The Current Situation and How We Got Here
THE HISTORY OF TUBERCULOSIS
Although accurate data that describe the incidence and prevalence of tuberculosis could not have been collected before the identification of Mycobacterium tuberculosis by Koch in 1882, “consumption” was well recognized, and deaths thought to be due to the disease were counted for a number of years before its etiology was determined (Koch, 1882). According to Thomas Young, writing in 1815, “The frequency of consumption in Great Britain is usually such that it carries off about one fourth of its inhabitants; at Paris the mortality by consumption has been estimated at one fifth; and at Vienna it is said to be one sixth of the whole. . . .” (Young, 1815). Young also noted that “it [tuberculosis] seems to be most frequent in poor countries, where the inhabitants are ill fed and thinly clothed.”
In the United States, tuberculosis accounted for a similar proportion of deaths in large cities during the mid-1800s: 24 percent in Providence, 23 percent in New York, and 15 percent in Philadelphia. As shown in Figure 2-1 , mortality rates were of a similar magnitude and followed a similar time course in the United States and Europe, peaking at approximately 400 per 100,000 population in Boston, New York, and Philadelphia combined (Dubos and Dubos, 1952).
The discovery of the etiological agent of tuberculosis provided the fundamental scientific underpinning for programs designed to prevent and control the disease. Organized comprehensive surveillance of mortality from tuberculosis began in the latter part of the 19th century and in the early 20th century although, as indicated in Figure 2-2 , data for various
geographical locales had been collected earlier. These mortality data, although crude, did provide important information about the magnitude of tuberculosis and demonstrated that, generally, rates were declining in advance of the application of any specific control measures. The progressive reduction in mortality rates was sharply and dramatically interrupted by World War I and less dramatically so but also clearly by World War II, as shown in Figure 2-2 .
Please refer to the page image for an unflawed representation of this content.
Obviously the equilibrium between man and the tubercle bacillus is very precarious. If war can so rapidly upset it, other unforeseen events might also cause recurrences of the tuberculosis epidemic in the Western World. (Dubos and Dubos, 1952, p. 196)
In the early 1900s the threat of tuberculosis caused by Mycobacterium bovis (one of three species in the Mycobacterium tuberculosis complex) was also addressed. The Tuberculosis Eradication Division was organized within the U.S. Department of Agriculture Bureau of Animal Industry in 1910 to organize a campaign for the control and eventual elimination of M. bovis in cattle to eliminate the threat to humans. This program required that all cattle herds be systematically tuberculin skin tested, that herds with infected cattle be slaughtered (with an indemnity paid to the owner), and that the premises be cleaned and disinfected after the infected animals were removed. This led to a very rapid reduction in the incidence of bovine tuberculosis. Currently, the positive reaction rate among cattle is less than 2 per 1,000 animals, with many of those reactions believed to be false positive. This program has also effectively eliminated tuberculosis in humans because of infection with M. bovis. In the late 1970s, an average of only 26 M. bovis isolates a year were reported among isolates sent to the Centers for Disease Control (now known as the Centers for Disease Control and Prevention [CDC]); this represents less than 1 percent of the total isolates sent to the CDC each year, and about 0.1 percent of the total number of tuberculosis cases reported each year (Good, 1980). The few new infections that occur seem to be associated with pediatric exposure to infected milk products outside the United States (primarily Mexico) (Danker et al., 1993) and from occupational exposures to infected game and zoo animals (Dalovisio et al., 1992; Fanning and Edwards, 1991; Nation et al., 1999; Thompson et al., 1993). Because of the low risk of disease following infection (Magnus, 1966; Moda et al., 1996), the occupational exposures are also unlikely to result in a barrier to elimination.
