The societal consensus that has supported the United States' universal childhood immunization programs for the past 50 years shows signs of eroding. This article proposes a systematic approach to evaluate immunization policy options. Through a unifying framework that combines epidemiologic, economic, and ethical concerns, this approach promotes a clearer understanding of underlying issues and inherent tradeoffs between alternative policies. Such a systematic examination of policy options could facilitate the public dialogue necessary to continually recreate a broad consensus on immunization practices and enable us to choose policies most in accord with our fundamental values.
- public health
- health care policy
- multi-attribute decision analysis
When Benjamin Franklin recalled, in the pages of his autobiography,1 the death of his 4-year-old son in 1736 from smallpox, he rued his decision to forego inoculation for young Francis. Years earlier, when Boston was being ravaged by the 1721 smallpox epidemic that killed hundreds, Franklin and his older brother had lambasted the Reverend Cotton Mather and Dr Zabiel Boyleston for advocating the “mischievous” practice of variolation; but decades later the still-aggrieved father reversed his previous position and counseled parents to pursue this “safer” course of immunization.1,2
Franklin's inner dialogue of pro and con regarding inoculation symbolizes an enduring societal debate that has embroiled immunization programs. Edward Jenner's development of “scientific” vaccination in the 1790s culminated in an 1867 British law mandating smallpox immunization, yet by 1869 organized political opposition had arisen and persisted unchecked (with children unimmunized) for decades.3 Louis Pasteur's immunization in 1885 of the boy Joseph Meister against rabies created an ethical uproar that turned riotous.4 Recent concerns about the safety of whole-cell pertussis vaccines led to a disruption or cessation of national pertussis immunization programs and a resurgence of epidemic disease in the United Kingdom, Japan, Sweden, and West Germany.5Sustained by the media, such turmoil roils while vaccination, acknowledged to be one of the most beneficial public health interventions ever, continues to prevent epidemics.6
Against this historical backdrop, 3 recent trends have made decisions regarding immunization policies even more complicated. First, the broad cultural consensus that has enabled the United States' universal childhood immunization programs of the past 50 years shows signs of eroding. With most parents and many grandparents no longer personally acquainted with the morbidity and mortality of vaccine-preventable diseases, many families have shifted their focus of concern to alleged vaccine reactions. Adverse events that occur in temporal association with immunization are presumed to be causally related, leading some to oppose mandatory immunization. Indeed, 15 state legislatures in 1999 considered bills to reduce or eliminate school entry immunization requirements (J. C. Dolins, personal communication, September 29, 1999).
Second, this questioning of mandated immunization occurs just as advances in molecular biology and immunology promise to introduce an array of novel vaccines. In the coming decade new immunizations will present many challenging policy and funding decisions for national committees that formulate immunization recommendations and determine federal entitlements, state legislatures that enact immunization laws and allocate some immunization program funds, and both public and private health plans that decide on coverage and payment or financing policies.
Third, for the past 2 decades the health care community has improved the methods used to evaluate medical technology and make policy decisions. Cost-effectiveness studies have weighed cost against some measure of benefit, effectiveness, or usefulness in improving health.7–15 Such analyses, however, have not formally considered ethical concerns, such as protecting individual rights or providing an equitable distribution of health care benefits. Because of this limitation, cost-effectiveness analyses, while necessary for policy decisions, are neither sufficient nor complete, because they currently do not address differences in values and perspectives that polarize our society.16–18
These 3 trends will make immunization policy decisions more numerous and complex. Because the powers granted to public health authorities are based on the public's trust, and in democratic societies this trust is founded on broad participation in formulating policy, health care decision-makers may well feel mounting pressure to include the diverse perspectives not only of physicians and immunization experts, but also of parent groups, politicians, special-interest advocates, economists and, perhaps, ethicists.
