Protection of Conscience Project
Protection of Conscience Project
Service, not Servitude

Service, not Servitude

The ethical minefield of COVID-19 vaccination: Informed consent and the obligations of doctors

Australian Broadcasting Corporation
10 May 2021
Reproduced with permission

Margaret Somerville*

The COVID-19 pandemic has raised a multitude of complex ethical issues — and new ones present themselves daily. These issues, including those related to vaccination, arise at four levels:

  • micro or individual (for example, when a doctor vaccinates a patient);
  • meso or institutional (regarding, for instance, a hospital’s or aged care residence’s policy on vaccinating staff);
  • macro or societal ( a government’s decisions or public health regulations governing distribution of vaccines and access to vaccination); and
  • mega or global (such as a nation’s obligations to provide vaccines to those in developing countries, which are without vaccines).
  • In many COVID-19 related decision-making situations at each of these levels, decision makers face what is called in bioethics a “world of competing sorrows” — that is, decision making in which there is no “no harm” option, but in which, instead, they must choose to whom harm will be allocated. The ethical difficulties are exacerbated when the harms and benefits do not accrue to the same people or, at least, not in equal measure. A striking example of such a situation at the macro or societal level would be the use of “lockdowns”, when the choice is between protecting public health and inflicting serious economic harm.

    What I want to focus on here is a particular micro- level issue: that of a healthcare professional’s obligation to obtain a person’s informed consent to COVID-19 vaccination.

    Failure to obtain an informed consent to, or an informed refusal of, medical treatment — which includes vaccination — is medical negligence (medical malpractice). Informed consent to or refusal of medical treatment has three requirements: competence, information, and voluntariness. There is a wealth of research on what is needed to establish each element, but here is a brief summary.


    The law presumes that all adults are competent, until proven otherwise. Children are not presumed to be competent, but those with the required mental capacity can be found competent to give an informed consent. Competence requires that the person must be able to understand the information they are required to be given in order to obtain their informed consent or refusal, and be able to appreciate the consequences of the decision they make.

    Medical treatment decisions for incompetent persons must be made by a competent “substitute decision maker”, who implements what they believe would be the incompetent person’s preferences and acts in the “best interests” of the incompetent person.


    All the information that would be “material” to a reasonable person in the same circumstances as the person seeking vaccination must be disclosed, along with any additional information the healthcare professional knows or ought to know the particular person would want to know, including honest answers to all of the person’s questions. Risks of death are always regarded as “material risks”, even if of very low probability. This explains the temporary suspension of using the AstraZeneca vaccine in Australia in order to add disclosure of the rare risk of fatal blood clots (cerebral venous thrombosis) to the consent form signed by the person.

    To obtain an informed consent to, or informed refusal of, any given treatment, the person must be offered all medically indicated treatments and the risks, harms, benefits, and potential benefits of each and of refusing all treatment must be disclosed. Failure to do so would negate informed consent to any vaccine given, even, possibly, if the required disclosure regarding that vaccine were made.

    Likewise, an informed refusal of vaccination would be negated by the inadequate disclosure of the same information — in particular, of the benefits of vaccination and of the fact that the risks associated with COVID-19 infection far outweigh the risks of vaccination.

    This extensive disclosure requirement can raise problems. What if a particular person is not told about the least risky COVID-19 vaccine for them and they develop serious complications from the vaccine administered that was not the least risky one? Have they given their informed consent to the vaccine given?

    If the least risky vaccine is not available — subject to its unavailability not being the fault of some authority with responsibility for providing vaccines — there is probably no ethical or legal breach for not disclosing its existence. That said, it is a controversial topic in medical ethics, in general, whether treatments, which are indicated as treatment or prophylaxis for the patient, but are not available, must be disclosed in order to obtain informed consent to the treatment provided. A further complexity is what if the least risky vaccine will become available, but not, let us say, for six months, as is true of sufficient quantities of the Pfizer vaccine? Does everyone at least have to be told of the Pfizer vaccine and its future planned availability in order to obtain their informed consent to the AstraZeneca vaccine at present?

    What if, however, the vaccine is available, but in limited quantities, and by government or bureaucratic decree is not being offered to the group of people to which the person belongs? Keep in mind that a doctor has a primary obligation of personal care to each individual patient. This means they must always put that patient’s “best interests” first — that is, before their own interests and those of other people. Let us take an example.