Countrywide mortality data for the United States were not available until 1933. Before that time data were collected from state “death registra
tion areas,” a project that began in 1900 and that comprised eight northeastern states plus Indiana and Michigan. The registration areas gradually expanded so that by 1933 the entire country was included. Tuberculosis morbidity data were also available for the entire continental United States beginning in 1933, when the National Tuberculosis Association (now the American Lung Association) began annual surveys. During the first 20 years of collection of countrywide morbidity data, both active and inactive cases were counted and, as shown in Figure 2-3 , the rates declined very little (Centers for Disease Control, 1978). However, beginning in 1953, shortly after the introduction of effective chemotherapy, incidence data began to be collected in much the same way as they are today, with only new active cases being counted. Coincident with the availability of antituberculosis drugs, the rate of decline in tuberculosis mortality rates increased and the incidence of active and inactive cases that had been nearly static for the previous 20 years began to decrease in parallel with the decline in the rate of active cases (Figure 2-3).
From 1953 to 1985 there was, in general, a consistent decline in incidence that averaged 5 to 6 percent per year, resulting in a total reduction of 74 percent, from 84,304 to 22,201 new cases (Figure 2-4). The exceptions to this trend were in 1975, when rates appeared to increase because of a change in reporting criteria, and from 1978 to 1980, when there was an influx of refugees from Southeast Asia, who at that time did not undergo screening or treatment for tuberculosis before they left refugee camps.
CURRENT STATUS OF TUBERCULOSIS
The leveling of the incidence rate curve that began in 1985 followed by the increasing case rates in the late 1980s and early 1990s represented the first true increase in incidence rates since case rate data began to be collected (Figure 2-4). The rate peaked at 10.5 cases per 100,000 population (26,673 cases) in 1992, a 12.9 percent increase over the 1985 rate of 9.3 cases per 100,000 population. The reasons for this increase have not been clearly delineated but are generally thought to be the result of several factors that are both biological and social (Bloom and Murray, 1992). These include the increasing prevalence of human immunodeficiency virus (HIV) infection, increasing rates of homelessness and incarceration, and, perhaps most important, a deterioration in the capacities of tuberculosis control programs to treat patients until cured. Although the natural tendency is for the numbers of tuberculosis cases and deaths from tuberculosis to decline over time, perturbations in various factors, as was seen during the world wars, can cause a reversal in the trend. The observation of René and Jean Dubos (see quote) was prophetic in this regard (Dubos and Dubos, 1952). Although many of the important factors that influence
the incidence of tuberculosis are beyond the control of the public health system, some are well within its control. Because of this tendency for tuberculosis to rebound, given favorable circumstances, the capacity to provide tuberculosis control services must be maintained even in the face of decreasing case rates.
The declines in the numbers of tuberculosis cases and in the case rates that began again in 1993 demonstrate the effectiveness of control measures. However, as will be discussed subsequently, considerably more money was required to regain control of the disease than would have been required previously to maintain control.
The regular decline in tuberculosis case rates that resumed in 1993, reached an all-time low of 6.4 per 100,000 population in 1999 (17,528 cases), for a cumulative 39 percent reduction from the incidence in 1992 (Centers for Disease Control and Prevention, 2000). Just as the reason for the increase in the number of cases was multifactorial, the decrease was as well (Frieden et al., 1995; McKenna et al., 1998). Largely, however, the reduction was caused by a comprehensive strengthening of control activities. A major factor in the resurgence of tuberculosis between 1985 and 1992 was a high frequency of transmission of M. tuberculosis within institutions including hospitals, correctional facilities, residential care facilities, and shelters for the homeless population (Daley et al., 1992; Edlin et al., 1992). Since then, improvements in the screening of new entrants to these facilities, more rapid diagnosis, the use of directly observed therapy to decrease infectiousness more rapidly, and more effective use of preventive therapy worked in combination to decrease the number of infec
tious sources and the potential for transmission. During this period of resurgence, the transmission of M. tuberculosis with rapid progression to clinical illness was an important occurrence that was demonstrated by several population-based studies that used molecular epidemiological analyses and that showed unexpectedly high proportions of cases caused by the same strains of the organism (Alland et al., 1994; Small et al., 1994). Conversely, as the number of cases again declined, clustering of cases decreased somewhat disproportionately, suggesting that the measures applied did in fact decrease the level of transmission of the organisms that resulted in rapid progression to disease (Frieden et al., 1995).
