Christopher Poynter on 22-11-2013
There appears to have been a lot of hyperbole surrounding the release of the TTM study over the 2 days since its release. Phrases like “game changer” or “paper of the year” have been uttered and some have indicated that the foundations of their critical care knowledge are shaken. Wow! Strong stuff. This merits further exploring and provides the perfect platform for my opening Crit-IQ blog. Any subsequent discussion is most welcome.
Some background for those not yet aware. The NEJM released a paper at the AHA conference on Monday comparing 33 vs 36 degrees Celcius after cardiac arrest. No difference was demonstrated between the groups for either the primary outcome of mortality or secondary measures including severe neurological disability. It appears to be a well conducted study and little of the subsequent discussion around the water cooler and online has revolved around issues with study design. 947 patients, international (Europe and Australia) multicentre trial with well chosen inclusion and exclusion criteria, good protocols and meaningful endpoints.
The reason for the hype centres around the change of practice that occurred in 2002 after what really were 2 “game changing” articles on cooling post out of hospital VF arrest. Both trials were relatively small (total pts less than half of this trial) but conducted on opposite sides of the planet (Australia and Europe) and showing similar results (a marked improvement in mortality and neurological disability with cooling to 32-34 degrees for 12-24hrs post event). It was impossible to ignore NNT=6 (even though most thought it likely an overestimate) and so cooling post OHCA became the new norm. At this time, life was good. Everyone seemed to agree on this area of practice and the data supported that we were making a real difference with a relatively simple intervention.
As is the way with intensive care medicine - nothing stays the same. There have been skeptics and zealots in this debate over the last decade and so we find that this new study has given everyone a chance to confirm their world view (or in some cases challenge the entire fabric of their universe). Certainly this is a great opportunity to reassess the issue of cooling post OHCA and throw a little uncertainty and debate back into the mix.
So....What does this study show? What issues do I see? Is it a game changer or just another step along the long, tortuous path to critical care knowledge?
I’ll start with my further questions/issues. Firstly, it appears underpowered. The study is powered to find an 11% absolute reduction in mortality. This is extremely hopeful and would represent an almost unparalleled treatment success (notwithstanding the results from earlier in this field). An excellent blog by the people at St Emlyn’s have substituted that goal with a more realistic (and still clinically relevant) 2% ARR and demonstrated that almost 20,000pts would be required. So, there appears to be a case of absence of evidence rather than evidence of absence of effect. Having said that, there was no trend to improvement in the 33 degree group or any subgroups.
The only other issue I have had is that there were very small numbers in each group with poor neurological outcome at 180days. 50% were dead, with only 4% having poor neurological function and the rest a fairly good neurological (Rankin score 1-3) outcome. This is less than I would expect and I’m not sure whether that is just because I worry about severe brain damage more than I should. If so, then that is good. Looking at the supplementary material, it appears that the protocol for reccommendation of withdrawal of therapy was reasonable.
So, what does this all mean? This is not a new revolution in treatment of OHCA. However, there is no doubt in my mind that this is an excellent and important paper in advancing our understanding of the management of OHCA pts. The outcomes for both groups were better than the controls for the 2002 studies, so we are helping patients. This may indicate that preventing hyperthermia is what improves outcomes, not mild hypothermia.
I’m not sure it will precipitate a massive practice shift but it certainly takes the pressure off in those instances where cooling leads to undesirable side effects. I look forward to our journal club on Friday and the discussion that follows but expect that we will perhaps shift to an easier to achieve 35-36degrees in our unit.
I look forward to your (and your unit’s) views and response post TTM. Will this change your practice? If so, how?
1 Nielsen N, Wetterslev J, Cronberg T, et al. Targeted temperature management at 33°C versus 36°C after cardiac arrest. N Engl J Med 2013. DOI: 10.1056/NEJMoa1310519
2 The Hypothermia after Cardiac Arrest Study Group. Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med 2002;346:549-556
3 Bernard SA, Gray TW, Buist MD, et al. Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N Engl J Med 2002;346:557-563
login below or Click here to create new account
Gemma from Australia wrote 11-25-2013 10:40:39 am
Really interesting trial, not just because of the answers it provides, but the questions it generates.