    Both AstraZeneca and Pfizer vaccines are being administered to Australians, but there are limited quantities of the Pfizer vaccine and, with certain exceptions, it is not being offered to people fifty years of age or over. What if someone over fifty could be at more risk of blood clots than the general over-fifty population? For example, what are a doctor’s obligations to a woman patient aged fifty-five years, who is taking post-menopausal hormone therapy, is diabetic, and has previously suffered a deep vein thrombosis (DVT)? Should the doctor offer her the AstraZeneca vaccine?

    A doctor’s ethical and legal obligations

    The first question is what are the doctor’s ethical and legal obligations to this particular patient? Just because the government has pronounced a policy or guidelines does not mean that a doctor following those guidelines is acting ethically or even legally. To avoid medical negligence (malpractice), the doctor must act within the parameters of what a reasonably careful and competent doctor would and would not do in the same circumstances — including with respect to the range of treatments offered and obtaining informed consent to the treatment chosen by the patient. That can require taking a different path from that articulated in the guidelines.

    The second question is, if this patient is offered the AstraZeneca vaccine, must she also be offered the Pfizer vaccine, assuming that it could be obtained, in order to obtain her informed consent to accepting the AstraZeneca vaccine? In other words, if she accepts the AstraZeneca vaccine without being offered the Pfizer vaccine, is her consent to the AstraZeneca vaccine ethically and legally valid? This question leads beyond information requirements for informed consent to the third requirement for an ethically and legally valid consent — that is, voluntariness and its requirements that the person’s consent not be tainted with coercion, duress or undue influence (which will be discussed shortly).

    Yet another question is, if the Pfizer vaccine is available (whether or not in limited quantities), and Australia has a large stockpile of the AstraZeneca vaccine, may the government still limit the availability of the Pfizer vaccine to those under fifty years of age? It is very difficult, perhaps almost impossible, to argue this could be justified. That difficulty is augmented by the fact that recent research found that mRNA based vaccines (Pfizer is one; AstraZeneca is not) seem to have additional benefits in that they are effective against at least some variants of the COVID-19 virus and possibly many of them.

    Disclosure and perceptions of risks

    The seriousness and probability of risks is always relevant to whether they need to be disclosed. The more serious and the more probable a risk, the more likely it is that it must be disclosed. The difference between knowing and appreciating the risk is also relevant. Many people have difficulty understanding probability — namely, that statistics are true for a group, but not for any one person, which can be a barrier to understanding the risk and appreciating the likely consequences of running or refusing to run it.

    In addition, the comparative or relative risk of the various options regarding vaccination is a relevant fact that should be disclosed. The risk of developing blood clots is, not only, a risk of vaccination, but also a risk inherent in being infected with COVID-19. One in four people with COVID-19 admitted to intensive care units develop blood clots, and around 25 per cent of them die. In an as yet unpublished study from Oxford University, researchers found that COVID-19 infection is up to ten times more likely to cause blood clots than vaccination with the Pfizer or Moderna mRNA vaccines. The AstraZeneca vaccine is not an mRNA vaccine, but it too is far less likely to cause clots than infection with COVID-19.

    Factors other than seriousness and probability of risk are also relevant in people’s decision making regarding whether to run a particular risk. People can view risks that appear to result from chance very differently from those resulting from a choice to impose the risk made by an identified decision maker. They are much more tolerant of risks seemingly arising by chance than by another person’s or institution’s choice. In short, they can have very different reactions to harm seen as resulting from “an act of God” as compared with harm resulting from “an act of a healthcare professional”.

    As I’ve already noted, a person who refuses vaccination with the AstraZeneca vaccine runs a greater risk of infection with COVID-19 and, if infected, of serious complications, than if they had accepted vaccination. However, if this risk of refusing AstraZeneca vaccination materialises, it is likely to be seen as having arisen from chance, whereas blood clots caused by the vaccine are likely to be seen as caused by an identified agent. This could be the doctor or nurse who gave the jab, or the government, which allocated the probably less risky Pfizer vaccine to others, or all of these decision makers and actors. Ethical responsibility is not like a football (you no longer have it when you throw it to someone else), it is more like a cake (everyone can have a slice).

    Risks have a dread factor: the more familiar the risk, the lower the dread factor; and the more unfamiliar the risk, the higher the dread factor; the higher the dread factor, the more likely the person is to refuse to run the risk. An example is the fact that people take the risks of being killed in a car accident every day, but they will refuse to fly, even though the risk of death from a plane crash is far lower than that incurred in travel by car. A plane crash has a much higher dread factor than a car crash.

    Risks also have a time factor: the more immediate the materialisation of the risk could be, the less tolerance there is for chancing it; the further away the consequences, the less weight the risk is given. The long-term risk of developing lung cancer from smoking is a good example of the latter. COVID-19 vaccination has an immediate risk of medical complications; the risks from contracting COVID-19 are possibly not as immediate, at least in Australia.

