Risk perceptions and ethical public health policy: MMR vaccination in the UK

Abstract

This paper is concerned with how public health policy makers should respond to the public’s perception of risks. I suggest that we can think of this issue in terms of two different models of responding to the public’s view of such perceived risks. The first model I will call the “public perception” view (PP view) and the second the “public good” view (PG view). The PP view suggests that the public’s perception of any risks is so important that public health policies should be formulated in direct response to knowledge about them. I will consider two possible ethical arguments that might be offered in support of such a view: the first argument is an “autonomy” argument and the second a “consequences” argument. I suggest there are serious problems with both arguments. I then outline an alternative model of public health policy formation that I call the “public good” or PG model. This model focuses on drawing distinctions between the clinical and the public health context, and argues that most of public health policy is primarily concerned with the creation and maintenance of various public goods. This latter fact means that the PP model is inappropriate for public health policy formation.

Zusammenfassung

Dieser Beitrag beschäftigt sich damit, wie Entscheidungsträger in der Gesundheitspolitik der Wahrnehmung von Risiken in der Öffentlichkeit begegnen sollten. Mein Vorschlag ist hier, diese Frage im Rahmen zweier verschiedener Modelle der Reaktion auf die öffentliche Sichtweise solcher wahrgenommender Risiken zu betrachten. Das erste Modell bezeichne ich als die Perspektive der öffentlichen Wahrnehmung (“PP (Public Perception)-Modell”), das zweite als die Perspektive des öffentlichen Nutzens (“PG (Public Good)-Modell”). Nach dem PP-Modell sind allgemeine Risikowahrnehmungen von solcher Wichtigkeit, dass Gesundheitspolitiken als direkte Reaktion auf das Wissen über solche Wahrnehmungen zu formulieren wären. Ich betrachte zwei mögliche ethische Argumente, die zur Stützung dieses Standpunkts angeführt werden könnten, das erste ein “Autonomieargument”, das zweite ein “Folgenargument”. Ich zeige auf, dass beide Argumente mit ernsten Schwierigkeiten behaftet sind. Dann skizziere ich ein Alternativmodell zur Gestaltung öffentlicher Gesundheitspolitik, das so genannte PG-Modell, in dem der öffentliche Nutzen im Vordergrund steht. Dieses Modell konzentriert sich auf die Unterschiede zwischen Fragen des klinischen Gesundheitswesens einerseits und solchen der öffentlichen oder Volksgesundheit. Das Argument lautet dann, dass es in der Politik zur öffentlichen Gesundheit weitgehend und in erster Linie um die Schaffung und Erhaltung verschiedener öffentlicher Güter geht, was allerdings bedeutet, dass das PP-Modell für die Politikformulierung zur öffentlichen Gesundheit ungeeignet ist.

Résumé

Le présent article s’interroge sur la manière dont les décideurs dans le domaine de la politique de santé publique devraient répondre à la perception des risques par l’opinion publique. Je suggère que nous réfléchissions sur ce thème dans le cadre de deux modèles différents de réaction à la vision de l’opinion publique des risques ainsi perçus. J’appelle le premier modèle le point de vue de la perception publique (point de vue PP (Public Perception)), et le deuxième le point de vue de l’intérêt public (point de vue PG (Public Good)). Le modèle PP suggère que la perception publique des risques revêt une telle importance que les politiques de santé publique devraient être formulées en réponse directe à ce qu’on en sait. Prenons deux arguments éthiques possibles pour étayer ce point de vue : le premier argument est un argument ≪ d’autonomie ≫ et le second un argument ≪ de conséquences ≫. Je montre que ces deux arguments posent de sérieuses difficultés. J’esquisse ensuite un modèle alternatif pour l’élaboration de politiques de santé publique, le modèle PG ou ≪ d‘intérêt public ≫. Ce modèle se concentre sur les différences entre l’aspect hospitalier de la santé et son aspect public, et argue que la politique de santé publique porte essentiellement sur la création et la préservation de biens publics. Ceci signifie que le modèle PP ne convient pas à l’élaboration d‘une politique de santé publique.