The issue I'm stuck on now is that this was not 33 v nothing, its 33 v cooling to 36. There appears to be little difference of cooling to 33, but DOES mean we now need to cool to 36. Or do we?
Both groups received sedation for 36 hours, plus or minus neuromuscular blockers. The authors can't quantify this though as data wasn't taken. So, its possible that the "control" arm got unnecessary cooling, sedation and consequent complications, resulting in cost, delayed prognostication and morbidity.
My question now is, do we need to do a trial (3 arms?) comparing 36, normothermia preventing temp >37.5, and placebo?
John from Australia wrote 11-25-2013 11:58:10 am
Yes Chris, the under powering is a problem. I just wrote under the paper that this assumption of an 11% ARR was overly optimistic, and results in an under powering of the study - clinically meaningful outcomes, like a 2-3% ARR, remain possible.
So what we know is that TTH isn't absolutely smashingly sensational but may STILL be a worthy treatment.
Hella from Denmark wrote 11-25-2013 01:43:30 pm
I agree with Chris - there are not enough patients with severe neuro injury. Table 3 seems to show that on modified rankin score only 8% of survivors has a severe or terrible neuro injury, another 8% has moderate. This is a total 34 pts in one group, 38 in other. So how can you prove TTM improves chances of better neuro with such small number?
Alan from Australia wrote 11-26-2013 03:12:22 pm
I was going to raise the same point as Gemma - the big question now is whether or not cooling to 36 is the same as not cooling at all (and only preventing temp rises). If this is true, there is potentially a marked reduction in resource consumption and earlier prognostication.
Another trial appears necessary - another job for the ANZICS-CTG?
Christopher wrote 11-27-2013 07:21:05 pm
Thanks for your comments, all excellent.
I agree with Gemma that this is a trial that raises more questions than it answers. That is the great (and frustrating) thing about science.
As for the resource implications, not only is whether therapeutic hypothermia (33 or 36) is required an important question to get an answer to, but how long for. This study had a 36hr protocol, which has potential resource implications compared to the 12-24hrs of the earlier studies.
I had not considered Hella's point that the study was certainly underpowered to see alterations in neurological outcomes. Does this mean that throwing out 33degrees may be premature? Some have argued that if there is a benefit for one subgroup for therapeutic hypothermia, then there must be a risk for another. However, if that subgroup is as small as the severely brain injured group in this study, then perhaps not as it will not show any signal.
I was more interested in the small numbers of severely brain injured survivors from an ongoing prognostication and decision making point of view. It appears that very few of those who survive are left with severe brain damage (if the withdrawal protocol from this study which is supported by best current evidence is used), therefore this can potentially be communicated in family meetings and potentially guide ongoing decisions.
Crit-IQ has managed to track Niklas Nielsen (Lead author for TTM) down for an interview to put this study into perspective and get his take on where to from here so watch this space...
Todd Fraser from Australia wrote 12-16-2013 03:48:16 am
I've just added our latest podcast, with TTM lead investigator Niklas Nielsen. Lovely guy, generous with his knowledge of this critical area.
Make sure you take the time to listen in to this important podcast - http://www.crit-iq.com/index.php/Podcast
Mac from Austria wrote 09-17-2014 03:20:07 am
I just wanted to congratulate you in your analysis at the time.
At the time it was published, there was a lot of twitting but few deep reads like yours. Recently Niklas Nielsen admitted the study was underpowered. It seems they will need circa 500 patients more to come to a better CI. It would be nice if there would be more information about this second step of his research. Do you happen to know more about it?
Christopher from New Zealand wrote 10-06-2014 06:51:57 pm
Thanks for your comments. Sorry, I am unaware of the ongoing research plans but will keep my ear to the ground and comment here if I hear anything.