American Journal of Law & Medicine

In tepid defense of population health: physicians and antibiotic resistance.


Antibiotic resistance menaces the population as a dire public health threat and costly social problem. Recent proposals to combat antibiotic resistance focus to a large degree on supply side approaches. Suggestions include tinkering with patent rights so that pharmaceutical companies have greater incentives to discover novel antibiotics as well as to resist overselling their newer drugs already on market. This Article argues that a primarily supply side emphasis unfortunately detracts attention from physicians' important demand side influences. Physicians have a vital and unavoidably necessary role to play in ensuring socially optimal access to antibiotics. Dismayingly, physicians' management of the antibiotic supply has been poor and their defense of population health tepid at best. Acting as a prudent steward of the antibiotic supply often seems to be at odds with a physician's commonly understood fiduciary duties, ethical obligations, and professional norms, all of which traditionally emphasize the individual health paradigm as opposed to population health responsibilities. Meanwhile, physicians face limited incentives for antibiotic conservation from other sources, such as malpractice liability, regulatory standards, and reimbursement systems. While multifaceted efforts are needed to combat antibiotic resistance effectively, physician gatekeeping behavior should become a priority area of focus. This Article considers how health law and policy tools could favorably change the incentives physicians face for antibiotic conservation. A clear lesson from the managed care reform battles of the recent past is that interventions, to have the best chance of success, need to respect physician interest in clinical autonomy and individualized medicine even if, somewhat paradoxically, vigorously promoting population health perspectives. Also, physicians' legal and ethical obligations need to be reconceptualized in the antibiotic context in order to better support gatekeeping in defense of population health. The principal recommendation is for increased use of financial incentives to reward physicians for compliance with recommended guidelines on antibiotic prescribing. Although not a panacea, greater experimentation with financial incentives can provide a much needed jump-start to physician interest in antibiotic conservation and likely best address physicians' legitimate clinical autonomy concerns.


Although once viewed as miracle drugs, antibiotics (1) have turned out to be fragile weapons in the fight against infectious disease. Antibiotic resistance (2) undermines a drug's ability to treat illness. Problems with resistance can develop insidiously, as bacteria evolve, adapt, and otherwise change over time so that a medication previously thought useful in controlling the bacteria no longer proves effective. Antibiotic resistance menaces the population as a dire public health threat and costly social problem. The Institute of Medicine estimates that antibiotic-resistant infections generate costs as high as $4 to $5 billion per year in the United States. (3) Antibiotic resistance appears to be not only on the rise, but accelerating. (4) Alarming increases in infection rates have been observed for methicillin-resistant staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE), drug-resistant forms of bacteria associated with hospital-acquired infection. (5) Indeed, a recent government prevalence study suggests that over 18,000 patients die each year in the United States as a result of MRSA, exceeding the number of deaths attributable to HIV-AIDS, Parkinson's, emphysema, or homicide. (6) Moreover, some resistant forms of infection simply cannot be treated, as is evident with recent concerns about possible outbreaks of extensively drug-resistant (XDR) tuberculosis. (7) Because resistance can develop across as well as within different classes of drugs, new antibiotics may face increasingly limited time windows before running into resistance problems. (8) The apparent vanishing supply of effective drugs and risk of return to the pre-antibiotic era make this a crisis period for public health. (9)

Effective strategies for combating antibiotic resistance unfortunately remain elusive. This reflects the fact that antibiotic resistance has numerous, complex root causes, including: weak surveillance for resistance; (10) aggressive promotion of antibiotics by pharmaceutical companies; (11) lax infection control practices; (12) patients' irrational demand for antibiotics even when they may not be effective; (13) unwarranted clinical variation in the way physicians prescribe and monitor use of antibiotics; (14) inappropriate patterns of antibiotic use in agriculture and food-animal products that may impact human health; (15) and a possible downturn in discovery and commercial development of new antibiotics. (16) Recent reform proposals attracting considerable interest focus a great deal on supply side solutions. For example, the influential Infectious Diseases Society of America and other groups advocate tinkering with patent rights to provide pharmaceutical companies stronger incentives to discover novel antibiotics and to resist overselling their newer drugs already on the market. These strategies generally aim to encourage stockpiling of newer medications for future public health threats and to extend the useful life of antibiotics. (17) Congress has also considered legislative bills that would leverage intellectual property rights in this manner. (18)