How can any decision-making process accommodate such a wide variety of concerns? Ben Franklin, no stranger to complex decisions, once proposed to a friend a simple method of “moral or prudential algebra” to sort through a host of competing objectives, weighing various pros and cons, to reach a good decision (B. Franklin, letter to J. Priestly, September 19, 1772). The science of decision-making has subsequently refined Franklin's method greatly, but the essential value of an explicit systematic approach remains: to protect against using unexamined biases, rather than one's best knowledge and deepest convictions, to make important decisions.19–22
In this article we propose a systematic approach to evaluate immunization policy options. Our model combines epidemiologic, economic, and ethical concerns into a unified analytic framework, thereby helping us to understand better the tradeoffs between alternative policies, and assisting us to choose a course of action that most accords with our fundamental values. We believe that such a systematic method of deliberation would foster a more explicit and morally relevant dialogue about diverse policy considerations than do current US policy debates and analysis.23
A SCENARIO OF THE PROBLEM
To illustrate our approach, imagine that we are public health officials on a distant island where the children are plagued by an endemic infection—Franklin Fever. Few children escape infection from this highly contagious virus. Most often the disease causes only a week of fever, cough, and an itchy red rash, but for ∼1% it results in a prolonged course of encephalitis, and ∼5% of these children either die or suffer brain damage with long-term disability, generating substantial social costs. Through the collaborative efforts of the island's academic and industrial research communities a safe and effective Franklin Fever vaccine is now available and an expert medical committee has recommended universal immunization of the island's children, noting that the cost of immunization would be roughly comparable to total direct medical and indirect societal costs of caring for afflicted children.
As physicians with responsibility for formulating public health policy, we must choose an immunization strategy. Stung by criticism that our previous decisions paid little heed to ethical concerns, we wish to consider several issues, including equal access to the immunization for all children, fair sharing of the benefits and risks of immunization, and due respect for those families who refuse to have their children vaccinated. How might we proceed?
I. CONSIDERING THE PROBLEM
The merits of particular immunization strategies can be clarified if we start by considering 3 broad domains of concerns (Table 1). First are the consequences of the disease: How much harm does it cause individual children and adults? How much does society fear these harms? Does the disease pose a substantial epidemic risk through person-to-person transmission? How much does the care of the acutely ill or subsequently disabled child cost the health care system directly as well as the family and society indirectly?
Second are vaccine considerations: How effectively does it prevent disease, both through individual and community (herd) immunity? What adverse events does the vaccine cause, how often, and how severe? Do particular immunization strategies pose other possible adverse consequences, such as altering the epidemiology of the disease and postponing infection into adult life? Might the vaccine eventually eradicate the disease, nationally or globally? Is the price of the vaccine, the costs of its administration and program implementation compatible with our valuation of other health care goods and services?
These considerations of disease, therapy, and certain aspects of cost constitute the standard focus for cost-effectiveness analysis. As public health authorities, though, we must also weigh important ethical considerations.24–27 Personal liberty—particularly the freedom to refuse medical intervention—may conflict with the right of vulnerable children to be protected from harm. This conflict requires authorities to strike a balance when specifying the degree of coercion the state should exercise to enforce a specific immunization policy. Achieving such balance is difficult when members of society value immunization programs quite differently.
Equally important considerations involve matters of liberty and justice. Ideally the benefits of immunization—namely, protection from disease—should be equitably distributed across the population. For instance, in the United States we believe that no child should fail to benefit from a universally recommended immunization program because of limited access to care, poverty, or discrimination attributable to race or ethnicity. Ideally there should be a fair distribution of the burdens of immunization. No segment of society should be placed at heightened risk of suffering a vaccine-related adverse event or falling ill because of vaccine failure, nor disproportionately bear the costs of disease care, adverse event care, or the care provided though the immunization program as a whole. This concern for fair burden-sharing typically is posed as the question of ‘free riders’: should any child (or their family) be allowed to take advantage of a common good (in this case, community immunity against a disease) and potentially put that collective good at risk, even if at an individual level this course might make sense, albeit from a selfish perspective?28,29 At the same time, Americans who prize individual liberty bristle or rebel whenever any authority encroaches on personal freedom, such as occurs not only with laws requiring immunization but also statutes enforcing the use of bicycle and motorcycle helmets or automobile passenger restraints.
