3 °C/h and 2.8 °C/h).29 and 30 This interesting finding may be explained by a Hawthorne effect in these earlier studies over a short study period versus the daily life experience of several years in our cohort study. This could also eventually Everolimus be explained by limited documentation of temperature in the prehospital setting in the cohort study. There was no difference between prehospital

and IH cooled patients, however our small cohort was not powered to address outcome, which represented only a safety outcome in this report. Nevertheless, the time to first CPR-attempts in the prehospital cooling group was significantly longer than in the IH cooling group, which may have represented a bias against cooling efficacy. We recognize that the minimal decrease of Tes of prehospital cooled patients on admission in comparison to IH cooling may need to be optimized to maximize potential gains in neurological outcome. Also, time of onset

of cooling after admission at our department likely needs to be reduced to increase the potential effect of early cooling. Based on animal studies36 and 37 cooling might be more efficient if already applied during CPR resulting in an immediate decrease in temperature. If this could be achieved with a cooling method that minimizes key side effects such as rearrest or pulmonary edema, it could represent the best possible chance to demonstrate benefit. Our exploration GSI-IX research buy has several major limitations: first, it was a retrospective, non-randomized single center study including only a minority of treated patients over a prolonged period of time. Therefore, the absence of a selection bias is very difficult to confirm. In more than half of the patients cooled in the prehospital setting,

initial Tes was not measured and onset of cooling was not documented. There were no data available whether or not ambulance crews applied pads in the field or in the ambulance. Also, other specifics of cooling pad application (number of pads used, % of body surface covered) C-X-C chemokine receptor type 7 (CXCR-7) in the prehospital group were not available. Based on a low number of equipped ambulances, only about 15% of all OHCA patients admitted during the study period were cooled in the prehospital period. An additional selection bias might thus influence the results. Prehospital Tes in the IH cooling group was not measured. Without being able to compare prehospital temperatures of both groups before cooling makes it difficult to differentiate between active cooling in the prehospital group and spontaneous heat loss during transport in the IH cooling group. Furthermore, there was a delay between admission and the start of IH cooling. This delay might diminish the results in outcome. It remains unclear, whether a shorter delay of IH cooling would change the effects on outcome.