This Article argues that a primarily supply side focus unfortunately misses the mark. Attempts to increase the general supply of effective antibiotics of course remain important. But reform proposals also need to do much more in directly confronting the powerful demand side influences. Ordinarily expected supply side effects do not always occur in the complicated health care market. Because of agency relationships and information asymmetries, physicians perform as the key intermediary, in a position to induce, control, or at least heavily influence, overall demand and utilization of many goods and services, sometimes seemingly irrespective of changes in supply. (19) Considerable attention needs to be paid, therefore, regardless of any changes in the supply of effective antibiotics, to the physician's necessarily significant, yet traditionally neglected and problematic, role in antibiotic conservation. It must be remembered that, as in many other areas of health care, the physician performs a critical role as gatekeeper to limited resources.

The ideal physician gatekeeper makes referrals and grants patient access to health care services and technology on a discretionary basis, considering effectiveness and cost as well as the patient's needs in an attempt to manage limited resources equitably for the benefit of patients as a whole. (20) Gatekeeping is often associated with health care cost control, particularly in managed care settings, a subject that continues to attract considerable debate and scholarly attention. (21) Yet there has been far less recognition in the academic literature of the important connection between physician gatekeeping and protection of public health. Indeed, in the antibiotic context, effective gatekeeping takes on increased importance for public health reasons because "[o]ther than certain vaccines, antibiotics are the only drug class whose use influences not just the patient being treated but the entire ecosystem ... with potentially profound consequences." (22) This arises due to the "antibiotic paradox"--prescribing an antibiotic can have dual, contradictory effects as it combats targeted bacteria while also possibly increasing selection pressures in the larger environment for bacterial strains that are resistant to that antibiotic, potentially jeopardizing the medication's effectiveness when used again for future health threats. (23) It remains critically important, therefore, that physicians prescribe antibiotics prudently, preserving the limited arsenal of effective medications for socially optimal use. Arguably, this becomes as important for public health purposes as ensuring that physicians use the drugs at all.

Dismayingly, the evidence indicates that physicians currently perform quite poorly as guardians of population health. The World Health Organization estimates that physicians in the United States overprescribe antibiotics by 50%, (24) as may occur when physicians recommend antibiotics for viral infections. However, the problems go far deeper than prescribing the drugs when completely ineffective. Even when the underlying illness is bacterially based, a surprisingly large degree of antibiotic prescriptions are inappropriate. (25) According to the Centers for Disease Control and Prevention (CDC), "inappropriate" prescribing occurs when antibiotic therapy is not limited to clinical situations where likely beneficial to the patient, the selected antibiotic will target the likely pathogens, and the antibiotic is used at the recommended dose over the recommended time interval. (26) Inappropriate prescribing also includes additional situations. A broad spectrum antibiotic may offer clear therapeutic benefit to the patient, but its use still considered inappropriate because narrow spectrum drugs could be used instead. (27) For example, a broad spectrum antibiotic such as azithromycin seemingly offers about the same cure rate for treatment of certain respiratory tract infections as narrow spectrum agents. When treating such infections, resistance considerations urge substituting the narrow spectrum drugs in order to best preserve a broad spectrum agent like azithromycin for future threats. (28)

At present, many physicians appear ill-equipped, ill-prepared, and unwilling to take on the public health responsibilities of antibiotic stewardship. This Article asks why. It also considers how health law and policy tools could favorably change the incentives physicians face for antibiotic conservation. At the outset, one must acknowledge that already overburdened physicians likely will not welcome more express obligations as guardians of population health. Antibiotic resistance presents a recurring, complex tension for physicians: how to act for the good of the larger community consistent with pursuit of individual health and obligations to individual patients. As has been seen with the many difficulties encountered in enlisting physicians in socially beneficial efforts to control escalating health care costs, many physicians, quite understandably, would prefer to evade complex gatekeeping responsibilities as "most physicians ... no longer aspire to the dual role of agent for society and for the individual patient ... [p]hysicians want to be on the side of their [individual] patients, advocating for more resources and better quality...." (29) Moreover, any attempts to encourage greater conservation by physicians will likely run up against powerful economic interests, especially pharmaceutical companies. Antibiotics represent a big slice of the health care market pie. At most hospitals, more than 10% of pharmacy expenditures are for antibiotics. (30) Antibiotics account for approximately 15% of the $100 billion spent annually on medications in the United States. (31)