For childhood immunization programs and safety-promoting policies, these deeper themes of liberty and justice play out through the duties that our society entrusts to parents to promote the health and safety of their children and assigns to governmental and other agencies to safeguard the welfare of children. Acrimony often erupts when parental and societal views about these duties differ, or when parents disagree with the course of action mandated by the policy. A mandatory policy with high immunization rates, to cite a well-worn example, would protect those vulnerable children for whom immunization is contraindicated or simply fails to elicit protective immunity, but would do so by placing many individual children at a minute yet measurable risk of severe adverse vaccine events. As troubling, though less strident, is the common problem of an immunization program failing to distribute benefits equitably because of unequal access to vaccination. An optimal immunization policy process strives to minimize both the contentious conflicts and these quieter pervasive problems.
II. SPECIFYING THE OBJECTIVES
Having outlined and organized our concerns, we can now focus our thinking by transforming these concerns into the following 7 objectives for our immunization policy:
Minimize the deleterious consequences of the disease.
Minimize the deleterious consequences of the vaccine.
Optimize personal liberty to choose or refuse vaccination.
Maximize the just distribution of benefits and burdens across society.
Promote the duty of families to protect their child.
Promote the long-term duty of society to protect all children, now and in the future.
Use limited health care resources prudently.
Are these the only or the ‘right’ objectives? Certainly other worthy objectives exist. Identifying the objectives that matter most is the first of several steps required to make any policy decision. Each of these steps requires some value judgments, either explicitly or implicitly. Making these value-laden choices explicit has the virtue of facilitating debate. For example, the set of key objectives tackled by cost-effectiveness analysis typically includes only the minimization of deleterious disease and vaccine-related consequences and the prudent use of health care resources: these are not merely the most important considerations, they are the only ones. Given the incompleteness of current cost-benefit analyses, our diverse society might be well-served by debating the degree to which our final decision should be influenced by cost-benefit information to the exclusion of all other concerns.
III. ENVISIONING ALTERNATIVES
With a clearer sense of what we are seeking to accomplish with our immunization policy, we now should develop a fuller list of alternative policies. To simplify our discussion, we will consider a single characteristic of the policy—the force with which vaccination will be promoted—and envision in some detail 3 alternatives: immunization with the new Franklin Fever vaccine will either be mandatory, recommended, or elective. The mandatory system of immunization would require that all children be vaccinated against Franklin Fever on entering school; failing to do so either unwittingly or though conscious refusal would result in the unvaccinated child being excluded from school during outbreaks (which is the prevailing practice in the United States for enforcement of other mandatory immunizations).30 The recommended strategy would strongly encourage immunization, using public education and expert advice as the chief persuasive means of raising immunization rates. The elective policy would likewise use public education to inform parents, but make clear that the choice to immunize or not is completely at the parents' discretion.
Deciding whether the immunization policy will promote the vaccine as mandated, strongly recommended, or entirely elective is linked to other policy decisions: under each program enforcement scenario, who will pay for the vaccine: a central payer, multiple payers, or self-payment by families? Will a special fund defray costs to families unable to pay? Who will pay for vaccine-related adverse events, or for disease care among the voluntarily unimmunized? If the mandatory policy is chosen, will ‘philosophical’ exemptions be granted under special circumstances? If so, exactly what circumstances?31 Beyond exclusion from school (or even preschool child care), might this mandatory policy be enforced through restriction of welfare benefits to those families receiving public assistance, as some regions of the United States currently are doing?32 Alternatively, if the recommended or elective policies were selected, could immunization rates be increased through the effective use of financial incentives for providers33–35 or even parents? Would policy initiatives that enhanced access and reduced barriers to obtaining health care, or facilitated voluntary compliance with these recommended or elective vaccination guidelines suffice to raise immunization rates to desired levels?36
The point of tracing out these interconnected considerations is this: as we proceed in our evaluation of these policy options, we may uncover issues that warrant our returning to this step of envisioning alternatives, developing new options, or refining and enhancing existing ones. This capacity to learn and improve our options in a reiterative manner is critical, because the quality of our ultimate decision is limited by the best alternative we create.
