Next steps
Download this section
Chapter 8

Ramanan Laxminarayan

When you want to cook a frog, they say, don’t throw it into boiling water—it will only jump out. Instead, place the frog in tepid water and, ever so slowly, increase the heat.

Much like the frog that is unaware that it is being cooked, our reaction to the antibiotic resistance problem has been to wait for a crisis before responding—but the frequency of resistance has been increasing slowly and steadily. When resistance reaches crisis levels, it may be too late. Meanwhile, thousands of people continue to die or suffer from a cause that does not show up on any death certificate. A crisis need not be a sudden, uncontrollable outbreak of a resistant pathogen. Many believe that the emergence and spread of deadly infections like community-acquired MRSA already constitutes a crisis. Perhaps we will see drug-resistant pneumonia and MRSA in large numbers of patients afflicted with avian influenza, or perhaps the prevalence of Clostridium difficile, which by itself is not a drug-resistant pathogen but whose survival and proliferation have been facilitated by widespread antibiotic use, will reach epidemic proportions. Many deaths during the influenza epidemic of 1918 are thought to have been caused by untreatable bacterial infections—bacterial pneumonia, and not just pneumonia caused by streptococci but also staphylococcus-associated pneumonia. The combination of today’s highly virulent MRSA with an outbreak of avian flu could have devastating consequences.

It is perplexing why so little attention is paid to finding solutions to the antibiotic resistance problem when it has such catastrophic potential. One is reminded of the years of neglect that led to the failure of the levees and the destruction of New Orleans during Hurricane Katrina. Even if policymakers are not motivated to act in preparation for such a medical eventuality, a more immediate concern—the increasing costs of health care and the consequent difficulty of bringing large numbers of uninsured people under the umbrella of pooled-risk financing—may spur action.

Regardless of when policy action is forthcoming, policymakers will need a playbook of carefully considered ideas. Our objective in writing this report has been to sketch the outlines of such a playbook, notwithstanding that more basic science and policy research may be needed on some of the ideas. A summary of policy actions, their pros and cons, and the actors involved is presented in Table 8.1.

This report has outlined a plan to change incentives to address antibiotic resistance in health care, not just in the immediate term (such as by changing Medicare reimbursement rules, subsidizing hospital infection control and diagnostics, or imposing stricter state standards for reporting hospital infections) but also in the longer term, to ensure a sustainable and affordable supply of antibiotics into the foreseeable future. After all, new drugs take at least 10 to 15 years to develop, and policies changing how antibiotics are used will take years to be implemented and have an effect on resistance.

Main messages

Our main conclusion is that antibiotic resistance is an important and growing challenge to health and health care systems. It raises the cost and lowers the effectiveness of health care in the United States and will have potentially serious consequences if not addressed now. Although the underlying causes appear to be, broadly speaking, overuse of antibiotics and inadequate hospital infection control, the deeper reasons relate to incentives. A policy solution will have to address incentives that affect how individuals, physicians, institutions, and pharmaceutical companies demand, use, and produce antibiotics. The changes in the behavior of humans must, in turn, effectively change the microbial world. These issues are not unique to antibiotics, however. Managing incentives is a challenge with the use of any resource, whether oil or fish, and the lessons learned in those contexts can be valuable here.

We have critically and objectively evaluated various policy options to address antibiotic resistance, and on the basis of this evaluation, we make five general observations about the policy solutions.

1. Policy solutions tend to focus either on changing incentives for how individual actors deal with antibiotic use or infection control (by changing how hospitals get reimbursed for hospital acquired infections, for instance), or on exercising federal or state government oversight (by requiring reporting of hospital infections, for instance). This report identifies the incentive problems associated with the latter type of regulatory policy and generally finds greater support for the former, the incentive-altering policies. Government action is needed but is more likely to be effective when focused on changing incentives (say, for new drug development) than on just mandating standards.

