In Reply Drs Peacock and Sterz comment that detailed information on cooling method, target temperature, and time to achieved target temperature would provide valuable insights into our findings. These data were not collected because Get With the Guidelines (GWTG)-Resuscitation is not solely a registry for therapeutic hypothermia treatment, but rather collects extensive data (>250 variables) on the totality of in-hospital cardiac arrest care. Although GWTG-Resuscitation has recently introduced a postresuscitation module to collect detailed information on temperature management, this optional module has been adopted by only a few hospitals and has enrolled few patients treated with therapeutic hypothermia. Nonetheless, we assert that cases of ineffective induction of therapeutic hypothermia (ie, not achieving temperature targets <34°C) in the study's hypothermia-treated patients would have biased the results toward the null and should not have resulted in lower survival rates in the hypothermia-treated group. It is possible that overcooling below recommended targets of 32°C (which occurred in approximately 21% of the study's hypothermia-treated patients) could have worsened outcomes in this group. As such, we believe that the observations represent a "real-world" estimate of the delivery and effectiveness of therapeutic hypothermia for patients with in-hospital cardiac arrest in the absence of clinical trials.Dr Bougouin and colleagues raise the possibility that the non-hypothermia-treated patients may have undergone active cooling because their lowest achieved temperature appeared similar to the 36°C group of the TTM trial. In fact, the lowest achieved temperature in the 36°C group of the TTM trial was a mean of 35.3°C, with mean hourly temperatures during the first 24 hours consistently between 35°C and 36°C. 1 In contrast, the lowest achieved temperature in our study's non-hypothermia-treated patients was higher, and we now report that the highest achieved temperature in this group was a mean of 37.4°C (SD, 1.3°C) and a median of 37.4°C (interquartile range, 36.8°C-38.1°C)-far exceeding the TTM trial's maximum 36°C target. Moreover, the TTM trial was not published until December 2013, a year before our study period ended. Given these 2 reasons, the proportion of our study's non-hypothermia-treated patients who were managed with an active cooling strategy with a target temperature below 36°C is likely small.We recognize the limitation of the observational nature of our study and reiterate the need for a randomized clinical trial to confirm the findings. Meanwhile, use of therapeutic hypothermia for patients with in-hospital cardiac arrest may need reconsideration pending evidence documenting clinical benefit.