Brain injury is the dominant cause of death for cardiac arrest patients who are admitted to an intensive care unit, and the majority of patients die after withdrawal of life sustaining therapy (WLST) based on a presumed poor neurologic outcome. Mild induced hypothermia was found to decrease the reliability of several methods for neurological prognostication. Algorithms for prediction of outcome, that were developed before the introduction of mild hypothermia after cardiac arrest, may have affected the results of studies with hypothermia-treated patients. In previous trials on neuroprotection after cardiac arrest, including the pivotal hypothermia trials, the methods for prognostication and the reasons for WLST were not reported and may have had an effect on outcome. In the Target Temperature Management trial, in which 950 cardiac arrest patients have been randomized to treatment at 33 degrees C or 36 degrees C, neuroprognostication and WLST-decisions are strictly protocolized and registered. Prognostication is delayed to at least 72 hours after the end of the intervention period, thus a minimum of 4.5 days after the cardiac arrest, and is based on multiple parameters to account for the possible effects of hypothermia.