A lady in her early sixties is out in the town centre with her husband doing some shopping on a crisp December afternoon. They stop for a break and get coffee.
The lady can recognise the subtle look of anxiety on her husband’s face.
“You look a bit grey, are you feeling alright?”
“Oh, yes. Yes, fine. My chest just feels a little tight.”
“Use your spray the doctor gave you.”
He rummages around in his pocket for it. She sees a glisten of sweat on his brow.
“Did you bring it out with you?”, she asks.
He doesn’t respond. His eyes are vacant, staring in to space. He sways to the side, vomits over the table and collapses to the floor.
Welcome to another instalment of #PHJC. This time we’re looking at the paper:
According to the paper, 70% of of out-of-hospital cardiac arrest’s are due to myocardial infarction or pulmonary embolism, therefore it sounds logical that giving a drug that can dissolve intravascular blood clots might help these patients. However, as we know, what might sounds reasonable on a biological level doesn’t always add up to helping patients after the research is done.
This is a RCT published in 2008 in the New England Journal of Medicine, one of the most prestigious medical journals in the world. However, this doesn’t mean we should just take it at face value, a good critical appraisal is still in order!
As always, don’t take my word for it, read the paper and come to your own conclusions.
Did the trial address a clearly focussed question?
The population studied – Adults with witnessed out-of-hospital cardiac arrest of presumed cardiac origin with initiation of basic or advanced life support within 10 minutes after collapse.
Patients with suspected non cardiac cause of arrest, known internal bleeding, neurologic impairment, coagulation disorders, pregnancy, participation in any other clinical study, hypersensitivity to the study medication, institutionalisation of the patient or any other condition that the investigator believed would place the patient at increased risk if included in the trial were excluded.
It’s not stated exactly how the research team that enrolled patients in to the study came to the presumption that the cause of cardiac arrest was cardiac and it wasn’t possible to do autopsies on all patients that were enrolled to see whether they did actually have a cardiac origin of their death, however I suppose this is what would happen in real-life clinical scenarios.
Intervention given – Tenecteplase, dosed according to estimated body weight.
No antithrombin and antiplatelet were given to patients as part of the protocol. I know in the UK, some Critical Care Paramedics will thrombolyse during resuscitation for presumed cardiac cause of death, using tenecteplase and heparin.
Comparator given – Placebo
Outcomes considered – Primary outcome was 30-day survival.
I think neurologic outcome is generally considered a more patient-centred outcome and would have been preferable as the primary outcome.
Secondary end points were hospital admission, return of spontaneous circulation (ROSC), 24-hour survival, survival to hospital discharge, and neurologic outcome of surviving patients.
There were also safety endpoints, such as symptomatic intracranial haemorrhage or major bleeding complications that were considered life-threatening, fatal or led to haemodynamic compromise requiring intervention.
Was the assignment of patients to treatments randomised?
This was a randomised controlled trial, the ‘gold standard’ of study design to investigate the effects of an intervention.
Patients with asystole or pulseless electrical activity as initial rhythm were randomised immediately after a cannula was in place. Patients in VF or pulseless VT were randomised if up to three shocks failed to achieve ROSC.
Assignment of study drugs were randomly generated and stratified according to individual ambulance in blocks of four (read more about block randomisation in this #PHJC blog).
Open-label treatment with thrombolytics was allowed if the care providers believed the patient’s cause of cardiac arrest was pulmonary embolism.
Were all of the patients who entered the trial properly accounted for at its conclusion?
12 months in to recruiting patients, the data safety monitoring board looked at the 30-day survival results from the first 300 patients and made a recommendation to discontinue enrolling asystolic patients due to an extremely low survival rate (1 of 103). After observing nearly identical survival rates in the two groups, the data monitoring board requested a futility analysis be performed, using data from 653 patients (1014 had undergone randomisation) and recommended suspension of the trial.
‘Futility’ refers to when a clinical trial is not going to achieve it’s objectives, for instance stopping a trial if the results suggest the study isn’t going to reach statistical significance . This can save more patients than is needed getting a treatment that might help them, or could even be harming them, and save additional resources in carrying on an expensive trial. At a final meeting, on the basis of 1050 patients who underwent randomisation, the data safety monitoring board recommended the trial be stopped.
No patient was lost to 30-day follow-up. Complete follow-up data were available for 1032 out of 1050 who were enrolled in the study.