During the late 1980s, not only did the number of cases of tuberculosis increase but for the first time large outbreaks caused by M. tuberculosis isolates resistant to both isoniazid and rifampin (multidrug-resistant tuberculosis) were recognized (Edlin et al., 1992). A survey of cases from 1982 to 1986 showed that about 0.5 percent of new cases were resistant to both isoniazid and rifampin, but by 1991 the rate of multidrug-resistant tuberculosis had risen to 3.5 percent among all new cases. This increase was the result of a failure to maintain patient adherence with therapy, the use of inappropriate regimens for the treatment of tuberculosis, and system failures, especially slow reporting or the unavailability of laboratory results.
A number of steps were taken to address the problem of multidrug-resistant tuberculosis and these are detailed in the National Action Plan to Combat Multidrug-Resistant Tuberculosis (Centers for Disease Control and Prevention, 1992). Several interventions have been especially important: an emphasis on treating all patients with a four-drug regimen to prevent the emergence of multidrug-resistant tuberculosis in individuals with isoniazid-resistant disease, increased use of directly observed therapy and the application of other patient-centered approaches to promote adherence and completion of therapy, and improved laboratory performance by the use of rapid techniques for the identification of drug-resistant organisms. As a result of these and other measures, the proportion of new cases of multidrug-resistant tuberculosis decreased to 1.1 percent in 1998 (Centers for Disease Control and Prevention, 1999). However, nearly every state in the union has reported at least one case of multidrug-resistant tuberculosis, and it is likely that even if no new cases of multidrug-resistant tuberculosis occur, cases will continue to occur for decades when individuals with latent infection with multidrug-resistant tuberculosis bacilli develop active tuberculosis.
Multidrug-resistant tuberculosis is also a serious problem internationally. The Global Surveillance for Antituberculosis Drug Resistance project, a survey conducted jointly by the International Union Against Tuberculosis and Lung Disease and the World Health Organization found
multidrug-resistant tuberculosis in all 52 geographical settings surveyed (World Health Organization, 2000). The estimated proportion of multidrug resistance in previously untreated cases was 1 percent, but the proportions were as high as 10.8 percent in China and 9 percent in Latvia and Russia. The same survey estimated the rate of multidrug resistance among previously treated cases to be 11.1 percent. Given these rates of multidrug-resistant tuberculosis throughout the world and the relatively poor state of global tuberculosis control programs, it is highly likely that there will be continued increases in the proportions and a more widespread distribution of multidrug-resistant cases. Thus, multidrug-resistant tuberculosis presents a threat to the entire world if steps are not taken to cure existing cases and prevent the emergence of new cases.
Recent Incidence Data for the United States
Closer examination of the overall data for the United States in 1998 and 1999 (preliminary data from CDC) reveals a number of important features.
Tuberculosis increasingly has a heterogeneous distribution. About 75 percent of the new cases occur in the 99 metropolitan areas that have populations of more than 500,000 persons and that account for about 62 percent of the total U.S. population. Nearly half the counties in the United States had no cases of tuberculosis in 1998 (Figure 2-5), although the populations of these counties represented only 11 percent of the total U.S. population. Five states—California, Florida, New York, Illinois, and Texas—had 54 percent of the new cases in 1998, but the decreased incidences in those states also accounted for 68 percent of the overall decrease between 1992 and 1998.
Tuberculosis is increasingly occurring in foreign-born populations in the United States (Figure 2-6). Although from 1992 to 1999 the number of cases decreased 49 percent among individuals born in the United States, it increased 2 percent among foreign-born persons, and the proportion of individuals with tuberculosis who were born outside of the United States but who now reside in the United States increased from 27 percent in 1992 to 43 percent in 1999. Individuals from three countries—Mexico, the Philippines, and Vietnam—accounted for nearly half of the foreign-born individuals with tuberculosis (23 percent, 13 percent and 10 percent, respectively), with the remaining 54 percent coming from 151 different countries.
