CVD is one of the major causes of death in the Western world, and it annually afflicts approximately 100 per 100 000 population. In Sweden, with a population of nine million, approximately 15,000 people die each year due to CVD, two-thirds of whom suffer a cardiac arrest outside of a hospital. Approximately 2,000 of these out-of-hospital patients undergo cardiopulmonary resuscitation (CPR), which is administered by ambulance staff when they arrive on the scene. Upon arrival at the hospital, 15–50% of these patients are still alive, while one month after their cardiac arrest, only 5% are still alive.^1,2 In spite of major efforts to improve these results, little has happened over the past 25 years.

It is commonly agreed that the human brain cannot survive much more than five minutes of normothermic untreated circulatory arrest. Recent statistics from Sweden reveal that the median time from witnessed arrest to alarm is three to four minutes, and there is a median time of four to 10 minutes from arrest to ambulance arrival, and six to 10 minutes from alarm to defibrillation, indicating that only a minority of out-of-hospital patients are within reach of effective treatment.² It has been confirmed that the majority of these patients are elderly, but it is equally true that 25% of cardiac arrests strike patients between 20 and 60 years of age. Similar data are available from most Western countries, which emphasise the obvious need for improvement in the care of these patients, such as by means of pharmacological augmentation of central nervous system (CNS) tolerance for anoxia and its consequences.

The question is why, despite intense research and in-depth education of both ambulance and hospital crews, the final decades of the 20th century did not bring about improved results in the resuscitation of cardiac arrest victims. Part of the explanation was recently revealed when two research groups monitored the efficacy of CPR as practiced in both ambulances and the hospital.^3,4 This group observed that for practical reasons, efficient resuscitation could only be delivered for less than half of the resuscitation process. This resulted in coronary perfusion pressure and a blood flow that was too low, which is known to result in a low incidence of restoration of spontaneous circulation (ROSC), i.e. death.

Furthermore, insufficient systemic blood flow secondarily involves a risk of increased neurological injury. However, this knowledge has already led to improvements in education and training and, above all, development of automated technical aids that ensure continuous resuscitative efforts of improved quality, resulting in better circulation as a basis for successful resuscitation.

The 1990s ended in a rather pessimistic way regarding the possibilities of inducing neuroprotection by means of pharmacological agents. Although several hundred chemical compounds were tested and efficacy was experimentally verified in small animals, very few of the successful results were obtained with candidate drugs in larger animals when administered after the ischaemic event. This was even more the case in the clinical setting. However, there were also exceptions, where more positive results were achieved. Therefore, it was not a total surprise when Bernard et al.^5,6 published the successful results of their clinical investigations where cardiac arrest victims were subjected to mild hypothermia (32°C to 34°C) during 24 hours after ROSC. A rather significant improvement in neurological outcome was demonstrated.