If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author at: Research Center for Emergency Medicine, Department of Clinical Medicine, Aarhus University Hospital, Palle Juul Jensens Boulevard 99, Bygning J, Plan 1, 8200 Aarhus N, Denmark.
Drs. Behringer, Abella, and Sunde are concerned about our decision to conduct separate meta-analyses for outcomes at different time points and about the selection of trials for inclusion in the meta-analyses.
The systematic review group and the Advanced Life Support Task Force a priori decided that outcomes would be analysed separately at hospital discharge/30 days and after hospital discharge/30 days. The selection of trials for each analysis was based on the available data. As stated in the manuscript, trials reported only as abstracts were not included. We did not include abstracts, as it is impossible to thoroughly evaluate a trial and perform bias assessment based on an abstract. We decided to include the Laurent et al. trial as we considered it unlikely that hemofiltration would be an effect modifier for the relationship between cooling and outcomes. We are not aware of a biological rationale nor data to support such a relationship.
The meta-analysis provided by Behringer, Abella, and Sunde in Figure 1b is similar to eFigure 3 in our manuscript.
The differences are 1) inclusion of the Mori et al. trial, 2) inclusion of outcomes at hospital discharge from Bernard et al., and 3) exclusion of the Laurent et al. trial.
As noted above, we do not agree with #1 and #3. In Fig. 1a, we have added the small Bernard trial to eFigure 3 from the manuscript. The result is a risk ratio of 1.15 (95 %CI: 0.97, 1.37) instead of 1.11 (95 %CI: 0.94, 1.31) in the original figure. We note that there are some differences in the included numbers for the trials by Hachimi-Idrissi et al. and Nielsen et al. Specifically, the numbers of events are reported as 8/16 and 2/17 for the Hachimi-Idrissi trial and 218/469 and 222/464 for the Nielsen trial. We have re-assessed the numbers and believe that those included in our meta-analysis are correct. Specifically, the results from the Hachimi-Idrissi et al. trial are presented in Table 2 in the original manuscript.
For the Nielsen et al. trial, it appears that Behringer, Abella and Sunde have used Cerebral Performance Category score to define neurological outcomes, whereas we have used the modified Rankin Scale. We used the modified Rankin Scale given the recent recommendation from ILCOR.
We note two important considerations regarding the above meta-analysis and the interpretation of the results.
(1)
The weights in the meta-analysis are agnostic to the quality of the trials. The most recent trials (i.e., Nielsen, Lascarrou, and Dankiewicz) clearly have a lower risk of bias than the previous trials. Furthermore, the external validity of these newer multicentre trials is likely higher.
(2)
The random effects approach used above puts relatively more weight on smaller trials. A fixed effects approach gives a different result (Fig. 1b).
The authors refer to two observational studies. As noted in the systematic review, we did not include observational studies due to a high risk of bias from potential immortal time bias and confounding and given the number of randomised trials.
Taken together, we do not believe that the above results change our conclusions.
CRediT author statement
Lars Anderson: Writing – original draft.
Asger Granfeldt: Writing – review & editing.
Mathias Holmberg: Writing – review & editing.
Jerry Nolan: Writing – review & editing.
Jasmeet Soar: Writing – review & editing.
Declaration of Competing Interest
JN is Editor-in-Chief of Resuscitation.
JS is an Editor of Resuscitation.
AG, MH and LA declare no conflicts of interest.
References
Behringer W.
Abella B.S.
Sunde K.
Meta-analyses of targeted temperature management in adult cardiac arrest studies – the big picture is dependent on study selection!.
00423-8&id=gr1a.jpg){kind=link}
00423-8&id=gr1b.jpg){kind=link}