Our overview of ethical frameworks shows that various efforts have been made to help policymakers and public health professionals deliberating about the ethical aspects of public health policy and programs. Kass offers a step-by-step procedure to weigh the burdens and benefits of a program . Childress et al. assist in evaluating programs that promote public health but that infringe upon other moral considerations. PHLS provides ethical standards to guide the practices of American public health institutions. Europhen gives insight in ethically relevant public health differences within the European Union and in ways to bridge them. The Nuffield stewardship model distinguishes acceptable goals and restrictions of public health programs, and its intervention ladder helps in balancing a program's benefits and its intrusion in people's lives. Finally, Tannahill's triangle assists in integrating ethics and evidence in such a deliberation.
However, all frameworks have limitations with respect to their practical value in the evaluation of programs to prevent overweight and/or obesity (see Table 1). Nuffield is the only framework that specifically addresses obesity prevention. Four frameworks can be applied to concrete programs related to overweight or other public health problems, but Europhen and PHLS cover a more abstract question, namely: 'what ethical values should direct public health policy?'[9, 10]
We found it remarkable that none of the frameworks specifies when and by who it should be used. This may stem from the desire to develop a framework that is broadly applicable and that can be used by anybody at anytime. We think that users of a framework would benefit from procedural guidelines for applying the framework. Especially professionals who have no experience with ethical consultation and who must fit the application of an ethical framework into their other tasks may profit from suggestions. Advice about the best time to apply a framework (before the implementation of a program or already during the designing phase) and about the number and background of the persons who are to use it, may save efforts and thus lower the threshold of using a framework.
Kass, Nuffield and Tannahill offer an analytic tool, which is an instrument to guide the evaluative process. These tools comprise a decision-making-triangle, a step-by-step-questionnaire and a ladder to indicate proportionality[5, 11, 12]. Such tools make a framework more practically useful for policymakers than merely a set of ethical values does. In addition, framing questions may contribute more to adequate deliberation of the ethical aspects of programs than providing fixed answers or guidelines. The Europhen policy recommendations, for instance, aim to help policymakers solving ethical issues by indicating the direction that policy should take[10, 11]. As opposed to this, Kass and Tannahill for instance frame the questions that should be raised and thereby encourage the process of deliberation. Kass leaves answering the question 'How can burdens and benefits be fairly balanced?' up to the public health professional or policymaker. Tannahill's triangle formulates the steps that are to be taken in the process of deliberation without filling in the decisions that should be made.
No simple solution seems to be available for dealing with ethical conflicts, although it is precisely the tendency of ethical principles to infringe upon each other that creates the need for frameworks. The designers of the frameworks agree that the principles cannot be ordered according to priority but must be weighed in concrete circumstances. Kass, Nuffield and Childress et al. identify criteria for this weighing process[5, 8, 11]. They agree on the fact that the burdens of a public health program should be in proportion to its benefits. Furthermore they refer to the 'harm principle', which implies that restrictions to people's freedom should be minimized and that they are only justified in case of a clear public health requirement. Childress et al. distinguish themselves from the other frameworks by putting ethical conflicts at the centre, rather than merely pointing out ethical values. They point out five justificatory conditions for public health programs that infringe moral principles, namely: effectiveness, proportionality, necessity, least infringement, and public justification. PHLS and Europhen do not articulate criteria for dealing with ethical conflicts[9, 10].
However, even with sound weighing criteria, disagreement about the outcome of a framework is inevitable. That is because personal, cultural and political perspectives affect the process of interpretation and weighing. Several frameworks recommend fair procedures for dealing with difference of opinion. Tannahill encourages an explicit use of the decision-making triangle, including documenting judgements. This may contribute to consultation and dialogue, and enables a discussion about disagreements on the basis of shared principles. Kass argues for a democratic process and public hearings to consider minority views. And Childress et al., to conclude, advocate a transparent process for expressing justice and sustaining public trust. Such a process requires both asking input from the public, as well as offering justifications for decisions that have been made.