Tuberculosis is relatively common among homeless people or in individuals who reside in congregate facilities, correctional institutions, or long-term-care facilities. Data on the rate of tuberculosis among indi
viduals in these populations were not routinely collected before 1993 and were incomplete before the past 2 to 3 years; thus, comparisons of current and previous rates of tuberculosis among individuals in these populations cannot be made. However, nearly 15 percent of the new cases were reported among individuals from what might be considered high-risk environments. Homelessness was an especially prevalent factor among individuals with new cases of tuberculosis in 1997, with, overall, 6.5 percent of the 1997 cases reported as being homeless. In Oregon and the District of Columbia, approximately 17 percent of the new cases, the highest proportions reported, occurred among homeless individuals.
Substance abuse is common in individuals with tuberculosis. Alcohol abuse was reported in 16.1 percent of the individuals with tuberculosis in 1997, noninjection drug use was reported in 7.8 percent, and injection drug use was reported in 3.3 percent. Again, comparisons with earlier years are not possible.
The rate of drug resistance, and, particularly, the rate of resistance to isoniazid and rifampin (multidrug resistance), is decreasing. The rate of resistance to isoniazid among new cases decreased from 8.9 percent in 1993 to 8.1 percent in 1998, and the rate of multidrug resistance decreased from 2.8 percent in 1993 to 1.1 percent in 1998. New York and California reported 49 percent of the 150 multidrug-resistant cases. Despite the small number of multidrug-resistant cases, individuals with multidrug-resistant tuberculosis consume a disproportionate amount of the resources in the areas where they live.
Molecular epidemiological studies have demonstrated that even in efficiently administered tuberculosis control programs, transmission of M. tuberculosis from individuals who are acid-fast bacillus smear negative is the cause of at least 15 percent of cases.
Detailed epidemiological analyses such as those described above are essential for determination of the control interventions that have been effective and that are likely to be the most useful in the future. However, similar analyses must also be undertaken at the local level to determine precisely and thoroughly local epidemiological circumstances. In some instances conventional epidemiological analyses can be supplemented by molecular epidemiological analyses to develop an even more precise picture of the local epidemiological circumstances (Hopewell and Small, 1996).
Tuberculosis Control amid Worldwide Complacency
The complacency that was a major factor underlying the resurgence of tuberculosis in the United States, and in many other low-incidence
countries as well, was not a new phenomenon. However, the context in which it occurred—a low and progressively declining incidence of disease—was new. In the past the frequency of tuberculosis and, thus, the familiarity that physicians and public health officials have had with the disease have led them to conclude that all that needed to be known about tuberculosis was already known. This was explicitly described by Keers (1978) in his description of the early career of Robert Philip, who in 1887 developed the “Edinburgh Coordinated Scheme,” considered to be the first organized effort at the prevention of tuberculosis. According to Keers, “His [Philip's] seniors made it quite clear that they considered tuberculosis to be an exhausted subject. Everything that was to be known about it was already known and understood and every sensible person realized the hopelessness of even thinking about its prevention and treatment” (Keers, 1978).
Tuberculosis and HIV: A Deadly Duo
Please refer to the page image for an unflawed representation of this content.
Throughout both the industrialized and developing world, tuberculosis and HIV disease are closely linked in mutually disadvantageous synergy: HIV infection promotes progression of tuberculosis infection to disease and tuberculosis accelerates the course of HIV disease leading to more opportunistic infections and earlier death. HIV infection greatly increases the likelihood that infection with Mycobacterium tuberculosis, either recent or latent, will progress to active tuberculosis. In fact, HIV infection may be the most potent risk factor for tuberculosis yet identified. Conversely, tuberculosis is the most common cause of death in persons with HIV infection throughout the world. The extraordinary, deadly interactions of HIV and M. tuberculosis have been amplified by the rapid spread of HIV in populations in which the prevalence of tuberculosis infection was already high. Increasingly, therefore, the control of tuberculosis requires dealing with HIV infection, and vice versa.
Although there are many factors that influence the rate of tuberculosis in persons with HIV infection, data from a multicenter study in the United States showed that, overall, the incidence of tuberculosis was 0.7/100 person years of observation, or slightly less than 1 percent per year. Not surprisingly, rates were higher among those with more severe immune compromise. Among those who had positive tuberculin skin tests the rate was much higher—4.5 per 100 person years. These rates are far in excess of those that would be seen in any comparable group of nonimmunocompromised persons.
