Medical Director, Respiratory Care, Harborview Medical Center, Seattle, Wash.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content