Most of the frameworks aspire not only to set ethical boundaries (such as restrictions to interference), but also to articulate positive ethical foundations for public health (such as the duty to diminish inequalities), which seems to contribute to their practical value. However, the usefulness for prevention of overweight or obesity requires that all ethical issues that are relevant for this field are clearly addressed. The majority of the frameworks frames abstract ethical values without outlining the concrete ethical issues they may give rise to. Most frameworks contain a set of ethical values. Some are articulated as principles, whereas others take the form of policy recommendations or goals. Only Kass' framework does not include a list of values, but her description of relevant ethical considerations does refer to them. These abstract ethical values do more or less cover the relevant ethical themes. For instance, the issues of liberty and responsibility that may occur in programs to prevent overweight are in all frameworks covered by the classical values of liberty and responsibility. Nuffield, Europhen, PHLS, and Tannahill explicitly mention social responsibility and stress the need for creating a healthy environment and facilitating healthy behaviour, which are both relevant for the prevention of overweight[9–12]. Europhen is the only framework that emphasizes that citizens also have duties, thereby paying attention to the debate about accountability for an unhealthy weight. It states that 'citizens consider themselves as consumers of healthcare who see health services as their right as taxpayers. However rights have reciprocal responsibilities, and the public must be reminded of these.'
Furthermore, all frameworks (except for Tannahill's) address the issues of privacy by mentioning the values of privacy and confidentiality. And all frameworks address the issue that the effectiveness of a program to prevent overweight may be uncertain or unfavourable by mentioning the values of well-being, and sometimes by mentioning the value of utility (producing the maximal balance of benefits over harms and other costs)[5, 8, 11, 12].
However, almost none of the frameworks describes the concrete ethical issues that may occur in programs. The issue of equality is covered in all frameworks except for the recommendations by Europhen. But knowing that equality is an important value does not specify that programs to prevent overweight may increase already existing health inequalities by being least effective among groups that have the highest risk of developing overweight. Likewise, the importance of providing adequate information is covered by the values of autonomy, transparency and trustworthiness that are mentioned in all frameworks. However, inadequate information is sometimes distributed by accident, and the frameworks do not provide guidelines about what adequate information exactly entails and how to prevent the accidental distribution of inadequate information.
Furthermore, two issues were absent in most frameworks. One issue, that interference may occur with cultural and social values of food and eating habits, is only covered by the PHLS framework, which articulates the need to respect cultural value pluralism: 'Public health programs and policies should incorporate a variety of approaches that anticipate and respect diverse values, beliefs and cultures within the community.' The other issue, namely the potential negative psychosocial consequences of programs to prevent overweight (such as uncertainty, fear and weight concerns about the health risks of overweight and obesity, stigmatization and blaming, and unjust discrimination), is by most frameworks only covered to a limited extent. Only Kass and Nuffield warn against the potentially stigmatizing effects of targeted messages[5, 11]. None of the frameworks goes into detail about how programs can reinforce the negative image of overweight people, how they may create unnecessary concerns about health risks, or how they may undermine self-confidence for people who do not succeed in losing weight. The lack of attention for cultural values, and for stigmatization and other psychosocial issues may be explained by the fact that these issues are particularly relevant for the field of overweight prevention and less for other fields in public health.
Designers of frameworks face the challenge of acknowledging the complex character of ethical issues, without loosing sight of their main task, namely guiding professionals in the process of articulating and dealing with ethical issues. Presenting a set of abstract ethical principles does not provide guidance to policymakers who are not familiar with ethics. This is not a shortcoming of the frameworks in themselves, since each has its own particular aims, but it does indicate that our last criterion is not satisfied by the available frameworks. Thus, it is questionable to what extent the frameworks facilitate deliberation among policymakers regarding the concrete ethical issues in the prevention of overweight and obesity.
Our study has several limitations. It is possible that we overlooked one or more frameworks that are suitable for evaluating the ethical aspects of programs to prevent overweight and/or obesity. Furthermore, our analysis of the usefulness of frameworks is restricted to self-developed criteria. We did not interview policymakers in the field of overweight prevention about the usefulness of frameworks and we did not test the frameworks on actual programs.