While it appears that the impact of HIV infection on the incidence of tuberculosis in the United States is declining, it is quite clear that HIV played a major role in the resurgence of tuberculosis during the late 1980s and early 1990s. At the height of the tuberculosis epidemic in New York City it was estimated that approximately 1/3 of the cases were HIV infected. HIV infection was also a driving force behind the rapid upsurge in cases of tuberculosis caused by multiple drug-resistant organisms (MDR) with much of the transmission occurring in hospitals.
Although the influence of HIV on the incidence of tuberculosis in the United States is decreasing, data from developing countries indicate a substantial and increasing rate of HIV infection among patients with tuberculosis. In 1992 the World Health Organization (WHO) estimated that worldwide there were approximately 4 million persons who were infected with both M. tuberculosis and HIV, nearly 80 percent of these being in Africa. Various studies have reported the prevalence of HIV infection among tuberculosis patients in sub-Saharan African countries to range from 20 to 67 percent. In 1998 the WHO and UNAIDS estimated that of a global total of over 30 million people living with HIV infection, 15 million were coinfected with M. tuberculosis, mostly in sub-Saharan Africa. Although this estimate is quite rough, it is indicative of the devastating impact that HIV is having and will continue to have on tuberculosis rates in developing countries.. In Botswana, for example, which has had a model tuberculosis control program using directly observed therapy (DOT) for more than 15 years, the incidence of disease has increased by more than 25 percent since 1995.
On a more positive note, data from many studies demonstrate that tuberculosis in persons with HIV infection responds well to treatment with standard antituberculosis drug regimens. Treatment regimens may, however, be more complicated to manage because of interactions between antiretroviral agents and the drugs used in treating tuberculosis. Even more encouraging is the fact that tuberculosis can be prevented in persons with HIV infection and positive tuberculin skin tests by using any of several proven preventive regimens.
HIV infection is now a fact of life in tuberculosis control. Clearly, any efforts to eliminate tuberculosis must take the effects of HIV infection into account. However, with good tuberculosis control and with effective antiretroviral therapy, the impact of HIV on the incidence of tuberculosis can be greatly reduced. During the past decade a great deal has been learned about both the nature and management of the deadly interaction of M. tuberculosis and HIV. This information must now be incorporated into effective interventions for tuberculosis control.
A somewhat paradoxical combination of complacency and defeatism led over the years to a near total dismantling of the World Health Organization's tuberculosis control program. The complacency was manifested as an attitude suggesting that all that was needed for control of tuberculosis was more assiduous application of existing tools at the local level. The defeatism was perhaps best expressed by Walsh and Warren (1979) who, in an analysis of “appropriate” primary health care interventions, concluded that tuberculosis control was too complex and costly to deserve being included as a high-priority activity.Funding and Mechanisms for Tuberculosis Control in the United States
With the global prevalence of attitudes and the consistently declining rates of tuberculosis in the United States, it is not surprising that funding
for domestic tuberculosis programs was progressively eroded to the point that by 1972, no funds from the CDC were specifically directed to tuberculosis control.
Beginning in 1970 the bulk of federal support for tuberculosis control was shifted from a categorical program “(i.e., funds for specific program categories, such as tuberculosis or sexually transmitted diseases), to block grants to the states for control of communicable diseases as a whole, and the states were not required to direct any of the federal funding to tuberculosis (U.S. Congress, Office of Technology Assessment, 1993). The formula by which the amounts of these grants were calculated tended to provide less money to cities of more than 500,000 population (those areas with the highest tuberculosis rates) than had been available under the categorical funding mechanism, also called project grants.” Within 2 years of the institution of block grants no categorical money for tuberculosis control was available from the CDC. To compound the problem, there was no way to determine the amounts of the block grants that the states were using for tuberculosis control. The general perception, however, was that funding for tuberculosis control was sharply curtailed.
