Date: November 6th, 2019
Reference: Lascarrou et al. Targeted Temperature Management for Cardiac Arrest with Nonshockable Rhythm. NEJM Oct 2019
Guest Skeptic: Dr. Laura Melville (@lmelville535) is an emergency physician in Brooklyn, New York, is a part of the New York ACEP Research Committee, ALL NYC EM, and is the NYP-Brooklyn Methodist Resident Research Director.
Case: A 59-year-old woman comes is brought into your emergency department (ED) by EMS in cardiac arrest. She had a witnessed arrest, and CPR was initiated by bystanders. Her initial rhythm in the field was reported as pulseless electrical activity (PEA) by EMS. The patient achieved return of spontaneous circulation (ROSC) on arrival to the ED. You call your hyperthermia team to initiate targeted temperature management (TTM), which in your hospital means 33C for 24 hours followed by slow rewarming for 24 hours. Your senior resident asks you “should we really be cooling our patient to 33C, doesn’t the data suggest 36C is just as good? And if she was not in a shockable rhythm at arrest, will she be likely to benefit from this treatment?” The patient’s family has separately mentioned they heard she might have a better chance of being “normal” if she gets cooled down. What do you say? Do you continue with the ICE Code? What do you tell the patient’s family?
Background: We have covered therapeutic hypothermia many times on the SGEM. This has been or out-of-hospital cardiac arrests (OHCA). Therapeutic hypothermia has not been demonstrated to have benefit in the pre-hospital setting (SGEM#54 and SGEM#183).
But two earlier randomized controlled trials (Hypothermia after Cardiac Arrest Study Group 2002 and Bernard et al 2002) showed benefit for good neurologic outcome when TTM was initiated in the hospital after ROSC was achieved. In those studies, the temperature goal was 32C-34C and 33C respectively.
The SGEM covered the targeted temperature management (TTM) trial published in the NEJM. It showed cooling patients to 33C was not superior to 36C for the primary outcome (SGEM#82). The most recent time we have looked at therapeutic hypothermia was SGEM#199. This was a trial looking to see if there was a neuroprotective effect of hypothermia in patients with status epilepticus. Unfortunately, that study failed to demonstrate a benefit of therapeutic hypothermia for adult patients admitted to the ICU with convulsive status epilepticus.
It seems like TTM is a good example of an intervention that “makes sense” but doesn’t always work. There are many examples like this in the literature where something makes sense from a pathophysiologic standpoint but is not demonstrated to work when properly tested.
Clinical Question: Does therapeutic hypothermia improve survival with good neurologic outcome in patients who achieve ROSC after non-shockable cardiac arrest?
Reference: Lascarrou et al. Targeted Temperature Management for Cardiac Arrest with Nonshockable Rhythm. NEJM Oct 2019
- Population: Adults (18 years and older) with OHCA or IHCA of any cause, with non-shockable rhythm and a Glasgow Coma Scale (GCS) score of 8 or lower.
- Exclusion: No flow time of more than 10 minutes (collapse to starting CPR), low-flow time of more than 60 minutes, major hemodynamic instability (continuous vasopressor infusion), time from cardiac arrest to screening >300 minutes, moribund condition, severe hepatic dysfunction, pregnant or breast-feeding, prisoner, lack of health insurance and decide not to participate (by next of kin).
- Intervention: Targeted temperature management to 33C (+/- 0.5C) was started post arrest and then maintained for 24hrs. Cooling protocol was determined by each of the 25 participating sites. Slow rewarming of 0.25-0.5C/hr to target of 36.5-37.5C, which was maintained for 24hrs.
- Cooling: Active internal cooling with a specific device, active external cooling with a specific device, or active external cooling without a specific device.
- Comparison: Targeted normo-therapy to 36.5-37.5C for 48 hours
- Primary Outcome: Survival to good neurologic outcome at 90 days as defined by Cerebral Performance Category (CPC) scale score of 1 or 2.
- Secondary Outcomes: clinicaltrials.gov listed 20 outcomes (NCT01994772). The methods section only mentioned six (mortality, mechanical ventilation, length of stay in the ICU and hospital, infection and hematologic adverse events).
Authors’ Conclusions: “Among patients with coma who had been resuscitated from cardiac arrest with nonshockable rhythm, moderate therapeutic hypothermia at 33C for 24 hours led to a higher percentage of patients who survived with a favorable neurologic outcome at day 90 than was observed with targeted normothermia.”
Quality Checklist for Randomized Clinical Trials:
- The study population included or focused on those in the emergency department. No
- The patients were adequately randomized. Yes
- The randomization process was concealed. Yes
- The patients were analyzed in the groups to which they were randomized. Yes
- The study patients were recruited consecutively (i.e. no selection bias). Yes
- The patients in both groups were similar with respect to prognostic factors. Yes
- All participants (patients, clinicians, outcome assessors) were unaware of group allocation. No
- All groups were treated equally except for the intervention. Yes
- Follow-up was complete (i.e. at least 80% for both groups). Yes
- All patient-important outcomes were considered. Yes
- The treatment effect was large enough and precise enough to be clinically significant. No
Key Results: The screened 2,723 patients for eligibility and 584 underwent randomizations. The median age was 67 years, two-thirds were male, and three-quarters were OHCAs.