IV. LINKING ALTERNATIVE POLICIES TO OUTCOMES
Next, we need to assess how the 3 policy alternatives would meet, or fail to meet, our stated objectives. We can do this by gauging the impact of each immunization policy option on 5 classes of outcomes: health outcomes for individuals and the population, ethical outcomes for individuals and the community, and total net costs to society. Arraying these outcomes into a flow diagram (Fig 1) helps us to see how they are interrelated, with process-oriented ethical outcomes occurring before health-related outcomes, and the collective outcomes resulting from the aggregation of individual outcomes. For example, if under an elective enforcement policy quite a few children were left unimmunized because of financial reasons, the program would not only have failed these children ethically; it also, by allowing them to remain susceptible to disease, would have failed to distribute benefits equitably, and would have raised to some degree the risk of epidemic disease. Conversely, a mandatory immunization policy that coerced certain children to be immunized might lower their individual risk of disease and share more fairly the burdens of maintaining a protective level of herd immunity, but violate their family's autonomy and place these children at some risk of experiencing an adverse event. These health and ethical outcomes, in turn, determine much of the health care and administrative cost outcomes of alternative immunization policies. (The price of the vaccine also determines the costs, but lies beyond the scope of this article.)
For many of these outcomes, surprisingly little data exist on which to build evidence-based answers. Considering this series of process and health-oriented outcomes, however, draws our attention to how our decision would be better informed if we had reasonable estimates of how the rates of immunization were likely to differ among mandatory, recommended, and elective immunization strategies, and how the marginal differences in coverage rates would translate into beneficial and harmful outcomes. Let us illustrate this point by returning to our Franklin Fever scenario. Suppose we estimated that an immunization program promoted by recommendation would result in 80% of eligible children vaccinated, whereas a mandatory program would achieve 95% coverage. Immunizing this additional 15% of children, we might further estimate, would diminish the annual number of cases of Franklin Fever by 100 000, but compel immunization on 500 000 unwilling participants. We would then be able to ask—under these hypothetical assumptions—whether our society should choose to immunize 5 children whose families are opposed to immunization to prevent a case of disease. Such information would advance our thinking beyond simply stating qualitatively how the programs differ, enabling us to measure the amount by which they differ—a move that will help us substantially when we come subsequently to examine tradeoffs.
V. ASSIGNING VALUES
Ethical analyses of health care programs usually wrestle with how to prioritize large overarching ethical objectives, such as whether securing the greatest good for the greatest number is to be preferred over protecting the rights of all individuals. In our scheme, however, the next step is smaller and more concrete, as we assign values to particular child health and ethical outcomes. Picking several important examples, we need to ask: How do we value differently, if at all, the loss of freedom when a family cannot afford to choose to have their child immunized (as might occur under an elective system) versus the loss of freedom that occurs when an unwilling participant is coerced into being vaccinated (because of a mandatory immunization policy)? Do ‘natural’ illnesses caused by an infectious microbe represent the same loss of value as a precisely equivalent degree of illness caused by a vaccine-related side effect, or is the vaccine-related morbidity somehow more costly? How should we compare the value of a case of disease prevented today from one prevented a generation from now? From a societal perspective, is a dollar in immunization-generated savings that returns to the pocket of a parent of equal value to a dollar that enters an employer's corporate coffer?