    Psychologists have found that in evaluating whether to run a particular risk, people make the opposite decision from their last worst risk-taking decision. So, for instance, engineers who decided that a dam wall was safe (even though there was a risk it was not), and the dam broke engulfing a small town and killing many people, the next time they evaluated a dam wall found it to be unsafe. Individuals’ last worst decision regarding the risks of a prophylactic medical treatment will vary from person to person.

    Then, beyond the obvious ones, which risks fall within the disclosure requirements? The risk of blood clots from the AstraZeneca vaccine is present for two weeks or more after vaccination (there have been suggestions of even up to twelve months). Is the psychological stress of living with that possibility and the chance of developing a clinically significant anxiety disorder as a result, a risk that needs to be disclosed? What about the fact that there is no established treatment for the vaccine induced blood clots? In addition, must the over twenty percent fatality rate of the people who do develop vaccine induced clots be disclosed? These are very difficult questions on which there will not be consensus regarding the answers.

    The impact of both mainstream and social media on the public’s perception of risk is also an important consideration, especially in a “post-truth” world. The media, too, must act ethically and responsibly in their communications and not exploit people’s fear regarding COVID-19 or vaccination just in order to increase their audience share.


    The doctrine of informed consent puts into effect respect for the patient’s autonomy and their right to self-determination. Voluntariness requires that the patient’s consent to treatment be free of coercion, duress, or undue influence.

    Mandatory vaccination raises many complex ethical and human rights issues — including the fact that it would bypass informed consent. In situations that clearly fulfil all the requirements of the doctrine of necessity, it can be justified as an exception to the basic presumption that a person’s (or their legal surrogate decision-maker’s) informed consent to medical treatment is always required. Protection of public health is a long-established domain of such exceptions. (I note that some people will strongly disagree that any exceptions are ethical.)

    However, duress or undue influence can be less direct. For example, a person can be told that their continuing employment depends on their being vaccinated. This demand is most likely to be justified where it is necessary to protect vulnerable people from exposure to COVID-19, those who are fragile or have disabilities, such as residents of aged care homes. The employee has an option of resigning, but the consent to vaccination is not entirely free. And what about people over fifty, who would choose to have the Pfizer vaccine, but are offered only the AstraZeneca vaccine — is their consent to the latter sufficiently free?

    Final considerations

    I began by pointing out the four levels of decision-making and that this discussion of the requirements of informed consent to COVID-19 vaccination would focus on the micro- or individual level. However, all the other levels must be concurrently taken into account; the protection and promotion of the “common good” can modify what otherwise would be required at the individual level — including what is required to obtain an individual person’s informed consent.

    It is important to recognise that the values, which apply at the different levels, are not the same and can be in conflict. For instance, at the micro-level, an individual doctor treating an individual patient must not put the “common good” before the patient’s “best interests”. In contrast, at the meso-level, if a hospital is to act justly, it might be ethically required to give priority to the “best interests” of all patients at the expense of the individual’s “best interests”. The most likely situation in relation to COVID-19 where the “common good” can take priority over respect for individual autonomy is, at the macro-level, when protection of public health clearly requires a coercive intervention, such as lockdown or “social distancing”.

    It is a truism in ethics, but no less important for being such, that good facts are essential for good ethics. Dealing with COVID-19 vaccines involves many situations of unavoidable uncertainty with respect to the facts. We need to keep in mind that many ethical mistakes are made when we are frightened of living with that uncertainty and convert it to false certainty — in which case we are certain, but we are ethically wrong. This is an ever-present danger in the political and public policy sphere, especially under the guise of the “we mustn’t panic the public” syndrome, sometimes used by government bureaucrats to avoid transparency and “being blamed” for harmful outcomes of their decisions.

    Intention matters in ethics. The same decision might be ethical if made for one reason, but unethical if made for another reason. Let us test this by looking at the federal government’s decision about access to the Pfizer vaccine.

    The Pfizer vaccine, with some exceptions, will only be offered to people in general under fifty years of age. A majority of those who have developed blood clots are women. Should government policy be to offer women the Pfizer vaccine, rather than people under fifty? This would be to substitute sex discrimination for age discrimination. Or was the decision based on reasoning that those under fifty are less at risk in waiting for the Pfizer vaccine than those over fifty? Or was the reasoning something else?

    In short, to judge the ethical acceptability of this decision we need to know the risk/benefit calculus on which it was based. Without all those facts, we cannot properly judge its ethical acceptability.