Categorical funding reappeared via emergency grants in 1980, but as shown in Figure 2-7 , this amounted to only $3.6 million in 1980 and $3.7 million in 1981. It was not until 1989 that funding reached the level at which it had peaked in 1969, before the institution of block grants, which
was $20 million. The largest increase in federal funding occurred in 1993, the year after case rates peaked, and there was a nearly comparable increase the following year. For the past 6 years (1995 to 2000) funding has been essentially flat at $142 million to $145 million per year. When these amounts are adjusted for inflation, there has been the equivalent of an approximate 15 percent reduction since 1995, the year with peak funding.
Funding at the federal level is only a part of the total financial resources devoted to tuberculosis control. However, because of the highly variable ways in which tuberculosis control activities are supported within states, it is impossible to determine the total amount. In California, for example, funding is from a combination of state and county budgets, as well as from CDC. Other states fund the program entirely from state budgets, usually including CDC support. Large cities are usually responsible for their own programs and received no assistance from the state.
In addition to the support required for public health tuberculosis control programs, other costs arise from the clinical care of patients with tuberculosis. In a comprehensive analysis of the health care expenditures for tuberculosis in the United States, Brown and associates (Brown et al., 1995) estimated that in 1991, direct medical expenditures totaled $703.1 million. Of this, $423.8 million (60 percent) funded inpatient care, a cost not usually borne by tuberculosis control programs. The remaining $279.3 million was expended as follows: $182.3 million for outpatient care, $72.1 million for screening, 3.4 million for contact investigations, $17.9 million for treatment of latent infection, and $3.6 million for surveillance and outbreak investigations. These costs are generally borne by the public health system. In 1991, the year analyzed by Brown and coworkers but not quite the peak year for tuberculosis cases, the amount from CDC for tuberculosis control was $25,274,000, or approximately 10 percent of the estimated countrywide costs, exclusive of inpatient care. If the same relative apportionment of costs is true today, taking into account the fact that in 1998 the number of cases has decreased by 30 percent in 1998 compared to the number in 1991, the costs borne by public health programs for tuberculosis control should have been approximately $195.5 million. If this is true, the $144.5 million provided by CDC represents approximately 74 percent of the public health costs of the disease. Given that the 1991 cost estimates were derived in part from actual data, if relative activity levels have not changed, it would appear that federal funding has largely been used to offset, not to add to, state and local support. However, it is likely that the costs relative to the total number of cases have increased because of the increased use of directly observed therapy, more active screening programs, broader contact investigations, and wider use of treatment of latent infection, all of which are more labor-intensive and, therefore, more costly interventions.
A Dramatic Impact from an Imported Case of Tuberculosis
Please refer to the page image for an unflawed representation of this content.
Young children are rarely suspected of transmitting tuberculosis to adults. Child contacts of individuals with infectious tuberculosis are usually examined to determine if they have been infected by the adult index case patient and to determine if treatment is needed. This conventional wisdom was turned on its head when a woman in North Dakota, who happened to be the guardian of two brothers from the Marshall Islands, was found to have tuberculosis involving her left hip. The two 9-year-old children living with her were examined to determine if they might have been infected by the same source. To the surprise of all involved, one of the brothers was found to have pulmonary tuberculosis with bilateral cavitation on a chest radiograph and sputum specimens strongly positive on smears and by culture. The other brother had a strongly positive tuberculin skin test, a normal chest radiograph, and a single sputum specimen culture that was positive for M. tuberculosis. The isolates from the woman and both of the boys all had identical 20-band DNA fingerprints, leading to the conclusion that the 9-year-old child had infected his guardian. Household contact investigations and contact investigation at the primary school attended by the boy resulted in the identification of 120 contacts among 432 individuals screened.