Many critics decry ever assigning such values, believing the judgments required are too subjective and contentious. However, every policy decision requires us to make such evaluations; the important choice is whether these evaluations are made implicitly as they are done today or in a more transparent, explicit manner. Returning to our previous example, cost-effectiveness analyses give health status and cost full sway, but largely omit concerns for respecting family decision-making autonomy and for distributing benefits and burdens fairly. Combining the private evaluation made by different policymakers into 1 ‘societal’ value is problematic—and in an objective formal sense perhaps even impossible.37 Nevertheless, every policy decision ultimately depends not just on information but also on an underlying structure of values and preferences that guide choices. Even if a ruckus ensues, we believe that a public debate about such values and their relative weights should be part of policy formulation to maintain the robust consensus required to support immunization policy.
VI. EXAMINING TRADEOFFS
We will now examine how the various policy options would or would not accomplish our objectives, constructing a table that arrays the alternatives across our objectives, then considering which alternative best addresses each objective (Table 2). Although judgments as to which options serve the various objects ‘best’ are debatable (and should be the subject of research and public dialogue), our major point is that this kind of table breaks down the much larger decision of ‘what policy to implement’ into more manageable smaller assessments, which highlight pivotal tradeoffs. Each assessment requires both factual information and value judgments. The minimization of disease-related injury versus the minimization of vaccine-related injury is 1 tradeoff dimension that at first glance has similar concerns, namely the minimization of harm. Informing the debate with numbers needed to treat for benefit and for harm might help clarify and thus promote consensus on this particular tradeoff. For each vaccine these considerations would differ, suggesting that a spectrum of policy enforcement strength is warranted, titrating the degree of coerciveness to the particular disease and vaccine-specific tradeoffs.
Underneath the debate regarding how to minimize various types of harm lies another tension, though, between promoting the just distribution of burdens and benefits and the protection of personal liberty—that is, quite specifically, for those who wish not to be immunized. Children who are left inadvertently unimmunized because of failure of poorly organized immunization programs (such as occurs under elective systems) represent instances of diminished family autonomy, not having had the chance to choose to immunize their child. Even if we wish not to call this lost opportunity a loss of freedom, certainly something of value has been lost. On the other side, compelling families to immunize their children against their wishes represents a clear loss of personal freedom. Evaluating these competing issues of justice and freedom, and striking a renewable and hence sustainable consensus, is a task as much of political dialogue as epidemiology.
These 2 levels of tension—one involving preferred health risks and the other involving civil liberties—raise the possibility of an additional ‘higher-level’ tradeoff between these different levels of concern. To address this tension, we should move beyond debating general philosophical questions or arguing over which objectives we care most about, and instead concentrate on how much the differing programs enhance or compromise each particular objective. Focusing on the amounts of benefit and harm at stake when choosing between options, we can make our value judgments more relevant to the policy decision by titrating a set amount of good against varying amounts of bad. Is a single case prevented worth 10 immunized unwillingly? Or is the threshold 100 or even 1000? What if we consider, on the benefit side, disease prevention through community-wide immunity for immunocompromised children for whom vaccination may be contraindicated; does our tolerance toward immunizing children of unwilling families go up, so that we might tolerate immunizing 5000 children unwillingly to prevent a case among these vulnerable children? Conversely, if we shift to consider preventing unwilling immunization as a benefit of a recommended immunization program, how many children are we willing to see be left unimmunized inadvertently to prevent an instance of unwilling immunization? Breaking down broad tradeoffs between different categories of concerns into a comprehensible series of smaller judgments clarifies our values and facilitates the dialogue about how to think about and make these complex tradeoffs, promoting a discussion that is itself a fundamental task for a transparent policy-making process.
VII. MAKING HEALTH CARE POLICY DECISIONS
Public health programs involve more than just issues of health. In recent decades the medical literature has reflected a societal emphasis on economic considerations, but public health is also a morally-laden medical venture. Concerns for individual liberty and social equity permeate public health policy, and should be incorporated into mainstream analyses of health care programs. Outcomes research must encompass these moral and political concerns. Worry that special interest groups might manipulate or abuse such considerations is likely well-founded; omitting moral considerations, though, will not protect against such abuses. Instead, leaving moral concerns as ‘gaps’ in our formal analyses of such decisions merely makes the policy-making process less transparent and the abuses harder to spot. Explicitness, a virtue of clearly-stated moral considerations and how they are to be measured, would help to foster constructive debate, which, in turn, may help to sustain the consensus required for effective public health programs.
What is the ideal immunization program? Certainly, no single answer exists. Yet our society still must decide a course of action, choose a vaccination policy, and pursue it. We intended this hypothetical case study to expose more clearly our areas of confusion and genuine disagreement. We believe that this framework, straddling the interface between moral and empirical reasoning, offers several key elements of a minimally sufficient public dialogue regarding vaccine policy. Such a dialogue must involve clinicians, public health authorities, legislators and the public, and must therefore take place not only in the deliberations of national committees but in the scientific and lay press, in the electronic media, and on the Internet. We believe that a broad dialogue is essential to sustain the societal consensus that empowered the immunization initiatives of the past half-century, and that only through such continuing dialogue can we be enabled to take full advantage of new opportunities to enhance public health through immunization in the century ahead.
Dr Feudtner was supported by the Robert Wood Johnson Clinical Scholars Program at the University of Washington and by Grant K08 HS00002 from the Agency for Healthcare Research and Quality.
We thank Noel Weiss, MD, DPH, Maria Silveira, MD, Thomas D. Koepsell, MD, MPH, and the anonymous peer reviewers for their thoughtful critiques, and Nancy Cochran for her assistance with manuscript preparation.
- Received July 17, 2000.
- Accepted November 30, 2000.
- Address correspondence to Edgar K. Marcuse, MD, MPH, Children's Hospital and Regional Medical Center, 4800 Sand Point Way NE, CH-03, Seattle, WA 98105. E-mail:
The authors alone are responsible for the facts, opinions, and conclusions expressed herein.
- Franklin B. The Autobiography of Benjamin Franklin. New York, NY: Barnes and Noble; (Originally published in 1791)
- Blake JB. The Innoculation Controversy in Boston, 1721–1722. In: Leavitt JW, Numbers RL, eds. Sickness and Health in America: Readings in the History of Medicine and Public Health. 2nd ed, rev. ed. Madison, WI: University of Wisconsin Press; 1985:347–355
- Geison GL. The Private Science of Louis Pasteur. Princeton, NJ: Princeton University Press; 1995
- Ten great public health achievements–United States, 1900–1999.MMWR Morb Mortal Wkly Rep. 1999;48:241–243
- Keeney RL, Raiffa H. Decisions With Multiple Objectives: Preferences and Value Tradeoffs. New York, NY: John Wiley & Sons; 1976
- Kenney RL. Value-Focused Thinking: A Path to Creative Decisionmaking. Cambridge, MA: Harvard University Press; 1992
- Hammond JS, Keeney RL, Raiffa H. Smart Choices: A Practical Guide to Making Better Decisions. Boston, MA: Harvard Business School Press; 1999
- Diekema D, Marcuse E; Ethical issues in the vaccination of children. In: Burgio G, Lantos J, eds. Primum Non Nocere, Today. 2nd ed. Amsterdam, Netherlands: Elsevier; 1998:37–49
- Ross LF,
- Lantos JD
- King S
- Hardin G
- Fine PE,
- Clarkson JA
- Orenstein WA,
- Hinman AR
- American Academy of Pediatrics, Committee on Bioethics
- Hillman AL,
- Ripley K,
- Goldfarb N,
- Weiner J,
- Nuamah I,
- Lusk E
- Arrow KJ. Social Choice and Individual Values. New York, NY: John Wiley; 1951
- Copyright © 2001 American Academy of Pediatrics