1 Introduction

A paper written by Andrew Wakefield and others was published in The Lancet in 1998 suggesting grounds for a possible association between the measles, mumps and rubella (MMR) combined vaccine and bowel problems and autism. ¹ The paper was based on a study of only twelve children and explicitly stated that the authors had no evidence strong enough to establish a causal link between the vaccine and the health problems. However, they suggested the need for further investigation of their findings. The hypothesis was widely reported in the media, not always accurately. The main author of the paper has been widely interviewing in the media, when he often proposes a stronger version of the hypothesis than that contained in the published paper. Uptake rates for MMR have dropped significantly in the UK since the publication of this paper, presumably as a direct result of the public’s perception of MMR based upon the media reporting of the paper’s content. ² A significant minority of the population are either not getting their child vaccinated with MMR or are seeking vaccination through single vaccines which they suppose are safer. ³ Despite these events a dispassionate consideration of the available empirical evidence about the safety of MMR is likely to conclude that the combined vaccine is the best option available to parents and the supposed side-effects are either rare or non-existent. ⁴

This paper is concerned with the following question: how should those given the task of formulating public health policy in relation to vaccination respond to such a situation? Should they insist upon a policy based upon the best possible empirical evidence or should they respond to public concerns about the alleged risks of harm by basing their policy on what many parents want (as they have been urged by some commentators)? ⁵ What will count as an ethical public health policy in such a situation? In this paper I will use the recent discussions about MMR in the UK as an example to illustrate this issue, but the argument should be relevant to vaccination policies in general and other aspects of public health policy. ⁶

In the course of this paper I consider two arguments for taking the public’s perceptions of such risks seriously. However, I argue that they do not prove to be decisive in suggesting the need to adopt a policy based upon such perceptions. Whilst such factors are relevant and certainly need to be taken into account, they should not simply be used to determine policy in this area. It is proposed that we should adopt a different model of public health policy formation within which the important factor is what will bring the greatest benefit to the public’s health. I argue that a vitally important aspect of this is to concentrate upon the production and maintenance of public goods such as herd protection. ⁷ I suggest that we can think of this issue in terms of the adoption of two different models of responding to the public’s perception of risk. The first model I will call the “public perception” view (PP view) and the second the “public good” view (PG view). In brief, the PP view suggests that the public’s perception of any risks is so important that public health policies should be formulated in direct response to knowledge about the public’s perceptions of the relevant risks. I will consider two possible ethical arguments that might be offered in support of such a view: the first argument is an “autonomy” argument and the second a “consequences” argument. I suggest there are serious problems with both arguments if used to support the PP view. I then outline an alternative model of public health policy formation that I will call the “public good” or PG view. This model focuses on drawing distinctions between the clinical and the public health context, and argues that much public health policy is primarily concerned with the creation and maintenance of various public goods. This latter fact means that the PP view is inappropriate for public health policy formation. I now turn to consider the first argument that might be used in support of the PP view.

2 The “autonomy” argument for the PP view

The “autonomy” argument is based upon the claim that when the public propose a change in policy the relevant authorities should respond to that request because that is what the public want. This view takes the role of public perceptions very seriously indeed and suggests that we can treat public health considerations as if they were in this respect just like individual clinical encounters. In both situations the central factor in decision making is what the autonomous patients request or wish for. The argument might go that given the respect deserving of such autonomous choices within a broadly liberal and democratic culture, we should use our knowledge about patient wishes to formulate the relevant public health policy. This appeal to autonomy will, of course, in turn have to be justified; but it is a truism that various different moral theoretical positions might support the idea that autonomy is an important moral consideration or value. However, it looks as though the appeal to autonomy in this argument must be to some sort of absolute principle of autonomy, where it is suggested that autonomy takes priority over all other moral concerns. It will be for the supporter of such a view to justify this rather implausible claim. ⁸

Putting this justificatory issue to one side, there are clear objections to such a view if applied in the public health context. One obvious difference between the clinical situation and the public health situation is that it is unlikely that in the latter case there will be a single view in a population in relation to such public health measures that can then be used as the basis of such an autonomy-based policy. However, I suggest we put this type of practical objection to one side as well. I am more interested in this paper in the possibility of more fundamental objections to the argument than this. I will consider two such possible “fundamental” objections here. The first is that there is an important difference between public health and clinical situations because of the difference in aims behind the interventions. The second is that what people want should not be the decisive issue in the formulation of public health policy.

