Equipoise

It has been argued that a core element for the ethical and epistemological justification of running a RCT is that a situation of clinical equipoise obtains, where clinical equipoise is defined as ‘a state of genuine uncertainty on the part of the expert community regarding the comparative therapeutic merits of each arm in a RCT’14 (p. 465). It was initially suggested that physicians who recruit to trials should be in ‘personal equipoise’, that is, they should be indifferent to the therapeutic value of the experimental and control treatments.15 16 Others have suggested that both clinical and personal equipoise are necessary for a truly unbiased RCT. However, discussions have primarily focused on ‘clinical equipoise’ as a state of genuine disagreement within a community of bona fide expert practitioners.17 There is a large and complex critical literature on equipoise, and it is, therefore, important to clarify that the argument in this section does not assume the strong position—supported by some commentators, including the authors of this paper—that clinical equipoise is always necessary for a trial to be ethical. However, we also do not assume that a distinction can be made between research ethics and the ethics of care such that equipoise is no longer relevant to the justification of randomisation.18 Rather, our analysis does, for the sake of argument, assume a weaker claim that, in situations where the equipoise position is uncertain, there is a need for a closer scrutiny of: (1) how randomisation to a treatment known or thought by the expert community to be inferior to the established standard of care can be reconciled with the physician’s duty to protect and promote the best interests of the patient19; and (2) whether the research-related risk created by the trial is acceptable.20 To clarify, we are not endorsing the view that this weaker claim should over-ride the claim for clinical equipoise, that is, we are not rejecting clinical equipoise in favour of closer scrutiny of research-related risk and the conditions by which randomisation can be reconciled with a duty of care. Rather, we aim to show that, even on this weaker account of the necessity of equipoise, the inclusion of molnupiravir in PANORAMIC raises very significant ethical questions. The simplest way of explicating the notion of ‘closer scrutiny’ is by drawing a parallel between: (1) standard research-related risks (ie, risks specifically caused by research procedures that are additional to the risks caused by the patient’s condition and the interventions conducted under the auspices of usual care); and (2) what could be called ‘randomisation risk’ (ie, the risk of being randomised to an arm that is known or strongly suspected to be less effective than the clinically prescribed treatment the patient could have received if they declined trial participation (see also Fully informed consent). It is generally accepted that participants can provide informed consent to research-related risk, even though the questions of whether there should be an upper threshold for such risks and how that threshold should be defined are hotly disputed. ‘Closer scrutiny’ could, therefore, be understood as paying specific attention to the risk inherent in being allocated to a known or suspected inferior arm and ensuring that: (1) this randomisation risk falls below the threshold for consentable risk; and (2) that potential participants are fully informed about that risk.

To what extent and for what indications can clinical equipoise be said to exist in relation to treatment with molnupiravir after the publication of the results of the MOVe-OUT trial, the conditional marketing authorisation provided by the MHRA and the drug being made available for prescription by the health system in the jurisdiction in which the platform trial is conducted (in this case the NHS)?

It might be claimed that the decisions of the MHRA and the NHS directly indicate that clinical equipoise has been disrupted. This would, however, be a mistake. These official decisions are, in themselves, only indirect evidence of clinical equipoise. They produce no new evidence as to the efficacy of molnupiravir but merely indicate that the relevant bodies find the existing evidence at a specific moment of time sufficiently convincing to: (1) allow the drug to be prescribed to patients (MHRA); and (2) make the drug available if they meet certain requirements (NHS). It could still be possible that the clinical expert community believed molnupiravir to be of no value. There are drugs that are currently available for prescription where the clinical equipoise position suggests that they have very little or no effectiveness. It is also noteworthy that both decisions allow for the extension of the use of molnupiravir beyond those uses for which direct evidence was provided by the MOVe-OUT trial. Both decisions, for instance, make no distinction between vaccinated and unvaccinated patients, even though the trial only recruited unvaccinated patients. In addition, the NHS criteria include adolescents between 12 and 17 years old, a group that was excluded from the trial, as well as those adults with underlying chronic health conditions not studied in the MOVe-OUT trial unless they happened to have been included as a result of being over the age of 60 years. Even though decisions made by regulators or by healthcare systems are not direct evidence of equipoise in themselves, they may influence clinical equipoise. For instance, since some members of the expert community may be aware that their own view is only based on partial knowledge of the available evidence, they may perceive these decisions as being based on a search for and synthesis of all the available evidence.