However, as a pragmatic matter, the alternatives to physician-focused conservation approaches seem worse. As health care industry observers know all too well, numerous reform initiatives, such as institutional efforts to improve quality, have floundered because of failure to obtain the dedicated commitment of physicians in the clinical trenches. (32) Further, it becomes near impossible to separate responsibility for developing prudent resource allocation guidelines from the physicians who must implement such guidelines when caring for patients ease-by-case. (33) All roads, it seems, lead back to the physician as gatekeeper.

Scholars distinguish between individual health and population health and acknowledge "the inevitable tradeoffs that arise in trying to optimize health at both levels." (34) Individual health has been the focus of traditional medicine. In this sphere, the physician works to identify and treat causes of disease in a specific patient. The physician provides individually tailored treatment within the confines of the special doctor-patient relationship, where a norm of patient fidelity rules. Population health, in contrast, concerns the aggregate health of the many individuals in the community at large. Promotion of population health involves activities such as epidemiological analysis and infectious disease control that aim to prevent risk and illness within the entire community. Population health has traditionally been considered part of the public health care system comprised of governmental health care programs, safety net providers, and public health agencies such as the CDC. (35) Another possible distinction is that individual health tends to focus on curing disease, whereas population health concerns itself more with disease prevention. But it is often difficult to draw a firm line between individual health and population health, as antibiotic resistance demonstrates. Prescribing decisions made between individual doctor and patient can influence resistance patterns in the larger environment that impact population health. (36) Meanwhile, antibiotic conservation, a population health goal, depends upon coordination of physician prescribing behavior, an activity traditionally relegated to individual health.

This, at times, false dichotomy between individual health and population health unfortunately exacerbates the antibiotic resistance problem. The individual health paradigm continues to dominate considerations in the doctor-patient relationship and it seemingly disfavors antibiotic conservation and may even encourage inappropriate antibiotic prescribing. (37) Of course, population health concerns about antibiotic resistance should not always trump medical concerns about doing what is best for the individual patient. A delicate balancing needs to be done. Yet, at present, many physicians appear to significantly underweigh, or seemingly prefer to evade serious consideration of, the population health interests. (38)

In light of these considerations, this Article explores legal and policy options for improving physician stewardship of the antibiotic supply. While disease transmission and antibiotic misuse that occurs across borders undeniably must be considered in combating antibiotic resistance, this Article's focus is domestic, intentionally limited to antibiotic resistance in the United States. Because of differing health care delivery systems and the wide variety of factors that drive antibiotic resistance in distinct regions of the world, it is unlikely that a one-size-fits-all approach would prove an effective response internationally. As background, Part I explains why, from a biological perspective, physician prescribing decisions can intensify selection pressures for antibiotic resistance. Part I further considers how antibiotic resistance can partly be viewed as a negative externality, and antibiotics as common pool assets subject to tragedy-of-the-commons type problems, but why such economic analysis does not fully describe the many confounding variables of what is ultimately a complex, dynamic biological phenomenon.

Part II reviews the medical literature to demonstrate the breadth and depth of the physician gatekeeping problem. Many physicians routinely ignore recommended practice guidelines for antibiotic conservation. Moreover, physician surveys suggest that physicians consistently underweigh obligations to combat antibiotic resistance as compared to more traditional medical goals.

Part III asks why population health concerns exert only a weak influence, at best, on physician decision-making with regard to antibiotic therapy. Acting as a prudent steward of antibiotics for the benefit of population health often seems to be at odds with a physician's commonly understood fiduciary duties, ethical obligations, and professional norms, all of which traditionally emphasize the individual health paradigm. Meanwhile, physicians face limited incentives for antibiotic conservation from other sources, such as malpractice liability, regulatory standards, and reimbursement systems.