The first objection to the autonomy argument is based upon a real difference between clinical and public health situations, in that the potential harms and benefits of any public health policy arguably extend beyond the sum of the harms and benefits for each participating individual. ⁹ If this is true then it can be argued that it is not clear why we should base policy making about public health matters upon the autonomous expressions of individuals in the sense proposed in the “autonomy” argument (even if we can make sense of what “population autonomy” might be). The idea is that such a judgment cannot be determined purely on the basis of adopting what individuals might want such a policy to be. On this critical view, considerations about the public’s health should be the driving force in the formulation of public health policies as we cannot make any sense of how individuals can directly contribute as individuals to such policy formation.

The second objection is to the way that the autonomy argument relies upon giving weight to the idea of wants. We can ask why we should choose to base a policy upon such a notion? It is clearly possible to want something inappropriate for your own health. It might be decided in the clinical case to respect your autonomy even to the point of allowing you to potentially harm yourself (perhaps after being fully informed about such likely consequences). However, in the public health realm it is not clear that we should adopt the same approach. Once again, this is because in the public health arena, at least in relation to vaccination for contagious diseases, individuals are not just deciding about something with consequences for their own health when they decide to forego a vaccination as their decision may have consequences for others. Your decision not to vaccinate your child for measles might mean that my child, too young to receive her scheduled vaccinations, is infected by yours. Your choice has resulted in harm to another. In terms of standard liberal theory you are required to take harm-to-others concerns seriously in your deliberations. ¹⁰ Just as individual choice has potential implications for others, so do the choices of policy makers in devising their policy. By definition, then, in shaping public health policy in relation to such vaccinations, policy makers are deciding a policy that has implications for all. This means that if there are fewer individuals vaccinated, chances are that it will not just be the individual(s) that requested that particular approach that are thereby put at risk. As I have just suggested it might well be “innocent” others unable to take any steps to protect themselves who are at greatest risk. ¹¹

Consider the example that I mentioned before. In the UK there have been many requests for separate single vaccines rather than the use of the combined MMR vaccine. The Department of Health has, so far, rejected these requests. Should public health policy respond to such requests for single vaccines (or offer a choice between MMR or single vaccines)? It might be argued by appeal to the autonomy argument that it should as this will respond to parental autonomous requests and this is a good thing. However, the counterargument is that adopting such a policy will impact upon all citizens, and that the public’s health will be worse off as a result. This is because a policy of encouraging single vaccines through state support is likely to result in lower coverage for at least some diseases as each child will need to have so many more injections on a number of different occasions. For example, MMR vaccination requires two sets of injections in pre-school children, whereas separate vaccines would require six sets of injections. This will increase the risk of harm to each individual child from these diseases as it will take much longer for immunity to be established, but it will also almost certainly result in poorer levels of immunity in the population as a whole as it is more likely that scheduled vaccinations will be missed, so increasing the risk of epidemics. ¹² This increased risk of harm will be a particular danger to those who are not vaccinated either because they are too young (e.g., neonates) or because they are allergic to one or more component of a vaccine.

For the sake of completeness I will consider two further possible arguments in favour of PP here, though in less detail, one based upon the importance of public trust in public health policies and another based on an appeal to parental perceptions of harm. Public trust in public health interventions is vital, as otherwise participation will fall, and the aims of any such programme will not be achieved. ¹³ In addition, once trust is lost it is an extremely difficult business to rebuild it. It takes time and effort. In the MMR case, it might be that ignoring the minority of the public requesting single vaccines means that such trust in the MMR vaccination programme or vaccination programmes as a whole may be reduced or lost. However, from a policy maker’s perspective this is a difficult issue as trust will also be lost if the vaccines used either cause harm through side-effects, fail to protect the individuals vaccinated due to vaccine failure, or if they fail to prevent outbreaks. If evidence exists that a combined vaccine such as MMR is safe and there is no such comparative evidence in relation to the corresponding single vaccines, it might be that public trust in the long-term is better protected by continuing to advocate the use of MMR even if some public trust is lost in the short-term.