It seems reasonable to assume that the original position in the relevant clinical community was one of equipoise in relation to molnupiravir and that expert members of the community only started to form a view on the efficacy of the intervention as research evidence gradually became available, which pointed to molnupiravir as a potentially efficacious intervention. From an epistemological point of view, the pre-existing equipoise should be more disturbed as more and better evidence becomes available and should be more disturbed in relation to those questions where the evidence is most directly relevant.14 For instance, this follows if we model the expert clinical community as consisting of a population of decision makers who update their knowledge and, therefore, decision making following a Bayesian approach.21 The members of the community will access different information and their updating will, in most cases, not be perfectly Bayesian. We should, therefore, expect a distribution of views to arise. However, this does not undermine the epistemological point that, as more evidence becomes available, we should expect a predictable evolution of the equipoise positions.

At the point in time where PANORAMIC started recruitment to the molnupiravir intervention arm, the best evidence for the efficacy of molnupiravir came from the MOVe-OUT trial, which had only included unvaccinated patients who were over the age of 60 years or who had at least one of the specified chronic health conditions (table 1). As noted, this evidence was crucial for both the MHRA and the NHS decisions relating to molnupiravir. We should, therefore, expect the equipoise position in the community to have moved the furthest in relation to the prescription of the drug to unvaccinated patients with the specified chronic health conditions detailed in the MOVe-OUT selection criteria who are at high risk of developing severe COVID-19. It is possible—perhaps even likely—that the community had moved away from equipoise towards a position whereby molnupiravir treatment was seen as clinically indicated for this group of patients. This would not require all members of the expert community to have formed this view but would require it to be the majority view. By contrast, given that the MOVe-OUT trial did not include vaccinated individuals, clinical equipoise regarding the efficacy of molnupiravir should have, in principle, obtained for this group of patients. Nevertheless, in practice, as we have already observed, the decisions to allow the drug to be prescribed to patients and to make it available if patients meet certain requirements may have contributed to a change to the equipoise position for vaccinated patients. Therefore, there is no guarantee of a state of clinical equipoise even for vaccinated patients. Given that an empirical analysis of the state of clinical equipoise in relation to vaccinated patients is beyond the scope of this paper, the following explicitly ethical discussion, for principled reasons, focuses on the equipoise position as it relates to specifically unvaccinated patients who are recruited to the PANORAMIC trial.

If the evaluation of the state of equipoise in relation to unvaccinated patients at the relevant time is correct, or even merely plausible, then it strongly raises the question of whether unvaccinated patients who are over the age of 60 years or who have at least one of the chronic health conditions specified in the MOVe-OUT inclusion criteria should be recruited to the PANORAMIC trial at all. The absence of equipoise means that most experts believe that treatment would be to the benefit of these patients.ii However, it is important to note that this particular expert belief is fully compatible with the belief, held by the same experts at the same time, that it would be desirable to have more and better evidence on the efficacy of molnupiravir in this group of patients. In less urgent, non-pandemic contexts, ‘more and better evidence’ would likely have been generated through additional phase 3 trials with slightly different patient groups before molnupiravir received marketing authorisation. However, as we acknowledged in the introduction, because the regulatory processes were (rightly) sped up to enhance access to clinical interventions in response to the pandemic, less evidence than usual was available at the point when the product became available for clinical prescription. This creates a situation where further evidence on the efficacy of molnupiravir for high-risk, unvaccinated patients that fall within the inclusion criteria of MOVe-OUT can no longer be ethically produced through placebo or ‘usual care’ controlled trials. If we take into account this equipoise situation on its own, then there is a principled reason to question the justifiability and permissibility of the randomisation risk for patients in this group who are recruited to the PANORAMIC trial and, as we shall discuss in the following section, to claim that PANORAMIC’s duty of care to unvaccinated patients is not being adequately met.

