In Reply:-In his letter, Dr. Maleck refers to previously reported cases of spontaneous recovery after unsuccessful resuscitation, including one of his own cases in which a patient was resuscitated for asystolic arrest for approximately 30 min and showed return of spontaneous circulation 5 min after cessation of resuscitation efforts. Dr. MacGillivray adds another case of asystolic arrest with spontaneous recovery after cardiopulmonary resuscitation (CPR). In both cases, resuscitative efforts were discontinued after an arbitrary time frame of 30 min.
The apparent question in the reported cases of the "Lazarus phenomenon," including my own, is whether resuscitation had been terminated prematurely. To date, there seems to be no consensus with respect to the timing of termination of CPR. The manual on advanced cardiac life support (ACLS) by the American Heart Association  states that resuscitation may be discontinued in the prehospital setting after an adequate trial of ACLS. An "extremely short attempt to reverse the arrest" is recommended for the intensive or critical care unit because of close monitoring by personnel and immediate attempts to reverse the arrest. In most clinical situations, CPR is discontinued after a full course of the relevant ACLS algorithm.
There are limited data on the predictability of outcome after CPR. Levine et al.  recommended the discontinuation of CPR if an end-tidal carbon dioxide level of <or= to 10 mmHg is measured 20 min after the initiation of ACLS. Dr. Maleck, in his letter, also makes reference to this approach. I would use end-tidal CO2as a parameter to gauge the effectiveness of resuscitation rather than to predict outcome. Nevertheless, it seems to be reasonable to discontinue CPR if effective circulation cannot be re-established over an extended period of time.
From my own and similar reported cases, I have learned that CPR should not automatically be discontinued if the end of the ACLS algorithm has been reached. Resuscitation may have to be continued until proven ineffective by parameters such as end-tidal CO2or metabolic deterioration.
Michael A. Frolich, M.D.
Department of Anesthesiology; University of Florida College of Medicine; Gainesville, Florida 32610–0254;email@example.com
(Accepted for publication March 25, 1999.)