Another possible argument in favour of PP is that all parents are seeking to do by not accepting MMR vaccination is, to protect their children from what they perceive as a significant risk of harm. Of course the formation of public health policies, just like any other medical practices, should involve a consideration of the risks of harm involved in any proposed intervention. But this does not mean that we should seek to avoid every single proposed risk (e.g., from possible side-effects), and if we did so, life would be impossible. Everything (including doing nothing) carries some degree of risk of harm. In addition, in this particular case, any advocate of such a view is almost certainly ignoring the (greater) risk of harm arising from non-vaccination. Public policy should indeed respond to evidence about any risks, but once again this is likely to be a difficult exercise, to be made in the light of all relevant evidence. Any such assessment of the risks of harm must consider a number of factors, such as not only the fact that possible risks might follow from that measure, but also that there are risks involved in not continuing with an existing policy or that risks will arise from changing an existing measure.

Of course, one potentially serious problem is that very often the relevant empirical evidence for assessing the various risks is either not available or is less than perfect in some way. ¹⁴ In addition, there has been a great deal of empirical work looking at the way that people understand and react to different risks. ¹⁵ What these factors together suggest is that when considerations involving risks are discussed, special care should be taken to ensure that the relevant risks are not either inappropriately magnified or reduced. There have been some efforts to take this issue to heart, such as the UK Department of Health’s advice on communicating risks to patients. ¹⁶ However, despite these efforts it might still be possible to argue that this provides another reason for decision making to be made by well-informed policy makers on the basis of the best possible medical evidence rather than on the basis of a response to the attitudes of the public. This is because if it is true that there are systematic biases in the way that people make decisions (such as tending to overestimate immediate risks and underestimate long-term risks), and the empirical evidence looks quite robust in relation to such “framing effects”, this provides a good reason to be cautious about basing policy on PP. ¹⁷ Of course, “experts” will not always be right either, as most would instantly concede, but they are less likely to make these systematic errors in judging the relevant evidence.

3 The “consequences” argument for the PP model

The “consequences” argument is very different to the “autonomy” argument. This argument does not rely upon the fact that (some of) the public have particular attitudes about certain risks as such, but rather the focus is upon what the consequences or results of such beliefs might be. It relies upon the idea that where policy does not change in response to public attitudes in relation to the relevant risks this may itself have consequences for the public’s health. Therefore, the focus here is upon the whole population in question, and what is best for that population’s health, rather than focusing on the attitudes of individual members of that population and their individual attitudes to the vaccination in question. For example, it might be argued that the policy of only promoting MMR, and not making single vaccines available, may over time actually reduce the effectiveness of the whole vaccination programme as individuals opt out of all vaccinations (even those that they would otherwise have agreed to). ¹⁸ This in turn might impact adversely upon immunity rates within the population and so increase the overall risk of harm to the public’s health.

I would argue that this argument is a stronger one than the autonomy argument and that it in fact gains in power as more people decline vaccination for their children. Indeed, there must come a point where vaccination rates become so low that the advocacy of a policy defending a particular vaccine (such as MMR) becomes untenable. In such a situation, the full power of the consequences argument kicks in and policy makers would need to think carefully about the best way to protect the public’s health. This might mean that in some cases defeat would have to be reluctantly conceded by policy makers, not on the basis of the empirical evidence, which may still support the vaccine in question, but in response to the practical realities about the need to maintain the maximum level of vaccination across the population and thereby protect the public’s health. Indeed, it might be argued that to do otherwise would be unethical, on the grounds that practical considerations should trump ideological commitments in such situations. Of course any policy maker is unlikely to publicly endorse such a view, but in reality, they might have to respond to such a situation. In the MMR case, this point would be reached if the current policy of only providing MMR itself threatens to bring about greater public health harms than the provision of single vaccines. However, it has been well argued by the UK government that the problem is that we have little evidence as to potential benefits and harms from single vaccines (whereas we have significant evidence for the effectiveness and safety of MMR).