Finally, it is worth acknowledging that the PANORAMIC trial includes unvaccinated participants with certain risk factors not covered by the MOVe-OUT trial. For these groups of patients, it is more difficult to ascertain with any certainty whether clinical equipoise existed, and whether it was already significantly disrupted when PANORAMIC started recruitment.

Standard of care

What is the standard of care or the best proven intervention at the point in time when PANORAMIC started recruitment to the molnupiravir intervention arm?

This question can be interpreted as either an epistemological question or a pragmatic question. The epistemological question about the standard of care is roughly equivalent to the question of the existence of equipoise in the whole clinical community combined with the intervention in question having successfully negotiated the clinical translation gap. We have discussed the epistemological issues involved in interpreting the state of evidence following the MOVe-OUT trial in the preceding section of the paper, so we will primarily consider the pragmatic question here. In the research ethics literature, there is a considerable body of argument that interprets the standard of care as roughly the treatments and interventions that patients in a particular healthcare system can routinely access, even if those treatments and interventions are known not to be the best available.23–25 This understanding is, for instance, the basis for the argument that if interventions that are routinely available in high-income countries are not available to potential participants in low-income countries, then it is acceptable to run placebo-controlled trials with those interventions in low-income countries. If, for the sake of argument, we accept this interpretation of the standard of care, then the standard of care in England is minimally defined by the interventions that are provided by the NHS. In the present context, this means that molnupiravir treatment for anyone who falls within the NHS criteria (table 1) is the standard of care or the best proven treatment.

This creates a problem for PANORAMIC given that the investigators declared in the protocol that they would comply with the principles of the Declaration of Helsinki10 (p. 28). According to Article 33 of the Declaration:

The benefits, risks, burdens and effectiveness of a new intervention must be tested against those of the best proven intervention(s), except in the following circumstances:

Where no proven intervention exists, the use of placebo, or no intervention, is acceptable; or

Where for compelling and scientifically sound methodological reasons the use of any intervention less effective than the best proven one, the use of placebo, or no intervention is necessary to determine the efficacy or safety of an intervention and the patients who receive any intervention less effective than the best proven one, placebo, or no intervention will not be subject to additional risks of serious or irreversible harm as a result of not receiving the best proven intervention.

Extreme care must be taken to avoid abuse of this option.22

It is important to note that this research ethics issue, which is based on a pragmatic understanding of standard of care as defined by availability, primarily arises because recruitment to the PANORAMIC trial takes place at a time when the intervention, molnupiravir is, in principle, already available to potential participants who fulfil the NHS requirements. Indeed, given that there is an element of general practitioner discretion allowed by the NHS in determining whether a patient is eligible to access molnupiravir (on the basis that they are deemed to be ‘at highest risk of getting seriously ill from COVID-19’),12 then, in practice, all potential PANORAMIC participants could access the standard of care treatment through the NHS.

It might be argued on the basis of the equipoise considerations above that there are ‘compelling and scientifically sound methodological reasons’ for generating further evidence by including molnupiravir in the interventions evaluated by PANORAMIC and that PANORAMIC is, therefore, compliant with the Declaration of Helsinki in the sense that it falls within the exception defined in Article 33. The compelling reason could either be the need for more evidence in general, or the need for evidence in relation to groups not included in MOVe-OUT but included in PANORAMIC, or a combination of the two. This argument seems to be undermined by the fact that the Article 33 exception only applies if patients in the non-intervention arm ‘will not be subject to additional risks of serious or irreversible harm’. In the absence of other evidence, we should accept that the MOVe-OUT estimate of efficacy applies directly to those participants in PANORAMIC who fall within the MOVe-OUT selection criteria. In addition, we should accept that for other groups included in PANORAMIC, we have little evidence to predict whether they are likely to have a higher or lower efficacy and that our best estimate is, therefore, the MOVe-OUT estimate for them as well. This entails that being randomised to the non-intervention arm (when they could access molnupiravir through the NHS) subjects the participants to an additional risk of severe COVID-19 with hospitalisation or death of, on average, three percentage points, which seems like a serious harm. This would entail that PANORAMIC contravenes the Declaration of Helsinki.

