Future Infectious Catastrophic Disease Outbreaks: Ethics of Emergency Access to Unregistered Medical Interventions and Clinical Trial Designs

The mass media excitement about Ebola has receded. The 2014-2015 West African outbreak has been brought under control not thanks to the deployment of successful treatment regimes, because there are none that are known to work. I participated recently in an international meeting of experts debating the ethical and methodological issues pertaining to trial designs for emerging infectious diseases like Ebola. It was both astounding and also immensely frustrating that to a large extent the controversies that exercised the minds of the delegates of this meeting exercised the minds of many an AIDS activist and clinical trials’ expert prior to the advent of highly active antiretroviral therapy, a good quarter of a century ago.[1] [2]Are placebo controls an ethically defensible methodological tool when patients face a terminal illness? Different alternative trial designs involving placebo controls, adaptive trial designs, and multi-stage approaches involving active controls were discussed during the meeting. The heated nature of some of these debates reminded me strongly of the passion that was on display during the early HIV trials. It turns out, despite decades of informed debate about these issues, a number of significant normative questions have not been settled.

A cluster of difficult ethical questions that engendered justifiably a lot of debate has to do with the use of placebo controls in trials involving patients facing a very high mortality risk (some in excess of 90%) and a fast-acting infection resulting in the death of these patient within 2-8 days after admission to a treatment centre. This scenario mirrors the sobering reality faced by a subset of Ebola Virus Disease patients. This issue was already highly contentious during the early HIV trials, and then patients and clinical investigators were faced with a virus that was nowhere near as fast-acting as the Ebola virus. The ethical conflict that arises here is this: We know that those randomized into the placebo arm face the same greater-than-90%-risk of death within a few days as those who receive the standard of care treatment. In some trial design the placebo control arm could be identical to the gold standard of (unsuccessful) clinical care provided in a particular clinical setting. Given that those who are randomized into the arm featuring the unregistered medical intervention might do better, or might do worse, or might do roughly as badly as those in the placebo control arm, the ethical question remains whether a trial design featuring a placebo control is ethically justifiable, given the almost certainty of imminent death faced by those randomized into the placebo arm. During the meeting I alluded to earlier a fairly contentious debate arose also over the question of whether trials producing less reliable results than placebo controlled trials might be acceptable under such circumstances.

What exacerbates the ethical challenges for those who undertake such trials is that their trial participants are arguably not true volunteers. Their – dying - trial participants are not given the opportunity to choose between participating in the placebo controlled randomized trial versus accessing the unregistered medical intervention on their own volition outside the trial process. It is perfectly conceivable that some patients might choose to participate in such trials in order to facilitate the development of a successful intervention capable of helping future patients like them. Or they might accept that there exists true clinical equipoise between the trial arms and they might be volunteering to be randomized under such circumstances. In the absence of alternative access routes to the unregistered medical intervention, we can never be certain that the patients agreeing to be randomized are not simply responding to what constitutes a coercive offer.

Clinical investigators colluding in this process, and arguably benefiting from it, are not absolved of their ethical responsibilities because they did not create the regulatory frameworks that gave rise to the problem. It is true that they did not create the regulatory framework under which they operate, but they undoubtedly benefit from its existence. We could respond to this kind of argument by pointing to the societal need for sound trial designs and the detrimental impact of permitting patients to access unregistered medical interventions outside the clinical trials’ system. The likely impact of permitting patients access, as a senior biostatistician attending the workshop rightly pointed out to me, would be a significant slowing-down in the trial recruitment process. Some trials might never be able to recruit sufficient patients, because most patients might be voting with their feet and opt to take their chances with the unregistered medical intervention. Surely that is not quite what is in the best interest of any society battling an emerging infectious disease such as Ebola. Does this justify coercing dying people into particular trial designs? I do not think so, but this is a contentious issue where reasonable, well-informed people can justifiably differ. A WHO panel looking at this question argued that while it would be ethically defensible to offer emergency access to unregistered medical interventions to Ebola patients, this should be subject to that emergency access not slowing down trial recruitment.[3] The panelists (not featuring a single expert or disease survivor from the affected countries) took a policy line here that mirrors US regulations. Other countries, including Canada and South Africa do not make this a threshold condition for emergency access. As it is with these sorts of panels, the advice it rendered on this controversial topic is not actually reasoned for, so policy makers and regulators as well as patient rights advocates aiming to balance the competing interests of access versus trial recruitment in a fair manner will be left wondering about the ethical reasons for this policy stance taken by the WHO panel, assuming there are any.

There are other ethical issues that arise in this context: Some experimental agents existed at the time only in insufficient quantities, for instance ZMapp, an unregistered medical intervention composed of monoclonal antibodies, was only available in very limited quantities. In light of this situation, is it acceptable to prioritize patients in comparable clinical circumstances who are willing to be randomized in a placebo controlled trial over patients clamoring for direct emergency access, given that the available quantities of this unregistered medical intervention would have been used up in the placebo controlled trial?