Survival with good neurological function was higher in the TTM group
- Primary Outcome: Survival to discharge with good neurological outcome as measured by a CPC score of 1 or 2 was statistically better in the TTM group compared to the Usual Care group.
- 10.2% TTM vs. 5.7% Usual Care (absolute difference 4.5%), p=0.047 which gives an NNT of 22
- Hazard ratio 4.5 (95% CI; 0.1 to 8.9) and a fragility index of 1
- Secondary Outcomes: There was no statistical differences in any of the secondary outcomes
- Mortality at 90 days 81.3% TTM vs. 83.2% Usual Care (95% CI; −8.0 to 4.4).
- No statistical difference was reported for mechanical ventilation, length of stay in the ICU and hospital, infection and hematologic adverse events.
1) Statistics: They based their sample size calculation on the assumption that there would be survival with good neurologic function in 23% of the TTM and 14% of the usual care. This would mean they expected a 9% absolute difference. The actual result was only 10.2% for TTM and 5.7% in usual care (4.5% absolute difference). This often happens in research. My EBM mentor Dr. Andrew Worster who taught me also told me if you want to make an outcome rare all you need to do is study that outcome. Here they thought they would have a higher prevalence of the primary outcome and yet in this population it was less than half.
Besides basing their sample size on this expected outcome, they also fell a few patients short of their target. The goal was to get 584 participants but three withdrew consent so left them with 581. Again, because their outcome of interest occurred less often than anticipated, and the difference between the two groups was half what was expected, the study was underpowered.
2) P-Values and Fragility Index: We have discussed the problem with being dichotomous about p-values and the utility of the fragility index. They did report a statistical difference between the two groups for the primary outcome, but the p value was 0.047. That does not mean therapeutic hypothermia works or does not work but rather needs to be interpreted as the probability of false rejecting the null hypothesis and making a type I error.
The fragility index is linked mathematically to the p-value and is another way of representing the data. In this study the fragility index was 1. This means that changing the outcome of one participant would have made the results statistically non-significant on a subjective outcome measure susceptible to bias due to lack of blinding and reliability of the CPC score.
3) Lack of Blinding: This is a huge limitation of this study. While the unconscious patients and the outcome assessor were blinded to group allocation, the clinicians were not blinded. This introduces bias that would probably be directed towards treatment. The hypothesis was that TTM would provide a patient-oriented benefit (superior to usual care). Patients could have consciously or unconsciously been treated differently by the clinicians. These potential subtle differences in management could be responsible for the fragile statistical difference demonstrated.
4) Temperature Management: The true duration of temp control seemed at first like both groups were exposed to temperature management for the same amount of time (48 hours). However, temperature management was performed between 8-16 hours longer in the 33C group, due to additional time required for re-warming.
In addition, a number of patients had their body temperatures rise above 38C, in particular after the period of TTM. This could have impacted the results and suggests the target should have been 36C for the usual care to prevent hyperthermia.
5) Outcome Assessment: This builds on nerdy point #3. The outcome was done by a single psychologist. They were blinded to the group allocation, but the patient was not. They were being questioned on their neurologic function and could have known if they were or were not in the treatment group thought to be superior. The CPC is a subjective assessment not objective and was done over the telephone. There would be a bias for the patient to say they were better if allocated to the TTM group.
In addition, the inter-rater reliability of the CPC score in post-arrest cases is known to be poor. One study from Grossestreuer et al (Resuscitation 2016) demonstrated disagreement between assessors more than 1 in 5 times (22%) giving a kappa of only 0.66. If another psychologist conducted the assessments, it is very likely they would not have assigned the same CPC score for the survivors. This nerdy point comes from our #FOAMed friend Dr. Josh Farkas from PulmCrit.
Comment on Authors’ Conclusion Compared to SGEM Conclusion: We would say that limited data suggests that patients with OHCA/IHCA with a non-shockable rhythm had a statistically higher chance of a good neurologic outcome with TTM.
SGEM Bottom Line: We do not have good evidence to routinely recommend TTM in patients with non-shockable cardiac arrests.
Case Resolution: You initiate your ICE protocol based on your current guidelines, not influenced by the patient’s initial rhythm.
Clinical Application: These results are too fragile to change practice. However, most institutions are applying their ICE protocols to all OCHA/IHCA patients that achieve ROSC and meet inclusion/exclusion criteria without regard to initial rhythm. The major take-away from this is that 37C may be a poor choice of target as it seems to allow for fever to develop. Otherwise we cannot say if TTM to 33C or 36C will improve outcomes for patient likelihood of survival with good neurological outcome at 90 days.
What Do I Tell My Patient? You tell the patient’s family that their loved one has experienced a cardiac arrest, but that her heart is now beating on its own. You explain she has been put into a targeted temperature control protocol, and that her prognosis is grave. Right now, there is no clear data to help us know if her initial rhythm will have any influence on how she will do, but the fact that she got bystander CPR could be very helpful.
Keener Kontest: There was no winner last week. The answer we were looking for is that the Butterfly IQ is to be available in Canada sometime in early 2020.
Listen to the podcast to hear this weeks’ trivia question. If you know the answer, send an email to TheSGEM@gmail.com with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.