A Systematic Review of Autoresuscitation after Cardiac Arrest*

K. Hornby, MSc; L. Hornby, MSc; S. D. Shemie, MD


Crit Care Med. 2010;38(5):1246-1253. 

In This Article

Abstract and Introduction


Objective: There is a lack of consensus on how long circulation must cease for death to be determined after cardiac arrest. The lack of scientific evidence concerning autoresuscitation influences the practice of organ donation after cardiac death. We conducted a systematic review to summarize the evidence on the timing of autoresuscitation.
Data Sources: Electronic databases were searched from date of first issue of each journal until July 2008.
Study Selection: Any original study reporting autoresuscitation, as defined by the unassisted return of spontaneous circulation after cardiac arrest, was considered eligible. Reports of electrocardiogram activity without signs of return of circulation were excluded.
Data Extraction: For each study case, we extracted patient characteristics, duration of cardiopulmonary resuscitation, terminal heart rhythms, time to unassisted return of spontaneous circulation, monitoring, and outcomes.
Data Synthesis: A total of 1265 citations were identified and, of these, 27 articles describing 32 cases of autoresuscitation were included (n = 32; age, 27–94 yrs). The studies came from 16 different countries and were considered of very-low quality (case reports or letters to the editor). All 32 cases reported autoresuscitation after failed cardiopulmonary resuscitation, with times ranging from a few seconds to 33 mins; however, continuity of observation and methods of monitoring were highly inconsistent. For the eight studies reporting continuous electrocardiogram monitoring and exact times, autoresuscitation did not occur beyond 7 mins after failed cardiopulmonary resuscitation. No cases of autoresuscitation in the absence of cardiopulmonary resuscitation were reported.
Conclusions: These findings suggest that the provision of cardiopulmonary resuscitation may influence autoresuscitation. In the absence of cardiopulmonary resuscitation, as may apply to controlled organ donation after cardiac death after withdrawal of life-sustaining therapies, autoresuscitation has not been reported. The provision of cardiopulmonary resuscitation, as may apply to uncontrolled organ donation after cardiac death, may influence observation time. However, existing evidence is limited and is consequently insufficient to support or refute the recommended waiting period to determine death after a cardiac arrest, strongly supporting the need for prospective studies in dying patients.


The physiologic transition from life to death is a complex process. The determination of death affects all physicians regardless of specialty, and modern, sophisticated medical technology has complicated rather than facilitated this process. The availability of life-sustaining interventions, such as cardiopulmonary resuscitation (CPR), mechanical ventilation, extracorporeal life support, ventricular assist devices, and other organ support or replacement technologies, has obscured our ability to distinguish between the seemingly discrete states of life and death. Yet, the practices of organ donation and transplantation necessitate this distinction. The ethical norm for organ donation is the "dead donor rule," which states that "vital organs should only be taken from dead patients and, correlatively, living patients must not be killed by organ retrieval".[1] For organ transplantation to be successful, the arrest of circulation and resulting warm ischemic injury (which occur at death and during organ procurement and transplantation) must be minimized. This conundrum is partially overcome when death is determined using neurologic criteria, because the brain-dead donor remains on a ventilator and circulation is maintained until surgical removal of organs.

Organ donation from brain-dead donors continues to be the preferred source of organs for transplantation; however, one of the responses to the persistent shortage of organs has been the re-emergence of donation after cardiocirculatory death (donation after cardiac death [DCD], which is also referred to as nonheart-beating organ donation). With advances in both transplant surgery and organ preservation techniques, the practice of DCD has progressively increased. DCD programs have developed throughout the world and now account for the largest incremental increase in organ donation in active programs in the United States.[2,3] There is an ongoing, focused attempt in the United States to increase the number of DCD donors.[4] Accompanying this renewed emphasis on DCD is the requirement to determine death as rapidly as possible after cardiac arrest to minimize any loss of circulation to the organs. This time pressure has forced the identification of a precise waiting period that is long enough to ensure the person has died but short enough to maintain organ viability for transplantation.

Death is generally understood to be based on the irreversible cessation of either brain function or circulatory and respiratory functions and the determination of death is a clinical matter that should be made according to widely accepted guidelines established by expert medical groups.[5] In the absence of organ donation, accepted medical practice for determining death after cardiac arrest has not included standardized diagnostic criteria or a specific time period of observation. In the setting of DCD, although recommendations exist for diagnostic criteria, there is a lack of consensus on how long circulation and respiration must cease for a person to be determined dead.[6–9] Internationally, this time period varies from 2–10 mins.[10] The historical influences on these timeframes include the 1995 International Maastricht Workshop (10 mins),[11] the 1992 Pittsburgh protocol (2 mins),[12] the 1997 US Institute of Medicine report (≥5 mins),[13] and the 2001 Report of the Ethics Committee, American College of Critical Care Medicine, and Society of Critical Care Medicine (≥2 mins but not >5 mins).[14] In a recent DCD pilot project in the United States, hearts from three severely brain-injured newborns were removed for transplantation soon after cardiac arrest (3 mins in the first case and 75 secs in the other two cases).[15] This variability in wait times required to determine death after a cardiac arrest, in part, reflects a lack of scientific evidence concerning autoresuscitation (AR).[16,17]

AR is the phenomenon of the heart being able to restart spontaneously and generate anterograde circulation.[18] Electrical activity of the heart (as measured by an electrocardiogram [ECG]) is essential to generate the contractile activity required for the heart to produce circulation. However, simply detecting the presence of some form of electrical activity is not an indication of contractile activity or of effective circulation. The determination of death after cardiac arrest is based on indirect measurements of circulatory arrest, including absent heart sounds, absent pulse, absent blood pressure, and cessation of breathing and neurologic function. Terminal ECG activity may persist in the absence of circulation and does not preclude the diagnosis of death.[10,19–23]

Given these considerations, if AR can occur, then the termination of circulation is not yet irreversible and the patient is not dead. The time limits of AR are uncertain. Therefore, we conducted a systematic review of the AR literature to answer the following question: how long after cardiac arrest can AR still occur? The primary objective of this review was to summarize the evidence on the timing of AR. We hypothesized that insufficient evidence exists to define the time limits of AR and the provision of CPR confounds these limits.

*See also p. 1377.