Part IV offers recommendations for improving physician gatekeeping. To have the best chance of success, initiatives will have to respect physician interest in clinical autonomy and individualized medicine, even while, somewhat paradoxically, vigorously promoting population health perspectives and greater standardization in prescribing patterns. Physicians' legal and ethical obligations need to be reconceptualized in the antibiotic context in order to better support defense of population health. While potential tensions exist as a matter of law and ethics, upon closer examination physicians seemingly should enjoy considerable latitude to engage in many forms of antibiotic conservation. The principal recommendation is for increased use of financial incentives that reward physicians for compliance with recommended conservation guidelines on antibiotic prescribing. Greater experimentation with financial incentives has the best chance of better aligning physician interests with otherwise neglected population health goals yet still respecting physicians' legitimate clinical autonomy concerns.



To appreciate why physician prescribing decisions matter so much for containing antibiotic resistance, it is necessary to briefly review the biological processes at play. The development of antibiotic resistance should not come as a surprise. Indeed, resistance problems date back to the first widespread use of antibiotics. (39) Antibiotic resistance results from a form of natural selection occurring at the microbial level. Bacteria with resistant genes may prove more fit by surviving exposure to a specific antibiotic. Resistant genes can be expressed when bacteria mutate or when latent genes with resistant qualities become expressed through reproduction. (40) Also, resistant genes may be passed on by conjugation, when genetic material is transmitted from one bacterium to other organisms. (41) Alternatively, bacteria can acquire new genetic material via bacteriophages through a process known as transduction. (42)

As resistant genes are expressed, they provide bacteria an ability to fight the select antibiotic in a variety of ways. An organism may develop the ability to neutralize the antibiotic directly. Alternatively, the bacteria can rapidly displace the antibiotic out before it takes effect. Or the antibiotic "attack site" changes so that the drug no longer affects the bacterial functions in the same manner. (43) In other cases, resistance occurs when the bacteria becomes impermeable to the antibiotic or through inactivation of certain enzyme activities. (44)

However, the emergence of resistant genes alone does not inevitably lead to a public health problem. The likelihood that a resistant strain of disease will take hold in an environment depends on whether an antibiotic acts as a strong selective agent. (45) Introduction of an antibiotic eliminates specific bacteria, but leaves resistant forms of the bacteria to grow and multiply. When an antibiotic acts as a strong selective agent, it disrupts the environmental background, favoring drug-resistant traits of the bacteria to replicate, and potentially leading to penetration in the environment of resistant disease strains. Typically, both the genetic factors and the antibiotic as selective agent must interact in this dynamic way for resistance to pose a true health threat to the community. (46) Accordingly, "the way that microbial agents are prescribed is a major risk determinant for resistance." (47)

Antibiotic resistance can develop in the general population through spread of community-acquired infections. It can also take hold in select institutional environments, such as a particular hospital or nursing home, through outbreaks of institutional-acquired infections that spread resistant strains among inpatients and staff of the health care facility. Whether an antibiotic will have a strong selection effect--either in the community/outpatient or institutional/inpatient setting--depends on several factors, including the amount of antibiotic used, the duration of use, the intervals between drug administration, the number of patients treated with the antibiotic, and the antibiotic's demographic influence. (48) Physician prescribing decisions directly impact all these elements. Moreover, the threat to population health can turn even on the prescribing actions of a few individual physicians. Antibiotic resistance tends to initially develop in local regions, such as in a single health care facility via institutional-acquired infection or in an otherwise relatively small geographic area, (49) environments where one or two physicians with sufficient patient load can introduce strong selection pressures. Once resistance strains of the bacteria are introduced, the threat of further spread to the larger population, through transmission of community-acquired infection and other pathways, remains very real. (50)

Once antibiotic resistance emerges, it is unfortunately not easily contained. Resistance effects "are almost inevitably cumulative." (51) An organism's development of resistance to one antibiotic seems to help it develop resistance to other antibiotics. (52) Exposure to a first antibiotic may select variants of bacteria that are much more likely to develop resistance to further drugs. (53) Meanwhile, a different kind of cross-resistance problem can also develop. Prescribing an antibiotic for one type of disease may cause organisms other than just the target bacterium to develop resistance to it, potentially undermining effective use of that antibiotic for entirely different illnesses in the future. (54)