2. Much of the public debate on dealing with antibiotic resistance has dealt with lowering antibiotic use. There is a need to broaden the discussion of policy alternatives beyond simply educating health care providers to reduce antibiotic use. We know that antibiotic use leads to resistance, but it is unclear to what extent education alone can lower antibiotic use and how much this will slow resistance. Dealing with resistance will require careful rethinking and restructuring of the incentives for infection control within hospitals and vaccination policies in the community. Lowering antibiotic use involves a tension between what is good for the individual patient—important from the prescriber’s perspective—and what is good for the rest of society. Resolving this tension between sound medicine and sound public health is one feature of the problem; preventing the spread of infections and using a diversity of antibiotics, in contrast, are policy options that do not require balancing the individual and the public good.

3. Our policy goal should go beyond minimizing resistance, since that may be best achieved by not using antibiotics at all. Antibiotics serve a useful social purpose, but we have to balance the benefits of their use to individuals and to the rest of society (by lowering the chance that one patient’s infection will spread to others) against the costs that are largely borne by society in the form of lower future effectiveness.

4. Successful policy solutions should incorporate an understanding of ecology and evolutionary biology. A sustainable antibiotics policy must recognize that drug resistance and new drug development are two facets of the ongoing process of coevolution between humans and microbes. New drugs can provide a temporary solution only until microbes catch up through the process of evolution. Moreover, new targets for antibiotics may be increasingly difficult to find, and there may be cross-resistance between old and new antibiotics. Antibiotic efficacy is a renewable resource, but only on very long time scales. Meanwhile, policy must focus on extending the useful therapeutic life of existing drugs, and this requires a change in human behavior that leads to change in microbial communities. To be effective, policy must consider population biology and microbial community ecology, and this will require new basic research, including research to identify microbial interactions that can be exploited to manage resistance.

5. We need to integrate our thinking of supply-side and demand-side policy objectives. Efforts to protect new antibiotics from drug resistance by keeping them on the sidelines potentially reduce incentives for new drug development by the pharmaceutical industry. Similarly, having a supply of new antibiotics that are fundamentally different from existing drugs expands our options by lowering selection pressure for resistance to evolve to existing drugs. Solutions that focus only on the supply side or only on the demand side may be less effective in the long term than solutions that are mindful of the interrelatedness between how we use existing antibiotics and incentives to produce new antibiotics. Empirical research can inform our current understanding of which policy solutions are most likely to improve sustainable antibiotic use. Much of the discussion of ways to change incentives for patients, physicians, and other agents to behave optimally with respect to resistance is based on a theoretical understanding of economics and the law. However, there are knowledge gaps that prevent progression to an implementation stage.

Future policy research and dialogue

This report provides an objective evaluation of various policy alternatives, but the assessment is challenged by important gaps in our understanding of these alternatives. Our call for more data and research is not just a nod to the established norm; our goal is to provide the biological, medical, and economic analysis that can directly inform policy decisions. Although we have evaluated incentives and motivating factors from a theoretical perspective, policymakers will undoubtedly need stronger evidence to act on such policies as subsidizing infection control in hospitals. Policy research is needed to empirically test, using pilot studies and modelbased approaches, the effects of some of the more immediate solutions related to changing prescribing behavior and hospital infection control. Policy pilots will be important for determining the impact of greater cost-sharing for antibiotic prescriptions and patient outcomes, and for calculating the effect of subsidizing substitutes for antibiotics that relieve symptoms, thereby reducing antibiotic use. These studies will be useful in understanding what proportion of antibiotic use can be avoided without harming patient outcomes. Modeling will have to be used for other approaches, such as the overall economic impact of antibiotic use and better reporting of resistance levels in hospitals. A natural outcome of this research will be prioritizing policy changes and identifying those most likely to have a significant impact on resistance.

Going forward, it is important not just to engage in policy research but also to reconcile diverse viewpoints among the broad range of stakeholders, ranging from consumer groups and physicians to pharmaceutical companies and health insurers. All of these stakeholders are committed to a long-term future for antibiotics: after all, no one is better off with drugs that do not work. However, specific policy proposals may be more or less palatable to different groups, and it will therefore be important to engage multiple stakeholder groups, such as the Interagency Task Force on Antimicrobial Resistance, the Infectious Diseases Society of America, the Academy of Managed Care Pharmacy, the American Hospitals Association, and the Joint Commission on Accreditation of Health care Organizations, in an expanded, multidisciplinary consultation process to develop consensus around policy solutions that will have a significant impact on how we use and develop antibiotics.