The issue identified here may be further exacerbated since there is a possible argument that a more expansive account of standard of care should be applied and that the standard of care should be defined as any intervention that has received conditional marketing authorisation by the competent national regulator. On this understanding, molnupiravir would count as the standard of care for everyone in the UK who falls within the PANORAMIC selection criteria, including those whose increased risk is only caused by having age ≥50 years as a risk factor.

Fully informed consent

As outlined in the introduction, molnupiravir is available in England to certain patient groups on the basis of clinical indication, and these patient groups overlap substantially or, in practice, completely with the group defined by the inclusion and exclusion criteria of PANORAMIC (table 1). This raises the questions of: (1) whether potential participants in the PANORAMIC study should be informed by the investigators in the participant information sheet or reminded about the clinical availability of molnupiravir as part of the recruitment process to the trial; and (2) whether the validity of consent is affected if they are not informed or reminded.

If they consent to participation in the trial, then the patients who are both eligible for recruitment to PANORAMIC and eligible to get a clinical prescription of molnupiravir run a 50% risk of not getting molnupiravir by being allocated to the ‘Usual Care’ only arm. If they do not consent to participate, then they can access molnupiravir if either: (A) they remember that they are eligible because they have been preidentified as being in the high-risk group and contacted by the NHS or (B) their physician knows/realises that they are eligible based on their risk profile. Therefore, potential participants in this overlap group have a choice between either accessing molnupiravir through the NHS (with certainty) or participating in the trial and having a 50% risk of not receiving treatment with molnupiravir.

In terms of the 50% risk of not getting molnupiravir by being allocated to the ‘Usual Care’ only arm, it might be argued that potential participants should be allowed to make the choice and consent to trial participation because this risk is of the type and magnitude that falls below whatever threshold we set for allowable research risk. Let us accept this for the sake of argument. It still leaves the problem that participants can only make this choice and voluntarily consent to assume the risk if they have knowledge regarding the choices that are available to them. A potential participant who is not aware at the time of consent that they can access the treatment outside of the trial lacks information that one may reasonably deem to be essential for making the informed choice whether to consent or not. In order to protect and promote that patient’s best interests and autonomous decision making, or to recognise the patient as a competent deliberator of equal standing,26 this also seems to be the kind of information that a physician has a professional and ethical duty to convey to a patient. Finally, it seems to be required by the first part of Article 31 of the Declaration of Helsinki (‘The physician must fully inform the patient which aspects of their care are related to the research’) as well as by the first part of Article 26:

In medical research involving human subjects capable of giving informed consent, each potential subject must be adequately informed of the aims, methods, sources of funding, any possible conflicts of interest, institutional affiliations of the researcher, the anticipated benefits and potential risks of the study and the discomfort it may entail, post-study provisions and any other relevant aspects of the study22 (our emphasis).

In the literature that argues for the relaxation of research ethics regulation and principles during the COVID-19 pandemic, there are a number of authors who aim to justify the claim that potential participants should be allowed to consent to greater research-related risks than is currently the case (eg, in relation to SARS-CoV-2 challenge studies).8 9 We are, however, not aware of any extant arguments for allowing researchers to not inform potential participants about known, significant risks prior to asking for their consent to participation.

Requiring potential participants to be informed about the clinical availability of molnupiravir complicates the information and consent process, a matter that is often simplified in large, platform trials to facilitate rapid recruitment. However, as we have argued, there are strong reasons to suggest that if patients are not informed of the choices that are available to them and the research-related risks, then they are unable to, or have not provided ethically valid, informed consent.