Note that it is unlikely that any advocate of the PP model will appeal to the consequences argument, but it is something for the policy makers to take into account. It will not be used by the supporter of the PP model because until the tipping point described above is reached, any argument appealing to consequences is probably going to be based upon the best available evidence of effectiveness (even in situations where a minority are allocating a disproportionate weight to a particular perceived risk). So, for example, in the MMR case a consequences argument in the current situation will be likely to support the status quo in favour of MMR. The power of the argument comes about because anyone truly committed to seeking to protect the public’s health will have to accept that at some point the existing policy may have to be abandoned as this could prove to be the best way of protecting the public’s health.

4 The “Public Good” model of public health policy formation

In response to a model of public health decision making that is actively shaped by patients’ perceptions of assumed risks, I would like to propose an alternative. This I will call the “Public Good” or PG model. The main idea behind this approach is that whilst those with the job of formulating public health policy should take into account the public’s perceptions of the relevant risks (as suggested in the “autonomy” argument) as well as the consequences of not doing so (as suggested in the “consequences” argument), they should not just note the public’s view and then infer from these facts their decision about the relevant policy. Such elements must inform the policy making, but other factors remain important as well. A vital consideration is what sort of policies will actually benefit the public’s overall health. This factor might mean that the other two elements are not decisive when it comes to policy formation.

In fact, in beginning to construct an alternative, we can offer a diagnosis of the problem that lies behind the PP model. One of the issues is that it is in danger of ignoring the differences between clinical medicine and public health. These differences are often lost in the application of principles constructed in the sphere of medical ethics to the field of public health. It is clear that the issues relating to the health of populations are different from those of the doctor–patient relationship. Whilst appeals to autonomy might be central to determining the right thing to do in relation to deciding an appropriate clinical treatment, it is not clear that the same is true in relation to public health decision making. ¹⁹ For example, it can be argued that when risks of harm are considered within a public health context, it is vital to give special consideration to what the wider impact of a particular measure might be upon the population as a whole. This is because the focus of attention within public health is not just upon individual patients alone but also upon the health of the public as a collective group. It can be argued that this means that consideration of such issues must go beyond the sum of interests of each of the individual patients. The easiest way to see the importance of this point is to see the centrality of public goods to most public health activity. What do I mean by public goods?

Vaccination for a contagious disease can be seen as an act that not only has potential benefits for the individual themselves but as also making a contribution towards a collective benefit as everyone profits where herd protection exists in the population, including those who are not vaccinated (for whatever reason) and those who do not show sufficient immune response despite vaccination (for example, because they were ill at the time of vaccination and their immune system was already stimulated by other infectious agents). One way to think of these two positive outcomes is as being both a private (or individual) and a public (or collective) good. A private good is a good that benefits only the individual concerned and a public good benefits all of the members of a group or population. It is because of such a shared population benefit from the existence of herd protection that we might think of the existence of such a state as being a public good. We can think of public goods in general, following Klosko, as being characterised by two main properties: non-excludability and dependence upon cooperation by a large number of people. ²⁰ “Non-excludable goods” are those where no one can be excluded from the benefits of the existence of the relevant good, even when they have not contributed towards bringing it about. ²¹ This is the case where herd protection exists: the non-vaccinated benefit from the high immunity in the population. In addition, herd protection can only be generated through collaborative endeavour as it cannot be created, obtained or controlled by any individual acting alone. In addition to these two aspects of public goods suggested by Klosko, I would add another proposed by Rawls. ²² He suggests that public goods must also be indivisible: that is, they cannot be broken down or divided into individual or private goods to be distributed amongst the members of a group or population. This is true of herd protection in the sense that the equal collective benefit can only come about as a result of collective action. Herd protection where it exists can therefore be seen to be a public good as it is indivisible, non-excludable and dependent upon the cooperative actions of a group. ²³