And here is another difficult question: While the AIDS activists of days gone by were highly educated about their disease and about the available unregistered medical interventions considered for expanded access programs, this is not quite the case with regard to the average West African Ebola patient. These patients were unlikely able to provide valid first person informed consent, because they were unable to demonstrate a reasonable person understanding of what was known about the unregistered medical intervention, about their options and so on and so forth. This is the case both because of educational limitations as well as disease progression. Are short-cuts to informed consent ethically justifiable under such circumstances? Given that time is of the essence and proxy consent might not be feasible due to family members being deceased or in a far-away village, are our informed consent requirements reasonable under such emergency circumstances?

The WHO panel suggests that evidence from nun-human primate experiments might be sufficient to justify offering a particular unregistered medical intervention for emergency access. Is that an ethically justifiable stance, given the high mortality rate and fast-acting nature of the infection?

Let me leave you with a final difficult question to ponder: Imagine you were running a medical NGO providing access to unregistered medical interventions to patients you care for in your emergency medical centre. By some fluke your unregistered medical intervention permits some of your patients to survive, but that survival comes at a high price, debilitating after-effects of the Ebola virus as well as of the unregistered medical intervention. Given concerns about your patients’ capacity to provide valid informed consent, should you accept responsibility for the patients’ future care and upkeep, given the lack of state infrastructure to assist these patients? If you accept responsibility for their care, say, by taking out an insurance package from some provider for them, you will expend a fair amount of donor monies on these patients (potentially for decades) that you cannot use to assist patients also facing life-threatening illnesses in other parts of the world. In other words, you face another ethical challenge, a resource allocation challenge. How should that medical NGO go about addressing this challenge?

---

[1] U Schuklenk, C Lowry. 2009. Terminal illness and access to Phase 1 experimental agents, surgeries and devices: reviewing the ethical arguments. British Medical Bulletin 89: 7-22.

[2] U Schuklenk, C Hogan. 1996. Patient access to experimental drugs and AIDS clinical trial designs: ethical issues. Cambridge Quarterly of Health Care Ethics 5: 400-409.

[3] WHO. 2014. Ethical considerations for the use of unregistered interventions for Ebola virus disease.  WHO: Geneva 2014: p. 5.

The Ebola response shows that we are learning nothing from history

Those of you who have followed the debates on AIDS clinical trial designs could be forgiven for thinking that humans must be unable to learn from history. I am not talking here about the current ongoing quarantine fiasco engulfing the USA. It's unworkable nonsense that has been condemned by pretty much every clinical expert under the sun. Panic is a bad guide for policy decision making. Perhaps next time the powers that are should try evidence based policy decision making. - Not terribly likely, unfortunately.

In any case, The Lancet has currently a debate going about the ethics of placebo controlled randomised clinical involving therapeutic and preventive experimental Ebola agents. The arguments pertinent to this have been - mostly - developed during the HIV/AIDS epidemic when it unfolded in the USA. Here is a review piece I had out a few years back on this subject. Remarkably, the arguments put forward in this context have not evolved at all since the late 1980s.

One the one hand you've those who are opposed to placebo controls in clinical trials involving post phase 1 experimental agents.

On the other hand you've the old battle axes from the US NIH bioethics department (Ezekiel Emanuel anyone?) reheating their undying support for benefit sharing in international health research and, of course, their undying support for placebo controlled randomised trials (presumably of post phase 1 experimental agents, but that ain't quite clear). It's the same arguments in favour of placebo controlled randomised trials involving catastrophically ill patients that are being regurgitated in the Ebola trials' debates. Its proponents write 'randomisation and placebo controls are the best means to control for confounding factors and determine whether interventions work or whether patients have recovered by chance.' Or, in another reheating of the same argument, writes David Shaw in a letter to the Lancet, 'but the best way to generate such data is in a randomised controlled trial'. It is entirely unclear what's meant by 'best' here. 'Best' presumably means a non-existent trial population on a planet where desperate patients and their loved ones have not killed health care workers and burned down treatment facilities. It's a planet, let's call it Emanuel-Shaw-landia, where catastrophically ill patients will be happily herded into placebo controlled trials. They are fully driven by a planetary sense of duty, no doubt. - On our planet, we know that such patients will do anything to subvert clinical trial designs they deem unfair. This is where our bioethicists might have taken on board lessons from the heydays of HIV/AIDS activism.  Catastrophically ill patients rightly consider a 50:50 chance of getting a placebo a lousy deal. They will go to great length to share the active agents, thereby subverting the placebo controlled trial design, dosing regimes and whatnot. Remarkably none of the historical evidence we have on this count featured in any of the Ebola publications. Zilch. Why bother? Academic memories are truly remarkably short-term these days. Learning from experiences from a time when desperate, catastrophically ill patients responded to coercive offers involving places in purportedly ethical, placebo controlled trial designs by cheating on such large scale as to render a whole lot of research that occurred in those years useless? Na.  Let's just stick to our story, the gold standard of placebo controlled randomisation, reality be damned.