Missing end-point data were imputed according to a worst-case scenario.
Were patients, health workers and study personnel ‘blind’ to treatment?
This was a double-blind study. Patients were masked to what they were receiving (as they were dead). Clinicians and study personnel were also masked to what intervention the patient had received.
Treatment assignment was unblinded in 33 patients who had received tenecteplase and 22 who had received placebo. This was mainly due to safety reasons.
Were the groups similar at the start of the trial?
The two trial groups were similar in almost all respects with regard to clinical profile, event intervals and concomitant medications before and during CPR. Only 21% of the participants were female.
There was a statistically significant difference in terms of the presumed cause of cardiac arrest, with more patients in the tenecteplase group thought to have died as a result of acute myocardial infarction (74.8% vs 68.5%, p <0.01).
Has ethical approval been obtained if appropriate?
The Declaration of Helsinki is a list of ethical principles regarding the use of humans in research. One of the fundamental principles is that people should have the right to make informed decisions about their participation in research.
This brings up a dilemma for this study. How can you make an informed decision about participating in a study when you’re dead?
In fact, can you make an informed decision when you’re in a lot of pain, or have an altered mental status due to an illness of injury, or just very anxious and worried? No, but we still need to do important research for patients in these horrible situations, which is why there are special allowances for recruiting patients in emergency circumstances to research studies. In the UK, the Heath Research Authority website states:
In the UK the law allows adults not able to consent for themselves to be recruited into Clinical Trials of Investigational Medicinal Products (CTIMPs) without prior consent in emergency situations if:
- Treatment needs to be given urgently;
- It is also necessary to take urgent action to administer the drug (IMP) for the purposes of the trial;
- It is not reasonably practicable to obtain consent from a legal representative;
- The procedure is approved by a NHS Research Ethics Committee;
- Consent is sought from a legal representative as soon as possible.
How large was the treatment effect?
14.7% of tenecteplase group survived compared to 17% of the placebo group (RR of survival 0.87, 95% CI 0.65 – 1.15, p = 0.36). These results would suggest that tenecteplase does not improve 30-day survival of patients in out-of-hospital cardiac arrest with suspected cardiac cause.
There were also no statistically significant differences in terms of any of the secondary outcomes, other than one of the safety endpoints – intracranial haemorrhage. The results showed an increased frequency of intracranial haemorrhage in the tenecteplase group (tenecteplase 2.7% vs placebo 0.4%, RR 6.95, 95% CI 1.59 – 30.41, p = 0006). The confidence interval is quite wide for this result, probably due to the the small numbers in each group.
This is another great example of why we need to be careful about making decisions on whether we should change our practice, based on lower level evidence. The paper cites a few other smaller observational studies that demonstrated a benefit in thrombolysis during resuscitation, which could potentially have been used as evidence to support a change in patient care. However, once this large, well powered, experimental study was conducted, it showed different results to the observational studies, that there was no difference in 30-day survival and an increased risk of intracranial haemorrhage.
I am definitely not suggesting we should disregard observational research as a useful means of researching certain topics, I’m just emphasising the importance of being aware of limitations of certain types of studies, when making decisions that could have serious implications for patients. There is a great paper by Professor Nick Black on why we need observational research to evaluate the effectiveness of heath care, you can read by clicking here.
It was suggested in the discussion section of the paper that had antithrombin and antiplatelet agents been given in addition to teneceteplase, increased benefit may have been demonstrated. This is possible, however we won’t know for certain unless another study that investigates this.
Can the results be applied in your context? (or to the local population?)
As this was a large multi-centre randomised controlled trial, this makes the results more generalisable. It was set in the prehospital environment amongst 66 different EMS systems in a variety of countries. The EMS systems involved may have different set-up in how they operate compared to the UK, for example being physician led or using Fire and Rescue service to respond. However, taking this in to consideration, I still believe the results can be applied in a UK EMS context.
Are the benefits worth the harms and costs?
The results of this study demonstrate no benefit in giving tenecteplase during resuscitation for patients with presumed cardiac cause of out-of-hospital cardiac arrest. However, a higher frequency of intracranial haemorrhage was reported amongst patients that received tenecteplase. Considering this, it would seem there is no benefit, yet an additional risk of harm.
If you’ve read the paper and come to a different conclusion, I would love to hear from you. Please post any comments below and I will try my best to respond!