A conversation on immunity passports for Covid-19

Joe Brew: I think that the “immunity passport” concept is worth exploring. It’s a potentially useful tool. The main technical arguments against it (imperfect tests, doubts about the degree and duration of post-infection immunity, difficult to operationalize) seem surmountable. And the main ethical arguments against it (potential discrimination) seem relatively minor: I don’t see it as “discriminatory” to allow non-infectious people to behave differently than infectious people. In fact, we already do this (quarantines, lockdowns, etc.). In other words, differential policies based on epidemiological risk are the norm, not the exception (and should be!). The immunity passport would simply help us to apply differential policies in a more precise way. And the more precise we can be in terms of risk of being infected or infecting others, the better. We are willing to apply differential rules based on age, where you live, and travel history, because these factors correlate (imperfectly) with risk of infection. So, why should we willfully ignore the presence of antibodies?

Carlos Chaccour: I think the main concern is certainty of knowledge relative to immunity. There is robust evidence that justify a differential protective approach for the most vulnerable i.e. protect the elderly and those with risk factors for severe disease. There is however, limited evolving knowledge on how protective post-infectious immunity is in terms of duration and magnitude, both against disease but also against the virus itself (i.e. we know that SARS-CoV-2 infection protects against rechallenge in rhesus macaques). Note the exact virus at the same dose was used for the re-challenge, in other words we do not know whether these macaques are protected against a challenge with a higher viral load or against a challenge with SARS-CoV-2 with a mutated spike protein. Another key knowledge gap is whether convalescent individuals could have transient and asymptomatic viral carriage in the respiratory tract making them infectious. I see potential value in differential protective measures for those at higher risk (an equal approach would be unfair here) but these knowledge gaps pose some serious risks. A good question is, how much risk are we willing to take?

Joe: So there are really two questions: First, the technical and scientific question of whether infection confers immunity, and the extent to which antibody tests are accurate enough to be useful. And second, the ethical question of whether we should operationalize post-infection immunity. You say that the first question is the “main concern”, but I’m not so sure: to many, the ethical questions are more important (i.e. this letter in Nature). Regardless of what the “main concern” is, the second question (on ethics) is only relevant if the answer to the first question (does infection confer immunity) is “yes”. So, if you agree, let’s start with the technical and scientific question of the degree and duration of immunity (ie, whether immunity is operationalizable) before going into the ethical implications. We have evidence (albeit limited) that previous infection reduces one’s future susceptibility (and by extension, infectiousness), correct? You know more than me about this - what do we know, and what don’t we know?

Carlos: I agree that breaking this into two separate questions would make this more “digestible” but it might be an artificial separation; after all, technical certainty on efficacy (or at least an equipose in case of a clinical trial) should support actions for those actions to be ethical. Otherwise implementation becomes pseudo-experimental, which might be acceptable in some cases but should be recognized as that.

Joe: So, we know some things, but not everything. From my point of view, this makes the matter of immunity very similar to other Covid-19 response methods, like hand-washing, social-distancing, mask-wearing, school-closing, etc. But for some reason, when it comes to antibodies, it seems that the scientific community’s communication with the public has really emphasized the things we don’t know, instead of emphasizing the things we do know. Why is that? Why such a hesitance to even suggest that antibodies should be considered in terms of behavior and policy-making? I don’t think that the hesitance is due to scientific uncertainty, because scientists have been _very_ quick to recommend certain policies which haven’t been “proven” via experiment to be effective. For example, mask-wearing _probably_ reduces one’s infectiousness, but I don’t think it has been demonstrated through RCT, right? Is it fair to say that post-infection immunity is similar? And if both (a) mask-wearing and (b) having antibodies _probably_ reduce infectiousness, why are we so quick to legislate around the former, and so slow to legislate around the latter? I think the reason is ethics, not scientific uncertainty. Do you agree? Or have I simply stretched the mask-antibodies analogy too far?