Many interventions in public health aim to create or maintain public goods as defined above. Herd protection in a population, where it exists, serves as an excellent example of such a public good. Herd protection is a public good because it creates a benefit that goes beyond that of the vaccinated individuals qua individuals. The whole population derives a substantial benefit from the existence of herd protection in addition to any individual benefit to the vaccinated individuals. Such benefits are felt directly by those that are unvaccinated, such as children too young to be vaccinated, but also indirectly by society as a whole. By indirect benefits here I mean that a society is better off if it knows that the threat of a potentially devastating disease is reduced. Such a benefit does not come about through the direct protection of the vaccination itself, but the knowledge that normal life is less likely to be affected by the consequences of any epidemic. ²⁴

It should be noted that one possible objection to the PG model of policy formation is that it might be held by the advocates of the PP position to be paternalistic. Paternalism is best defined as acting so as to promote the overall best interests of another individual. If I intervene to prevent you from smoking by removing your cigarettes then I am being paternalistic. We might then go on to argue about whether my action was justified or not. This might depend upon a range of factors such as whether you are competent or not, whether you are aware of the risks you run to your health, perhaps even whether you have asked me to do this in the past whenever I see you smoking. However, doubts might be expressed about whether the PG model of public policy formation is a form of paternalism because the focus of action is not merely upon the good of the relevant individuals in society, but it also takes account of the good of society as a whole resulting from the existence and maintenance of the relevant public good. The intention to create or maintain public goods is very different to a decision relating to an individual’s health. It might well be that the existence of such a public good as herd protection is in any individual’s best interest in that particular society where it exists, but this does not mean that action in relation to creating and maintaining public goods is paternalistic in any straightforward sense.

It should also be noted that the PG view provides a different perspective upon public health from that of the consequences argument, despite sharing some similarities. The consequences argument is a purely consequentialist argument concerned with all of the relevant outcomes of an act. Such an approach may be used to justify many public health actions. ²⁵ However, the PG view provides a different type of justification: one focused on the creation and maintenance of public goods (as defined above). Whilst the PG view can be supported by consequentialists (and often will be as public goods are likely to contribute to human welfare) it may also be supported by more deontologically inclined individuals (because public goods are also likely to contribute towards non-maleficence and beneficence). Arguments about public goods are therefore distinct from arguments about consequences.

If we return to the MMR issue, the PG view will produce an outcome as follows. Worries about the supposed risks of MMR and the possible subsequent actions by parents (e.g., non-vaccination) mean that policy in this area must take account of these concerns, but these must be weighed against the important benefits to the public health following from the maintenance of the existing MMR programme (e.g., fewer injections; greater immunity in the relevant population, etc.) and the judgment on the basis of the existing good quality medical evidence about what the best possible policy to protect the public’s health will actually be. This view suggests that the relevant “experts” will have an essential role in deciding the relevant policy. Whilst such experts are, of course, open to bias and manipulation they are also trained to make judgments about the relevant material medical evidence and not base their policies on partial media reporting or selective reading of the relevant literature. Any such judgments and policies are then, of course, open to public scrutiny and debate.

5 Conclusions

In conclusion, I have argued that ethical public health policies need to be formulated in the light of evidence about public attitudes, but this does not mean that those same attitudes should directly determine public health policies. ²⁶ Far more important to the formulation of ethical public health policy is the creation and maintenance of public goods such as herd protection. Such a basis to policy requires action at the level of government and should be informed by the best possible evidence about the risks and benefits of the relevant vaccines. Such policy should only consider taking public perceptions of risk as the basis of policy when what I have called the consequences argument becomes so strong that the credibility of vaccination in general is threatened. Otherwise the right thing to do is to keep an eye on public perceptions, take them into account, but work with the PG model of policy formation. The supreme moral obligation of government in relation to public health is to create and maintain relevant public goods such as herd protection. ²⁷