North Dakota is a state with a very low incidence of tuberculosis, with a rate of 1.1 cases per 100,000 population in 1999. In the 5 years before the outbreak described above only one case of tuberculosis in a child under 15 years of age had been reported and only two cases of tuberculosis of any kind had occurred in the two county area in which the woman and the children lived. These cases were, in part, a failure to properly screen the two brothers when they entered the United States. Both received a multiple-puncture tuberculin skin test but the results were not read and they underwent no further screening for tuberculosis. This incident indicates the dramatic impact that a single imported case of tuberculosis can have on a low-incidence community, the importance of screening new arrivals from high-risk areas, and the importance of contact tracing and identification of the source patient. This incident also demonstrates how, as the incidence of tuberculosis declines, conventional wisdom concerning tuberculosis transmission will consistently be challenged and how creative thinking will be needed to identify contacts and manage outbreaks. SOURCE: Adapted from Curtis et al. (1999).
It is clear that as the number of tuberculosis cases decreases, the costs per case will increase, as long as an adequate infrastructure is maintained. The fixed costs of tuberculosis control are not sensitive to the numbers of cases. Although it is tempting to think that funding can decrease in proportion to the decrease in the number cases, below a certain point (number of cases), reductions cannot be applied without damaging the tuberculosis control infrastructure. In view of the natural history of tuberculosis, generally with a long period of latency between the time of infection and the occurrence of disease, and given that the country has not yet
seen the results of the wave of earlier infections, including infections caused by multidrug-resistant organisms that presumably occurred with the resurgence, there is an ongoing, long-term need to maintain effective mechanisms for tuberculosis control. Additional reasons include the fact that it will likely become more difficult to treat patients with tuberculosis because of their social situations and that treatment will require more labor-intensive schemes. Also given the current upheavals in the world, it is likely that the United States will continue to be the destination for large numbers of people from countries with a high incidence of tuberculosis. Even with effective screening, many of these new arrivals will be infected with M. tuberculosis. Thus, they may develop tuberculosis at any time in the future. Maintenance of Control amid Decreased Incidence
A major challenge that needs to be addressed now is how to maintain effective tuberculosis control as the incidence of the disease decreases. This is especially true in parts of the United States where, currently, there are very few cases and a comparably low prevalence of infection. For example, in 1997, seven states had less than 20 new cases each. Yet, outbreaks of tuberculosis may occur in low-incidence areas, such as the outbreak that took place in a rural area along the Kentucky-Tennessee border in 1995 (Valway et al., 1998).
That the downward trend in the number of cases of tuberculosis and the number of deaths from the disease can be reversed by external circumstances has been clearly demonstrated by data from earlier in the 20th century as well as the more recent resurgence. Likewise, the consequences of not having an effective control system in place when unanticipated events occur—for example, the HIV epidemic—recently have been experienced. Although reduced funding was not the only reason for the failure of the system in the late 1980s, it played a major role. It is now necessary to focus on maintenance of an effective control program as the incidence declines and with it the level of concern about tuberculosis.
THE CURRENT LEGAL SYSTEM
Writing in 1993, Lawrence Gostin noted that his survey of state tuberculosis control statutes found “the majority . . . remain antiquated and incompatible with modern standards of constitutional law and public health” (Gostin, 1993, p. 260). Indeed, many statutes were enacted at the turn of the century, and some were enacted in mid-century. Only 10 states had enacted statutes within the previous seven years. Most significantly, the older laws failed to accord individuals with tuberculosis the protec
tions that had increasingly come to characterize the standards applied to those who faced the possible deprivation of liberty. These older statutes had been subjected to judicial review, and characteristic of the deferential posture of the courts to public health enactments for much of the 20th century, they had passed constitutional muster. As late as 1966, a California appellate court had upheld the confinement of a tuberculosis patient pursuant to a law that provided virtually no procedural protections. In its ruling the court declared:
Health regulations enacted by the state under its police powers and providing even drastic measures for the elimination of disease . . . in a general way are not effected by constitutional provisions either of the state or national government. (Application of Halko, 54 Cal. Report 1966)