Carlos: I definitely see your point, but I think recommendations are not based solely on the probability of something being efficacious but on a balance between potential protection and other factors such as economic and social cost. Masks have an unproven efficacy but they entail low economic cost and relatively low social cost through personal discomfort. Confinement has better evidence to support it but it also has enormous economic and social costs. Immunity passports have several uncertainties regarding efficacy but carry the risk of immunoprivilege or immunodiscrimination which has led in the past to people having difficulties obtaining life/health insurance, obtaining or even retaining a job. This was particularly visible in New Orleans while yellow fever was endemic there. Beyond the high societal cost of these unintended effects there is also the risk of a high demand for “immunity capital” which can easily lead to "corona-parties". Would this offset any benefit obtained from the immune passports themselves? That is a question worth pondering.

Joe: As I see it, you’re expressing two main worries, both of them legitimate. First, the issue of discrimination (ie, “immunoprivilege”); second, the risk that discrimination might generate perverse incentives to self-infect. These seem like reasonable worries, albeit without a firm evidence base (I might be wrong, but nobody has actually done any research, theoretical or practical, on whether people would actually intentionally infect themselves so as to get immunity passports in the case of Covid-19). Regardless, I think there is a misguided underlying premise in both of these worries, and it goes like this: since restricting activities for only some people would be “discriminatory”, let’s instead restrict activities for all people. In other words, as a response to this pandemic, many are treating basic rights (the right to move around as one pleases, for example) as a “privilege”. Moving around is not a privilege; it’s a right. And our goal as a society should not be to reduce the number of people with that basic right in the name of some false “equality”, but rather to minimize the number of people with restricted rights. In the case of a second wave of Covid-19, would it be better to restrict the rights of 100% of the population (in the name of “fairness”) or 90% of the population?

Carlos: I think we agree that equal treatment provided different conditions is far from being just. That is, justice demands equity, not equality. A blanket approach given different levels of immunity would be wrong. But we are back at having to balance three aspects: (a) risks, (b) benefits and (c) unknowns. I also agree that restricting the rights of those that are immune solely on theoretical grounds would be wrong, provided they are not endangering themselves or others (see knowledge gaps above). I recognize that basic rights are being treated as privileges by many policy makers right now, but here again those making the decisions are balancing the right of movement with the risk, cost and consequences of contagion. If our main worry is avoiding the restriction of rights then it is our duty to protect the right to live, which is the right _par excellence_. I would tend towards erring on the safer side here.

Joe: I like your description of the three components to balance: risks, benefits, and unknowns. It’s clear that any policy approach should take into account all three of these components. And since all of them are constantly changing, our position on the matter (ie, the solution we consider “optimal”) should be subject to changing over time, too. I’ve heard a lot about the “risks” of discrimination, perverse incentives, insufficient accuracy, false confidence, moral hazard, etc. But what about the risk of not issuing immunity passports? If the State does not issue them, do you worry that private initiatives will? This is my worry. In the case of a second lockdown, I think that there will be a natural “demand” among the previously infected to show that they should not be subject to disease control measures. This demand could lead to a “supply” from many non-State actors. In other words, there could be a splintering of documentation showing “immunity” - from labs, from doctors, from companies, etc. So, by doing nothing, the State is delegating the role of the immunity passport to the private sector, which will do it more poorly, with less “verifiability”, and with less checks on privacy and ethics. I can imagine a scenario in late 2020, during a second-wave lockdown, in which an elderly man goes out for his daily walk, is stopped by police, and says “I’m immune, here’s my hospital discharge documentation, now leave me alone”. And the police, in this case, might not know what to do…

Carlos: I agree 100% on the need to revise decisions as new evidence emerges or the situation changes. In the case of COVID-19 the revision rate is vertiginous. So settling the “issue” of immune passports forever based on an assessment done at a concrete point in time is not good practice. I think you bring up valid points regarding non-state and less-transparent actors taking advantage of a surge in demand during a second wave in autumn-winter.

Joe: Just as you’ve pointed out the potential advantages, for the sake of balance, allow me to point out why I see as some potential disadvantages in operationalizing immunity passports: