The committee’s sole recommendation is that decisionmakers use, when possible, objective and quantitative tools, such as that developed for this report for assessment of vaccine development, to guide and inform their actions. This particular cost-effectiveness model can help the vaccine research and development (R&D) community think systematically about the investments it might make. The model can also be used by vaccine program policymakers to inform their decisions about investments in delivery programs. The model is not static and may be modified. As new data emerge, specific elements can be changed or new candidate vaccines can be assessed. Major components of the model can also be modified if the needs or interests of a user differ from those envisioned by this committee. Should individuals wish to modify the model, the committee urges them to study carefully the chapters on ethics (Chapter 6) and methods (Chapters 4 and 5) so that the model is not used for purposes for which it is unsuited and inappropriate.
In the course of developing and illustrating this model, the committee discussed general issues related to the funding of research, neglected opportunities for vaccine R&D, the qualitative judgments integral to this modeling exercise, and vaccine program concerns. The committee thus closes this report with a series of observations that it hopes are considered as seriously as the analytic model that was the focus of the project.
THE FUNDING OF RESEARCH
Vaccinology exemplifies the centuries-old experience that science moves ahead in ways that one cannot always predict. Since the publication of the 1985
Institute of Medicine (IOM) report on priorities for vaccine development, vaccines can now be envisioned for the treatment and prevention of diseases not previously considered to be potential vaccine targets: for example, therapeutic vaccines for noninfectious diseases such as multiple sclerosis and melanoma were absent from consideration in the 1985 exercise, and in the future, preventive vaccines for these diseases will also likely be studied. The role of hepatitis B virus in liver cancer was recognized in 1985, but the current report includes many more examples of vaccine-preventable infections as causes of chronic conditions; for example, hepatitis C virus infection and liver damage, including cancer; Helicobacter pylori infection and gastric ulcers and cancers; and human papillomavirus infection and cervical cancer. Furthermore, scientific studies are emerging indicating a role for infectious agents in the pathogenesis of coronary artery disease and in a predisposition to asthma.
Stable and sufficient funding of basic research by the federal government, the use of creative funding mechanisms, and the creation of alliances between the public and private sectors are crucial to ensuring that effective, safe, and needed vaccines will be carried through the development stage into licensure. Funding of basic research in fields such as immunology, virology, and micro-biology can also lead young investigators into more applied research on vaccines. In addition to basic research in molecular and cellular biology, progress in vaccine development and program implementation depends on research in fields such as epidemiology, health services research, health economics, human behavior, and even ecology. The lack of data and research in these fields, information that would have been useful to the committee in assessing disease burden, was surprising. In some cases, no significant new data had been published since that referenced in the 1985 IOM report on vaccine priorities, particularly national data on disease characteristics such as morbidity states and patterns of care.
R&D is an expensive enterprise currently supported through a natural and fluid mix of public and private funding. The federal funds used to support intramural research within government laboratories or dispensed by the extramural programs to researchers, most of whom are in academic institutions, have traditionally supported the vast majority of basic science research. New knowledge resulting from basic research is the essential first step that allows applied R&D to move forward into the private sector. Chapter 3 discusses how the lack of fundamental understanding of immune responses to Treponema pallidum, for example, led the committee to not consider the development of a vaccine against this agent, which is the cause of a very important public health problem, syphilis. Although it is sometimes difficult to demonstrate the benefits of investment in basic research with a direct link to a health intervention, the reader is referred to a classic paper for examples (Comroe and Dripps, 1976).
The National Institute of Allergy and Infectious Diseases (NIAID) of the National Institutes of Health (NIH) provides the majority of the approximately $250 million of public money spent annually on vaccine research (Mercer Management Consulting, 1995). Much of this is investigator-initiated basic research.
Some research, however is very targeted, and funding mechanisms such as Cooperative Research and Development Awards, sponsorship of centers for clinical trials such as the Vaccine Treatment and Evaluation Units, or support for acellular pertussis vaccine trials are used. Other NIH institutes, such as the National Cancer Institute and the National Institute of Child Health and Development, and other federal agencies, such as the Centers for Disease Control and Prevention (CDC) and the U.S. Department of Defense, also fund research related to vaccine R&D. Private philanthropic organizations such as the Rockefeller Foundation, the Burroughs Wellcome Fund, and the Josiah Macy Foundation also support basic and applied R&D related to vaccines.
Although private industry supports basic research, the most important role it plays is to assume the costs of applied R&D. The impetus for a company to invest in the development phase of a vaccine begins with the establishment of proof in principle, which is evidence that the vaccine could protect against disease. Such proof in principle results from the basic research findings of researchers funded by either public or private money. Another impetus is the potential of a return on investment by a company, which depends on the likelihood of product licensure, the market for that product, and the predicted costs of development and production. Manufacturer’s profits from the sales of existing vaccines contribute approximately twice the amount of money to R&D as the federal investment in R&D. Another important source of funding is risk capital invested in small biotechnology firms. Once private industries invest significant amounts of money in the development of a product, they stand to make or lose money on the basis of the quality of the product, the size of the market, the purchase price of the product, and the profit associated with the sale of the licensed product.
However, R&D opportunities frequently come to fruition only if the government or other nonprofit organizations leverage their resources in partnership with private, for-profit organizations to develop products or gather data in phase III clinical trials to support efficacy and safety claims sufficient for approval of the products by the Food and Drug Administration (FDA). Program staff at NIAID, for example, work to stimulate creative and targeted research and collaborations at critical periods in the natural life cycle of some products to keep the R&D cycle moving until private sponsors are prepared to take on the project.
Many of the challenges to commercial interests in vaccine R&D are fairly well documented (IOM, 1995; Mercer Management Consulting, 1995). It is widely believed that vaccines do not generate large profits—either as a percentage of revenue from the sales of an individual product or as a percentage of the total profit of the parent pharmaceutical firm. In contrast to many pharmaceuticals, which are sometimes taken several times a day for years, most vaccines are used only a few times in each person’s lifetime. Because of these market forces, the committee has heard compelling arguments that federal investment in vaccine R&D—as well as in fields such as health services research and health communication, which are necessary to understand how to
stimulate the appropriate use of vaccines once they are developed—is all the more necessary.
The cooperation and synergistic activities of the public and private sectors have led to remarkable successes. The development of the acellular pertussis vaccines is one such example. NIAID invested significant funds to support clinical trials of acellular pertussis vaccines, which are now licensed and recommended for use. Tensions, however, are inevitable and probably healthy. For example, the committee is aware of tensions between the needs of FDA for mechanisms that it believes are necessary for ensuring the safety and efficacy of the vaccines and the effect that some of these mechanisms have on the financial requirements for licensure. Examples of these discussed by the committee include the financial burdens of pilot production and the high costs of the complex clinical trials required by FDA to demonstrate efficacy as well as safety. Although the committee did not study in detail the relationship between public and private research on vaccines, it is clear that improvements can be made to foster collaboration when market forces cannot guarantee that a for-profit vaccine manufacturer will risk an investment in the development and manufacture of a particular vaccine or a particular type of vaccines.
The private companies involved in vaccine research, development, and manufacturing expect and are entitled to a reasonable return on their investments. In fact, market forces have led to the development of many vaccines, such as the Haemophilus influenzae type b (Hib) vaccine. The burden of disease caused by Hib infection was significant, the target population (the annual birth cohort of approximately 4 million infants) had regular contact with the medical community, vaccinations were already an integral part of health care for this population, and a guaranteed market was confidently predicted. The Hib vaccine was, in fact, recommended for routine use by major advisory bodies such as the Advisory Committee on Immunization Practices of CDC and the Red Book Committee of the American Academy of Pediatrics, and immunization with the Hib vaccine was required for entrance into most day-care centers.
NEGLECTED OPPORTUNITIES FOR VACCINE R&D
In an ideal world, every vaccine of medical or public health importance would be developed. It is not clear that 20 years prior to licensure of polio vaccines either the disease burden or the basic science knowledge required for vaccine development would have been sufficient for inclusion in a modeling exercise such as the one undertaken for this report. However, it most certainly would have been a compelling candidate for analysis by 3 to 7 years prior to licensure, owing to the increase in paralytic disease and to advances in tissue culture and virology. This underscores the need for a dynamic research program that is not limited to the candidate vaccines discussed in this report and for the dynamic use of a model for ongoing evaluation of R&D priorities.
Sometimes this development is impeded by other than scientific or technical obstacles. Development can be abandoned if a return on investment seems unlikely because the vaccine would not be used optimally due to a lack of resources available to purchase and administer the vaccine, or to low acceptance of a vaccine. Candidate vaccines intended primarily for use in less-developed countries often are not pursued due to concerns that they will be unaffordable. Additional impediments to a return on investment include a very small target population or financial risks for litigation to a manufacturer. Three specific examples are discussed: candidate vaccines to be used primarily in less-developed countries, in pregnant women, or in very small target populations.
The committee focused on the burden of disease in the United States. There are many diseases of great importance to other countries that do not currently pose a significant burden on domestic health. However, policymakers in this country should factor these diseases and the international burden to health into their decisionmaking for both parochial considerations related to the potential threat to the United States from new and emerging infections, as well as the altruistic considerations involved in aiding other countries.
The experiences of recent years regarding new and emerging infections can only serve as a stern reminder that the United States cannot afford to be complacent about the potential threat to the health of its population from infections currently of importance mostly to other countries. Infectious agents know no boundaries, and a threat to another country’s population today could be a threat to the U.S. population tomorrow. An example considered by the committee, but ultimately not included in the full analysis, is the threat in the United States of dengue hemorraghic fever. In addition, throughout the world the increasing levels of resistance of a variety of bacterial diseases to antibiotics make the need for disease prevention, including the use of vaccines, more compelling than ever. It is thus to the benefit of the United States to help protect the population of other countries from infectious diseases that could someday become a threat to its population.
In addition to the threat that emerging diseases could pose to the United States, it is incumbent upon the United States to assist in R&D for vaccines against these diseases, because without U.S. help, the products will not come to market. Developed countries frequently directly or indirectly assume the burden of financing advances that can benefit people in less-developed countries. The basic R&D for all vaccines used in the Expanded Programme on Immunization and the Children’s Vaccine Initiative was supported by the United States and other developed countries, and mechanisms are in place to use these vaccines to the benefit of less-developed countries in Africa and Asia. When the development of vaccines against diseases in the United States is compelling, as evidenced by the results obtained with the model whose development has been de-
scribed in this report, it would be prudent to at least qualitatively consider the nature of the international burden of disease potentially averted by use of the vaccine outside the United States.
Small Target Populations
Some infections threaten the health of only a small group of people, but that threat is nevertheless quite serious for that population. Chapter 6 discusses ethical considerations and provides some insight into how this situation can be considered. These groups can be identified by geography, age, or chronic health conditions. The committee felt that a Pseudomonas aeruginosa vaccine was not necessarily compelling for the general population, but that the threat to the health of people with cystic fibrosis was severe. Although the committee did not analyze the cost-effectiveness of a candidate vaccine for P. aeruginosa, it recognizes the benefit to be gained from a vaccine for that targeted group of people. The committee did analyze the cost-effectiveness of a candidate vaccine against two conditions of relatively low incidence but serious morbidity. These are the geographically-confined infections of Histoplasma capsulatum and Coccidioides immitis. As the results demonstrate (Chapter 4), these candidate vaccines fall into Level IV, despite the very serious morbidity associated with these infections. One significant difference in the consideration of a candidate vaccine for P. aeruginosa compared to one for H. capsulatum or C. immitis is that the health status of the target populations is very different. The impaired baseline health status of the P. aeruginosa target population would affect the cost-effectiveness ratio in ways that are unfavorable. Equity considerations, such as those discussed in Chapter 6, might have an important qualitative influence on policy-making regarding development or use of a P. aeruginosa vaccine.
As described in other sections of the report, concerns about liability have influenced vaccine R&D over the past two decades. The creation of the Vaccine Injury Compensation Program (VICP) successfully stabilized and encouraged the development of vaccines primarily intended for use in children. However, many existing vaccines are not covered by that program. The lack of compensation or indemnity against liability is perceived as a serious impediment to the development of some new vaccines; in particular, vaccines that would be beneficial if given to pregnant women. Vaccine manufacturers and other researchers interested in the use of vaccines directed against group B streptococcus and other infectious agents believe that some form of legal protection from lawsuits is imperative before these vaccines can be developed and licensed. The rationale for the immunization of pregnant women as a crucial strategy for reducing the rates of morbidity and mortality from group B streptococcus in both mothers and
infants is presented in other sections of the report. The committee notes here, however, that the analysis described in this report identifies many infectious agents whose disease burden could be prevented most effectively with a strategy of immunization of pregnant women, and it hopes that serious consideration will be given to addressing the significant impediment to all vaccine development brought on by liability concerns. The committee did not believe it was the appropriate group, however, to recommend a specific policy solution.
Models put a framework around previously incomparable data, and imperfect as the data can be, the committee nonetheless encourages the use of such evidence-based tools as aids, not mandates, for decisionmaking. If the results—the relative ranking of vaccine candidates—make intuitive sense (that is, if they conform to the informed judgment of the health care community) the model is probably correct. If the results are surprising (that is, a vaccine candidate ranks much higher or much lower relative to others than one would have predicted), a decisionmaker might ask if either the model or the data inputs are suspect. If not, then the model has been particularly useful. The experience in prioritizing reimbursement for health care interventions described in Chapter 6 is instructive. At times relative rankings run counter to the community’s beliefs, and thus, this model, like any model, should be viewed as malleable. There is always a role for informed judgment when deciding to what degree the results of a modeling tool drive policies, particularly when the limitations of the model have been made explicit. The 1985 IOM committee on vaccine priorities, in fact, included an acellular pertussis vaccine mostly because of the pivotal role that its development would play in increasing confidence in vaccine safety and ensuring a supply of vaccine. This was a qualitative consideration that the committee valued but that it could not enter directly into the model.
Ethical concerns are another such consideration. Health status measurements are discussed in detail in Chapters 4, 5, and 6. The committee used quality-adjusted life years (QALYs) in this exercise. The disability-adjusted life year is used by some researchers in these kinds of analyses, but for reasons explained in the chapters on methods (Chapter 4) and ethical considerations (Chapter 6), QALYs are a respectable and valid choice preferred by the committee. Although QALYs are a quantitative measure, they embody ethical considerations that cannot be directly quantified or weighted. The committee struggled with applying the health utility index in its QALY calculations, but in the end, it noticed remarkable inter- and intraperson consistency in the values obtained for similar health states. The committee expects and encourages continued research on this and other measures of health status.
VACCINE PROGRAM CONCERNS
The committee discussed barriers in addition to those identified above in the section on R&D. Vaccine delivery poses significant barriers to the effective prevention and control of infectious disease. Children in the United States can receive up to 16 injections and three oral doses of vaccine delivered against 8 infectious diseases before the age of 2 years. The rate of compliance with the recommended immunizations at 2 years of age is below that achieved for children a few years older due to compliance with vaccination requirements for school entry. Combination vaccines promise to reduce the number of vaccine doses that must be administered separately, but these will not be a panacea. The combinations will help to increase the level of acceptance and the rates of utilization of vaccines, but clinical trial design issues are not trivial. The committee hopes that the Federal government and state governments, the medical community, the public, and vaccine manufacturers carefully think about rational approaches to combination vaccines and vaccination schedules. Furthermore, noninjection routes of delivery (for example, oral, intranasal, or cutaneous routes of delivery) should receive serious consideration. At the same time, the committee knows that market forces and corporate alliances will influence the availability of combination products.
Despite the impediments of delivering so many immunizations during infancy, mandatory childhood immunization is a fundamental part of health care for children in the United States. Adolescents and adults have received less information about the importance of vaccines to protect their health and are less accessible than children to health care providers, especially for preventive health care services. Patient and provider education about the benefits of new vaccines will be crucial.
Vaccines are one of the few preventive measures that save money. It is not clear that people are willing to accept, use, and pay for vaccines that do not save money. For example, it is the opportunity cost savings for a parent who does not need to take time off from work to care for a sick child that has helped make the varicella-zoster virus vaccine a marketable preventive health intervention. Saving a child from illness and the very rare cases of death due to chicken pox was not enough to convince some in the medical establishment and some parents that the varicella-zoster virus vaccine was important.
The model described in this report demonstrates that not all vaccines will save money. Some new vaccines might be very expensive. However, the health benefits might still be compelling. Use of these vaccines will require a shift in thinking from an expectation that vaccines always save money to an acknowledgment that the health benefits of some vaccines might be worth the cost.
Many vaccines are not covered by health insurance, under either indemnity plans or managed care plans. Financial incentives might be crucial for encouraging the use of vaccines. The Vaccines for Children Program and other public health initiatives have helped provide childhood vaccines to those who can not afford them. Federal and state governments need to prepare now to work with
insurers, providers, and communities to ensure that all who need the many vaccines that will be developed in the next two decades can receive them.
However, the cost of vaccines to the individual and to insurers is not the only impediment to vaccine use. Vaccines, like other public health successes, such as clean water, fluoridation for the prevention of caries, and food safety measures are victims of their own success: people forget how dangerous vaccine-preventable diseases can be and become complacent. This false sense of security strikes individuals, communities, health care providers, and policymakers. It is not until the medical and public health systems fail and illness from infectious disease surges (for example, as a result of antibiotic resistance, nosocomial infections, measles outbreaks in the late 1980s, or food-borne illness) that society pays the price for interventions, such as vaccines, not yet developed, maintained, or implemented.
The committee urges careful consideration but not rigidity in the use of evidence-based approaches, such as the qualitative framework and quantitative model developed for this report, for prioritization of research, development, and use of vaccines, as well as other preventive and therapeutic interventions. The committee, while acknowledging the limitations of modeling exercises in general and of the one it developed and used in particular, does believe that modeling is useful and important when attempting to compare widely divergent vaccine-preventable conditions. It hopes that the inferences derived from the model will be useful to the vaccine science community, the vaccine manufacturers, and research and program policymakers.