Skip to content

“Please randomize me – but don’t tell my family that you did”

Last week various newspapers (see here and here) reported on a planned research study of adrenaline for patients suffering a cardiac arrest outside hospital. The PARAMEDIC 2 trial (full protocol here) involves ambulance officers randomly giving patients either the traditional resuscitation drug adrenaline, or a salt-water solution (placebo). The trial has been strongly criticized by Ruth and Lindsay Stirton, writing in the Journal of Medical Ethics.

There are two main controversial elements to the trial design. The first involves the lack of consent for involvement in the trial, the second involves the researchers’ plan not to inform families of patients who died that their loved one had been in a research trial.

First, it is important to appreciate the context of this trial. 50,000 people in the UK each year have a cardiac arrest outside hospital. Contrary to public perception, the chance of survival if this happens is extremely low – less than one in ten patients leave hospital alive. Adrenaline is a naturally occurring stress hormone that has been used for a long time for patients whose hearts have stopped. But there is actually very little good evidence that adrenaline is helpful in this situation. There has only been one relatively small trial, which didn’t show a definite benefit. And there is some reason to think that adrenaline may be harmful. Very large studies (not randomized trials) that compared patients who had received adrenaline to those who had not received adrenaline suggest that adrenaline may increase the chance of patients’ hearts being restarted. But these studies also suggest that those patients who received adrenaline had a lower chance of surviving to leave hospital, and may have a higher rate of long term disability.

Is it ethical for a trial like this one to take place without consent? Normally research studies ask patients (or their surrogates) beforehand whether they would like to take part. But the nature of this sort of emergency treatment study is that consent beforehand is impossible. There is no way to predict which patients are going to have a cardiac arrest out of hospital, and no time to ask family members. Either the trial takes place without consent, or it does not take place at all.

The PARAMEDIC2 trial is a very good example of a study where it is ethically justifiable to perform research without consent. Research ethics guidelines (eg see here or here) allow research studies to go ahead without consent for emergency , life-threatening conditions like cardiac arrest. The question addressed by the study must be important, with no other way to answer it. There must be enough available evidence from animal studies and other types of research to suggest that a trial is needed. (In this case, the large international consensus group that provides guidelines for resuscitation (ILCOR) have reviewed available evidence and recommended that a trial take place urgently). And the trial must be carefully reviewed by relevant research ethics committees.

All of those conditions are met by the PARAMEDIC2 trial. And if I were to have a cardiac arrest out of hospital I would want paramedics to include me in the trial, to randomize me to receive adrenaline or not. This is not a self-less altruistic choice, wanting to contribute to science or the wellbeing of future patients. Given the existing uncertainty about whether or not adrenaline is beneficial or harmful, it would be in my best interests to be in the trial. It would be a rational way to manage the risks of resuscitation – adrenaline might be better at restarting my heart or it might not, it might give me a better chance of leaving hospital alive, or it might not, it might leave me in a more disabled condition if I survive, or it might not. The uncomfortable fact is that no one knows whether or not adrenaline is the right treatment. In that situation it is better to be in a trial where you have a 50% chance of receiving one of two treatments that might be helpful.


It is ethical to involve patients in the trial without consent. (Indeed, given the existing uncertainty, it is arguably unethical not to do the trial, and not to randomize). The second question is what to do afterwards. Should the patient or their family be told that they were in a trial? For survivors, the answer is straightforward. The patient or their family should be told – this respects that patient’s autonomy in a way that is possible and meaningful. It means that the patient/family can decide  whether or not to take part in ongoing follow-up that is an important part of the study. Some patients may choose to discontinue involvement in the research at that point.

The more difficult question is what to do where the patient has died. Sadly, most trial participants will be in this category. As noted above, 9 out of 10 patients who are enrolled in the trial are likely to die before leaving hospital. In this situation, informing family members that the patient was involved in a trial cannot benefit the patient, nor respect his/her autonomy. At that point they cannot decide not to take part in the study – the die is already cast, and no further follow-up is required. It does not give them any meaningful choices. Furthermore, informing families at this stage, grieving and distressed, is likely to be harmful to them. Although it is (as argued above) completely rational to be randomized to receive adrenaline or placebo in this situation, grieving is not by its nature a rational process. Families may well blame the research study or the researchers for the death of their family member. Guilt, denial and anger are normal parts of grieving. No matter how carefully explained by researchers, it would be difficult to avoid exacerbating some families’ distress and loss. David Hunter has argued that it is a “a matter of respect that you inform people”. But the key issue is how much weight we give to that respect compared with the risk of harming and distressing families by informing them about the trial.