Antibiotic resistance imposes high costs on the health system and society as a whole. Patients hospitalized with drug-resistant strains of infection face increased risk of death. (55) Estimates suggest that between 14,000 to 63,000 individuals die in the United States each year from drug-resistant infections acquired in hospitals. (56) The true mortality risk associated with antibiotic resistance is likely greater. For many sick patients, it becomes difficult to determine whether natural disease progression or a resistant strain of infection was the cause of death, making mortality risk estimates of antibiotic resistance inherently imprecise. (57) Also, medical records likely understate the number of patients hospitalized with resistant strains of infection because commonly used diagnostic codes may indicate the type of infection but do not always indicate when a patient has a resistant strain. (58) Further, estimates of the mortality risk may sometimes fail to account for occurrences outside of health care facilities, even though antibiotic resistance occurs in outpatient settings.

Apart from mortality risks, antibiotic resistance increases medical costs in a number of ways: longer hospitalizations; extra physician visits; testing for resistant strains; more follow-up care and surveillance; and higher rates of complications. (59) The FDA has conservatively estimated that approximately 150,000 hospital-acquired infections per year involve resistant strains of bacteria, leading to additional hospital charges of $375 million per year. (60) As a more specific example, it can cost one hundred times more to treat a patient with drug-resistant tuberculosis compared to a tuberculosis patient responsive to the standard medications. (61) Also, in the face of resistance problems, physicians may have to deploy second-line and third-line antibiotics that are more costly and that have stronger toxicity, further increasing medical care costs. (62) As for the outpatient setting, where good data is harder to obtain, one study suggests that the annual deadweight loss associated with reduced effectiveness of antibiotics due to resistance problems ranges from $378 million to $18.6 billion. (63)

Patients with resistant forms of disease can also have more significant problems concerning lost productivity and disability. The FDA estimates that extended hospitalizations due to resistant bacterial infections result in $43 million in lost productivity costs each year. (64) Also, the true costs of antibiotic resistance include the replacement costs arising from the need to develop newer classes of drugs to substitute for formerly effective antibiotics. Such replacement costs can be substantial and end up diverting limited resources away from other needed public health efforts. (65) In addition, given the high rates of inappropriate prescribing, "waste" costs alone could be significant. (66)

Perhaps the broadest picture of the true costs associated with antibiotic resistance comes from the Institute of Medicine's (IOM) 1998 study. Accounting for medical costs in the hospital and outpatient settings, as well as indirect costs such as lost productivity and disability, the IOM estimated the costs generated by antibiotic resistance to be as high as $4 to $5 billion per year. (67)


In economic terms, antibiotic resistance appears to be a classic situation of a negative externality. A patient's direct and immediate costs arising from inappropriate antibiotic therapy may be minimal, yet the introduction of the antibiotic increases selection pressures for drug resistance in the environment, potentially reducing the effectiveness of the medication for other health threats. Externalization of resistance costs in this manner can be expected to lead to inefficient, suboptimal levels of antibiotic consumption. In particular, more instances of "low value" antibiotic consumption will likely occur, where the marginal benefits to the patient do not justify the marginal resistance costs imposed on the population. In other words "what is individually rational, to use antibiotics whenever they might help even a little, is socially irrational." (68)

However, labeling antibiotic resistance as a negative externality problem does not fully capture the true complexities of the phenomenon and the resulting implications for policy. Complicating matters is that antibiotic usage also has positive externalities. (69) Appropriate antibiotic usage helps combat serious infections that otherwise could spread to the community, so policy responses must be careful to avoid curbing quite beneficial antibiotic use. Also, distinguishing between antibiotic consumption that is low value and of high value to the individual patient can often be quite difficult. (70) Further complicating the externality analysis is that physicians acting as proxies for their patients, not the patients themselves, ordinarily direct the course of antibiotic therapy. Thus, to internalize costs appropriately, physicians should theoretically bear some of the resistance costs associated with antibiotic consumption, or at least there should be coordination between the costs borne by patients and their physicians. The externality problem also becomes particularly difficult because of time spans involved; the true costs of antibiotic resistance may be borne by future generations of patients, yet future generations are not in a position to bargain for a current change in antibiotic usage patterns. (71)