In Western countries we have myriads of access schemes designed to ensure that whoever signs up today to participate in a placebo controlled randomised trial does so as a true volunteer, not a desperate patient who has run out of options in terms of accessing experimental agents legally by other means. One of those lessons we learned from HIV/AIDS. No word on any of this in current Ebola papers.

There's another issue that bothers me a bit about this debate. Much is made by everyone of the high mortality rate among patients infected with the virus. Things seem more complicated.

A recent paper in The Lancet reports this, 'evidence suggests that many Ebola infections are asymptomatic,1, 2 a factor overlooked by recent outbreak summaries and projections.3 Particularly, results from one post-Ebola outbreak serosurvey1 showed that 71% of seropositive individuals did not have the disease; another study2 reported that 46% of asymptomatic close contacts of patients with Ebola were seropositive. Although asymptomatic infections are unlikely to be infectious,2 they might confer protective immunity and thus have important epidemiological consequences.'

Much of the hype driving the reheating of the placebo debate is a result of projections likely overstating the future spread of the disease as well as the mortality rate associated with an infection. I wonder whether we will look back at short-cuts to informed consent, and the fast-tracking of ever more experimental agents as a terrible mistake. The little bit of data that we have from the USA suggests that early detection plus good clinical care can bring down the mortality very significantly. In fact, nobody who was treated competently and in a timely fashion died. This sheds a different light on trial justifications flagging the high mortality rate. Trial justifications flagging the high mortality rate that do not take into account that this rate is likely an artificial (ie human created, as opposed to disease created) result of the lack of efficient health care delivery essentially propose that economic reasons are good ethical justifications for clinical trial designs. After all, the mortality rates in question would have been caused by the economic conditions leading to the lack of timely and efficient clinical care for those who actually get sick from the Ebola virus. Perhaps bioethicists should focus again international justice issues. Not unlike in the days of HIV/AIDS, it's about prevention, prevention, prevention, treatment, treatment, treatment. Good clinical care will permit us to dramatically reduce the mortality rate among those getting seriously sick. Perhaps that's where our focus ought to be. Not on a regurgitation of the placebo debates of years-gone-by.

Or, if you really must, take note of the history of this debate. I thought that's a minimum requirement in the academy.

(Hopefully) the UK's last homeopathic 'hospital' to close its doors soon

The Observer newspaper reports that the UK's last homeopathic 'hospital' might be about to be closed because the NHS's Primary Care Trusts will refuse to reimburse patients 'treated' there. That's pretty good news. I doubt anyone other than avowed homeopathy fans such as the Queen (yep, HRM etc etc) and her son Prince Charles will miss the outfit. Surely, as a matter of public policy it cannot be that tax monies are being spend on 'treatments' of truly unknown benefit. There is no (I am not exaggerating here) independent scientific evidence supporting the idea that homeopathy works. So, if patients are keen on placebos, we should be able to find a slightly cheaper way than a costly 'hospital' requiring millions of tax monies every year for its up-keep. It is surely unethical to waste health resources that could be poured into buying and distributing medicine proven to work.

Beware: AIDS microbicide makes you more vulnerable to HIV infection

An HIV prevention trial was halted in various Southern African countries and in India. The BBC reports that the investigators discovered that instead of reducing the risk of HIV infection the microbicide in question actually increased the risk. In other words, more participants (it being a microbicide, all women, of course) became infected in the active agent arm then in the placebo controlled arm of the trial.

There are several interesting questions with regard to this trial, none of which was flagged in the BBC news report:

1. The women who participated in the trial, by definition were HIV negative when they enrolled. They became infected (almost certainly due to unsafe sex) during the trial. The question is whether these infections constitute a trial related injury that ought to be subject to compensation. Many of them might well have had a therapeutic misconception, that is they might have thought that the stuff they got in the clinic gave them some (a lot - complete?) protection against HIV.
   
2. Indeed, one wonders whether these women will be provided with access to HAART when this could be clinically beneficial to them.
3. Sexy as the idea of an HIV microbicide is, none of the trials undertaken so far have led anywhere. As a non-expert I wonder whether this is a dead-end type concept that is never going to work. You might say, reasonably so, that only further research will tell. The trouble is that we will only find out when women become infected during a trial. That's arguably a pretty high price to pay, even if the women were volunteers and even if they gave truly informed first person consent.