VF and VT account for only 25 % of OHCA cases [3]. For the remaining 75 % who experience non-VF/VT rhythms, the indications for receiving TH after ROSC are less clear. Although earlier randomized trials [4, 5] only examined OHCA due to VF, it can be reasonable to think that the effect of TH on brain injury after circulatory arrest would be the same regardless of the cause. This hypothesis was tested in 15 observational and two randomized studies. Regarding the randomized trials, both were not dedicated to study benefit of TH (one was a feasibility study on a helmet device for inducing hypothermia [15], and the other assessed whether high volume hemofiltration alone or with TH improve survival after cardiac arrest [16]). These trials included only 44 patients with non-VF/VT rhythms and found a nonsignificant survival benefit in the hypothermia group.

Among the 15 observational studies [17–31], the majority showed a nonsignificant trend toward better outcome with mild TH, but statistically significant survival benefit was shown only in few studies. In a multicenter observational study that included data from 19 centers, among which 197 developed non-VF/VT cardiac arrest and 124 received mild TH, the rate of survival to hospital discharge was significantly higher in mild TH-treated patients (p = 0.023) [18]; however, only univariate analysis was performed. Also, selection bias was a concern, because decision of hypothermia treatment was at the discretion of the treating physician. A meta-analysis evaluated these 17 studies (two randomized and 12 observational) that included 1,336 non-VF/VT patients, out of which 30.8 % were treated using mild TH [32]. The quality of evidence in all studies was low with a substantial risk of bias and high degree of imprecision due to small sample size, and therefore the results should be interpreted cautiously.

Some studies showed benefit of implementing TH in OHCA due to VF/VT, but not in non-shockable rhythms. In a retrospective study that included 491 patients with OHCA (of whom 74 % had non-VF/VT cardiac arrest), there was no significant improvement in patients resuscitated from non-VF/VT rhythms, but there was a significantly higher rate of survival and favorable neurological outcome in the VF/VT group who received TH [19].

Several questions were raised after a recent trial showed no benefit for TH in VF OHCA. In the study by Nielsen et al. [10], the median time for ROSC was 25 min, with a wide range from 18 to 40 min. One may expect that the reduction of neurological metabolism by hypothermia will not benefit the already damaged neurons by prolonged cardiac arrest. Also, up to 4 h were permitted to start the cooling process after OHCA, and four more hours were allowed to achieve a mean temperature of <34 °C. The delay in starting hypothermia protocol and achieving target temperature might have affected the outcome. The neurological outcome was determined by the cerebral performance category and modified Rankin scale, which assess the patient’s capability of daily activity but does not evaluate for fine cognitive impairment. Also, the rapid rate of rewarming from 33 to 36 °C in a 6 h period can be harmful and may abolish a potential benefit from hypothermia. It should be noted that in this study by Nielsen et al., the TH group was compared with a group of normothermic patients with a targeted temperature of 36 °C, while in the earlier studies [4, 5] the control group did not receive thermal control. There is evidence for the negative effect of hyperthermia in the 72–96 h post-ROSC, as it was found to be associated with increased mortality and poor neurological outcome [33].

Regarding the value of prehospital compared with in-hospital cooling, Kim et al. [14] found less than 1 °C temperature difference at hospital admission between both groups. One can assume that the small difference in temperature did not allow for an expected change in outcome. Moreover, the use of cold fluids to achieve hypothermia can be associated with pulmonary edema. In a review of experimental studies, the utilization of cold fluids to achieve intra-arrest hypothermia was associated with a poorer outcome compared with other cooling strategies [34]. In the same study, there was an 11 % higher absolute rate of pulmonary edema on arrival to the emergency department in the interventional group. The 2 L of saline given rapidly after ROSC caused negative hemodynamic effects. This conforms to prior animal studies that demonstrate a reduction in coronary perfusion pressure when saline load is given to achieve cooling [35]. This adverse outcome was not observed when cooling was achieved with other methods. Therefore, the outcome of this study might be an effect of the method used rather than an effect of hypothermia. The trial was powered to show a 30 % improvement in outcome, so a modest treatment effect may have been missed. Also, the quality of cardiac arrest care is very high in Seattle (site of study conduction), which might have masked a subtle benefit of hypothermia.

In conclusion, the benefit of TH in OHCA demonstrated initially by the two studies in 2002 was not reproduced by the recent larger trial. How should this influence our current practice? First, many questions still need to be answered. What is the ideal target temperature? Is intra-arrest cooling superior to later cooling? How fast should the target temperature be reached? How long should hypothermia continue? Does cooling work for patients with asystole or pulseless electrical activity? Is intravascular cooling or surface cooling more effective? Therefore, before abandoning TH for OHCA, further rigorous randomized trials should be performed targeting possible concerns that might have attenuated its benefit in recent studies.