Antibiotic resistance can likewise be viewed as a variation on the "tragedy of the commons" problem that occurs with shared resources. (72) Individuals deplete common pool assets inefficiently and suboptimally, such as the paradigm case of overgrazing that can occur in the common pasture, when no party has property rights over the assets and consumption decisions are made with little regard to preserving the assets for longer-term, socially more efficient use. (73) Individual antibiotic prescription decisions similarly can lead to socially inefficient, suboptimal depletion of effective antibiotics. However, there are differing views as to whether antibiotics are truly analogous to common pool assets. (74) Such resources, such as the common pasture or fishing stock, typically can be renewed and replaced over time. In contrast, the number of effective doses of an antibiotic may be finite due to biological inevitability, making antibiotics more appropriately characterized as exhaustible resources. (75)

Antibiotics actually may have paradoxical qualities, given the complex biological dynamics of resistance at play. They share renewability characteristics with common pool assets in some respects, yet they also are finite like exhaustible resources in other situations. (76) A further complication is that the supply of effective antibiotics can deplete in non-linear steps, as resistance can occur very quickly, in cyclical patterns, in varying phases of renewal and decline, or even when the antibiotics are not widely used. (77) In the end, antibiotic resistance does not fit neatly into any one model and resists easy categorization because it is ultimately a complex, dynamic biological process.


Many reform proposals attracting current political, legislative, and academic interest seek to address the diminishing supply of effective antibiotics through significantly changing patent system incentives. For example, the Infectious Diseases Society of America (ISDA) has advocated that pharmaceutical firms enjoy extended market exclusivity for certain medications. (78) Related supply side reforms recommend the use of "wild card" patent extensions. Under this approach, firms that develop and receive FDA approval for a 'high-priority" antibiotic would be allowed to extend the market exclusivity term for one of their other drugs, allowing a company to reap windfalls and perhaps lessening financial pressures to oversell the needed antibiotic early on during its patent exclusivity term. (79) Congress has also considered legislative bills that would leverage intellectual property rights in this manner. (80) Professor Eric Kades' proposal to lengthen patent periods for antibiotics to extremely long terms is representative of such supply side strategies. (81) Kades suggests that a longer period of time-limited monopoly will encourage pharmaceutical companies to postpone sales and reserve drugs for future public health threats, when even higher prices can be obtained. In Kades' view, firms with lengthier patent terms and enjoying monopoly power will more likely price their drugs in a manner that will discourage initial misuse and low value use, thus helping to preserve the medications' effectiveness for more socially optimal applications. (82)

Pragmatic considerations suggest that this current enthusiasm for such supply side approaches should be tempered. Among other concerns, it is questionable whether pharmaceutical firms, responsive to investors demanding short-term gains, will really be willing to delay aggressive promotion of newer antibiotics (and the resulting immediate profits) even with longer patent exclusivity terms. Also, even under the best incentive structure, there is no guarantee that biological innovation will be able to find new and effective replacement drugs. In addition, these reforms do very little to address inappropriate use of antibiotics already off patent and heavily in use on the market. (83) Moreover, cross-resistance problems between different antibiotics can seriously complicate these supply side strategies. Even if revised patent rights encourage a pharmaceutical company to reserve a perceived blockbuster drug and not rush to market it aggressively, inappropriate use of another antibiotic already on the market within the same therapeutic class may nonetheless create resistance problems for the new drug as well. To address cross-resistance problems effectively, it seems patent extensions would need to be imposed on sets of drugs within a therapeutic class that share cross-resistance effects, rather than on individual antibiotics. (84)

This could require more comprehensive changes to patent law and difficult coordination between competing firms that have property rights in different drugs.

In short, pragmatic considerations suggest there are limitations as to what patent-based, supply side approaches can effectively accomplish. Antibiotics remain a fragile societal resource, with great replacement and renewal difficulties. (85) Moreover, the current interest in supply side reforms detracts attention from physicians' pivotal role, as trusted agents and learned intermediaries for their patients, in influencing demand. Even if the supply of antibiotics is expanded, this only goes so far. Any advantages of increased supply can nonetheless be continually undermined by inappropriate management of the available drug arsenal. …

Log in to your account to read this article – and millions more.