“Patient autonomy” is not a good way to think about childhood vaccination policy

The Trump administration has made sweeping changes to its vaccine policy over the past year, justified by invoking the “personal autonomy” of patients. In January, Department of Health and Human Services head Robert F. Kennedy Jr. Trump reduced the number of routinely recommended childhood vaccines from 17 to 11 after an HHS report highlighted “personal autonomy and self-determination” as key principles requiring reconsideration of the childhood vaccination schedule.

That same month, the new head of the federal government’s vaccination panel led by Trump estimated that all childhood vaccines could be optional in schools, based on similar reasoning: “What we’re doing is putting individual autonomy back first,” Kirk Milhoan said, emphasizing that we allow people to make various choices, such as unhealthy alcohol consumption, even if there are risks.

And, late last year, Florida’s top health official and Trump ally, Joseph Ladapo, also called school vaccination mandates an encroachment on “bodily autonomy” as he moved to eliminate some vaccines required for school entry, although the efforts failed.

The pushback against such policies from those who support vaccination requirements in schools – like me – generally emphasizes that failing to vaccinate a child can endanger not only their own health, but also that of other children in the class, including some who may not be able to get vaccinated. It’s not only a question of personal autonomy, the argument could be this: our choices have an impact on others.

But in reality, it is not a question of autonomy at all, because young children do not refuse vaccination: their parents do. And while parents rightly have broad rights and responsibilities regarding their children’s health, this does not constitute “autonomy.”

This important distinction has been elided in this debate, which is unfortunate: recognizing that a parent’s responsibilities are not equivalent to a patient’s autonomy broadens the scope of appropriate policies to ensure that young patients receive the care recommended by doctors.

Medical ethics is traditionally presented as a framework of interlocking principles that can sometimes come into conflict. For example, “beneficence,” or the principle of doing what is best for a patient, may in some cases conflict with “autonomy,” or the principle of respecting a patient’s sovereignty over their own body, which is the basis of informed consent. Among adult patients who have the capacity to make their own medical decisions, autonomy (with rare exceptions like suicide) generally prevails – even in scenarios, such as in my field of medicine in the adult intensive care unit (ICU), where the consequences can be dire. For example, in the New York Times, physician and writer Daniela Lamas recently wrote about the nightmarish case of a young man she treated in intensive care with a dangerous heart rhythm abnormality that could have been corrected with a relatively minor procedure (placement of a pacemaker). But he steadfastly refused despite the insistent urgings of doctor after doctor, and died as a result. This was an avoidable loss of life, but the only alternative would have been to physically impose an invasive procedure on an adult deemed capable.

Things are of course different for patients – whether adults or children – who cannot make decisions for themselves. In these cases, medical decisions generally must be made by some sort of surrogate (usually a family member) acting on the advice of a doctor. Adults, for example, can designate “health care proxies” who are legally authorized to make medical decisions on their behalf – from approving treatments to withdrawing life support – if they become incapacitated. The use of such proxies is commonplace in my field: these people are often instructed to act in accordance with what they think the patient would want, not what they themselves would want.

On the other hand, for children, the default decision maker is the parent.

Even among adults, equating vicarious decision-making with true “personal autonomy” is something of a pragmatic fiction. On the one hand, studies show that discordance between surrogate decisions and patients’ “wishes” is common, that our wishes fluctuate over time, and that we may not even necessarily want them to guide decision-making as we approach the end of life.

More fundamentally, it is difficult, if not impossible, for any of us to know precisely what we would “want” in an ill-defined future state: the matrix of all imaginable medical conditions that we might one day face – and the potential interventions that might be considered for each – is vast. So attorneys do their best, in collaboration with doctors, to make decisions that are sensible and advance their loved one’s interests and wishes (to the extent that they can be understood). Yet, because of these limitations, the rights of agents do not have the same ethical preeminence as the actual autonomy exercised by an individual themselves. For example, imagine that the young man suffering from a heart rhythm problem is temporarily unconscious and therefore his medical team must consult his proxy regarding the placement of the pacemaker. Would doctors have refused a relatively minor procedure that would have prevented the imminent death of a young adult with (potentially) a full life ahead of him because the proxy refused to give authorization? I doubt it. Likewise, a parent’s directive that doctors should not give antibiotics to a child arriving in the emergency room with a rapidly fatal infection like bacterial meningitis would be rightly ignored.

But there are other distinctions between surrogate decision makers for adults (i.e., proxies) and children (i.e., parents). In the case of adult patients, surrogates can at least try to act on what they understand about a patient’s life philosophies, such as their previous thoughts about interventions such as life support, and sometimes even their written directives (although “living wills” tend to be very open to interpretation). But this is obviously not possible with infants: when a parent refuses a universally recommended vaccination for their child, no one even claims that this is the choice the child “would make” if they could speak for themselves. Thus, when it comes to young children, the ethical conflict is not between the principles of “beneficence” and “autonomy,” but rather between “beneficence” and “parental rights.”

I do not dispute the central responsibility of parents in medical decisions concerning their children: after all, who else would have it? But it is essential to clarify that, for children, distinct principles are in tension.

For example, when I previously saw (adult) patients in clinic, I would review the indicated vaccines and make recommendations. But if a patient refused a recommended vaccine, that was it and we would move on, even if she (for example) had advanced emphysema, which made her decision against the flu vaccine seem reckless to me. There would be no world in which I would or could “fire” a patient for such a decision: that is autonomy.

The principles are different in a pediatric clinic: a doctor who is not authorized to vaccinate a child against measles is asked to provide substandard care to a patient who cannot speak for himself. Autonomy did not enter the chat.

It does not follow that pediatricians should not accept patients whose parents refuse vaccination: nudge we may be sympathetic, but ultimately every child needs a pediatrician, so I’m against it. But again, the role of autonomy in the first scenario, and its absence in the second, helps to explain why these two situations have a totally different valence.

School vaccination mandates can be viewed in a similar light. They are important for the health of all children given the communicable nature of many vaccine-preventable diseases, and this provides a sufficient basis for many of them. But even setting aside these public health prerogatives, such mandates can be powerful protections that advance the interests of children who themselves otherwise, they would not be vaccinated. Such protections do not infringe on autonomy: on the contrary, their removal could infringe on it.

This may seem paradoxical, but it is not. It is of course ridiculous to think that we can know anything about an individual infant’s opinion, for example on the measles vaccine (the only given that seems to be the omnipresent hatred of needles). But the vast majority of adult Americans support vaccination against preventable diseases like measles. At least today, the average infant, when they reach an age where they can consider the evidence, will have supported their own vaccination. More importantly, when autonomy is absent and individual “wishes” are unknowable, the guiding principle should be best interest.

The principle of autonomy is rightly sacrosanct in medicine, although it even has limits, giving rise to a multitude of ethical dilemmas in my field of intensive care unit medicine and in other disciplines. But such controversies are incidental to the debate over school vaccination mandates. When it comes to recommendations or requirements for childhood vaccines, “personal autonomy” is little more than a red herring. Proponents of such protections should reframe the debate accordingly or risk ceding some ethical position.

Adam Gaffney, MD, MPH, is a pulmonary and critical care physician at Cambridge Health Alliance and an assistant professor of medicine at Harvard Medical School.