If families are not informed, there is another risk – that down the track some families will discover about the trial anyway, and that delayed discovery of this fact would be even more distressing. This might lead to significant negative media attention, and could lead to the trial being stopped. It is possible that this would threaten future, important research studies of emergency treatments, and mean that they do not take place. These risks are important to consider too. One way of mitigating them is to have open discussion about the trial in the media, and information about the study widely disseminated within the press, internet and the community. That would help to avoid any perception that the trial was being done secretly or that there was an attempt to unethically experiment on vulnerable patients.


For myself, I would not want my family to be informed if I had been in the PARAMEDIC2 trial. It cannot help them, and it cannot help me. The researchers have chosen this path after carefully reflecting on the risks and benefits of various alternative approaches. The ethics committees that have reviewed the studies have presumably judged that most patients would feel as I do. It is important to reflect on the ethical questions around studies raised by trials like this. But it is even more important that the studies go ahead.



Image:  Wikimedia commons

Share on

2 Comment on this post

  1. Hi Dominic,

    This is, of course, a tricky issue. Two things, though. One, Stirton and Stirton aren’t exactly against the consent-less trial per se, and recognize the necessity of research without consent in some circumstances. But they do propose an interesting compromise that you seem to overlook: opt-out. The idea would be to bombard the community with information about the trial, and set up a system (like organ donor cards?) where people who wanted to actively avoid the trial could be flagged as opt-outers; they would just get the standard of care if there’s a cardiac arrest. Not that this is necessarily practicable- I believe the main downsides are a) low uptake (how many people would really opt out?) and b) expense – info bombardment and bureaucracy cost money, and this might be a big waste if not many would opt out.

    Two, the presumption against informing the family strikes me as problematic. There seem to be two worries: 1) harm to the family; and 2) irrational blame of research enterprise. On 1: this is partly an empirical question, but I believe the family’s desire to know the circumstances of the patient’s death should count as something strongly in their interest, and overriding that interest for their welfare is problematically paternalistic (more controversially, I think the failure to inform is in this case a form of deception-by-omission that should also be considered a harm). To be sure, empirical study of this (family reactions to learning loved ones died in a study, or even speculative interviews on whether family would want to be told) would help. But in the absence of compelling evidence of harm, we should side with the preferences of the family to be informed. (partial solution: ask family if they want to know certain further details of death, which the docs have reason to believe may be distressing…then again, just asking that further question might be distressing/trigger distress!)

    On 2 (irrational reaction): It is part of what it is to respect others’ autonomy, to respect potential irrational reactions. Maybe if you thought they would actively go out and harm someone, the info should be withheld (e.g., they’ll try to kill the researchers)- but that doesn’t seem likely. The reaction might lead to a worsened state of grief (in that case, see above – and again, I’d like to see empirical evidence of this). Or maybe the worry is, trust in research/medicine will be undermined. But then, we’d be in the situation of hiding information concerning circumstances of death from people in order to ensure their support for something – an extremely problematic manipulation at odds with liberal democracy. (and in any case, only a few people would be affected by the study – not too many votes lost)

  2. Hi Owen,

    1. Ruth and Lindsay do not propose ‘opt-out’ consent for the Paramedic2 trial – rather they suggest that this is what the trial involves, and criticise it on the grounds that most who end up in the trial will never have received information about it. “In order for opt out consent to be real, potential participants must be properly informed and there must be a real opportunity to opt out. ”
    I think they are right to suggest that the Paramedic2 trial isn’t adopting an opt-out consent framework, at least in the way that it is often used. Rather, a better characterisation of the trial consent arrangements is that the trial does not include patient consent at all – they have sought a waiver of the normal consent arrangements. However, they have provided the additional option of opt-out consent to satisfy those in the community who might have strong feelings about the trial and not want to take part. The opt-out element is not the study’s main approach to consent. The key question is whether a waiver of consent is appropriate – and I have argued above that it is.

    2. I agree that not informing the family is paternalistic, and that it could feel, or be perceived as a form of deception by omission.
    As for the paternalism – that would be a much stronger objection if we were talking about withholding information from a patient in order to avoid distressing him/her. (For example, the family member who asks a a doctor to conceal a cancer diagnosis to avoid upsetting the patient). Withholding information from a patient in order to prevent them from an irrational conclusion does seem to compromise their autonomy. But this criticism has much less force when it comes to information provision to family members who are not acting as surrogate decision-makers. If the patient is still alive, giving families information might be relevant to the decisions that they make for the patient. But where the patient has died that is no longer the case. Families have every right to know what has happened to the patient – what has caused their illness (where it is known), and why treatment was not successful. Families do not usually seek, and they are not usually provided with every detail of every medical intervention that was given.
    For those families who do have questions, and clearly express a desire to know more specific details – these should be provided, including information about the trial. As I understand it, the Paramedic2 trial does have a process to providing information to families who seek it out. But they do not plan to actively give information about the trial to all families of deceased patients.

Comments are closed.