This broad deference to the legislature and to the exercise of public health powers would come to look archaic just a few years later as the jurisprudence of confinement underwent a radical revision in the wake of a series of far-reaching constitutional challenges to the power of states to confine patients with psychiatric disorders to mental hospitals. By 1979, Chief Justice Burger would state in Addington v. Texas:
The Court has repeatedly recognized that civil commitment for any purpose constitutes a significant deprivation of liberty that requires due process protection. Moreover, it is indisputable that involuntary commitment to a medical hospital . . . can engender adverse social consequences to the individual. Whether we label this phenomena “stigma” or choose to call it something else is less important than that we recognize that it can occur and that it can have a very significant impact on the individual. (Addington v. Texas, 441 US 418, 425–426 (1979))
Central to constitutional developments was the assertion that before an individual could be deprived of his or her liberty the state had to exhaust other means that were not as intrusive upon the right to liberty. In 1960, the U.S. Supreme Court held in Shelton v. Tucker:
Even though the governmental purpose to be legitimate and substantial, that purpose cannot be pursued by means that broadly stifle fundamental personal liberties when the end can be more narrowly achieved. The breadth of legislative abridgment must be viewed in the light of less drastic means for achieving the same basic purpose. (Shelton v. Tucker, 364 US 479, 488 (1960))
It was on the basis of such developments in mental health law, rather than as a result of challenges to the actions of public health officials in responding to tuberculosis cases, that it became possible to assert successfully that tuberculosis patients be accorded the procedural protections guaranteed by the Constitution. In 1980, in an appellate court decision
that upheld the procedural rights of a tuberculosis patient, the Supreme Court of Appeals in West Virginia articulated a standard that expressly followed the developments in mental health law (Greene v. Edwards, 263 S.E. 2d 661 (W.Va. 1980)). The state's tuberculosis control act was ruled unconstitutional because it did not guarantee the right to counsel, did not provide for the right to cross-examine, confront, and present witnesses, and failed to hold the state to the stringent “clear and convincing” standard of proof required by the U.S. Supreme Court.
The social transformation in the legal and political context within which issues of confinement were considered in the 1970s and 1980s shaped policy and practice in the United States. In 1993, when the Advisory Council for the Elimination of Tuberculosis (ACET) recommended changes in state tuberculosis control laws, it declared
As in commitment proceedings under state mental health laws, any law under which a person may be examined, isolated, detained, committed and/or treated for TB [tuberculosis] must meet due process and equal protection requirements under state and federal statutes and constitutions. Also, all patients who are subject to these legal proceedings should be represented by legal counsel. (Centers for Disease Control and Prevention, 1993)
Reflecting the doctrine of the least restrictive alternative, the Council recommended, “Before committing TB [tuberculosis] patients for in patient treatment, states should adopt step-by-step interventions beginning with DOT [directly observed therapy] and supplemented by incentives and enablers” (Centers for Disease Control and Prevention, 1993, p. 8).
The Council's incorporation of both procedural due process protections and the doctrine of the least restrictive alternative into its recommendations was especially crucial, because it was calling for the expansion of existing tuberculosis laws to “permit the involuntary isolation and detention of non-infectious patients” who “refuse to adhere to a treatment regimen or to complete treatment” (Centers for Disease Control and Prevention, 1993, p. 8).
This expansion on the conceptions of who posed a threat to the public health, driven by concerns about multidrug-resistant M. tuberculosis, represented a move of great significance. No longer did the person to be confined have to represent an immediate threat of transmission. Rather, it was the prospect of reactivation and the prospect of the development of drug resistance that provided the grounds for state intervention. Here the concept of “threat” was informed by the population-based concerns of public health. It was concern about the collective consequence of permitting many individuals to conduct themselves in a way that posed some threat that motivated the extension of public health powers to reach pa
tients who had begun therapy for their tuberculosis but who would not or could not complete their treatment. It was inadequate treatment that posed the threat of multidrug-resistant tuberculosis. This was, however, a calculus far different from one that would center on the potential risk posed by a given individual. Nevertheless, legal commentators have, by and large, agreed that confinement until cure would probably be found constitutional for noninfectious patients who did not adhere to treatment (Bayer and Dupuis, 1995; Gostin, 1995).
A striking reflection of the expansion of the concept of what constitutes a public health threat was the adoption in 1993 of treatment-until-cure regulations by the New York City Board of Health (Ball and Barnes, 1994). The board had concluded “[f]or patients who are unable to complete treatment [until] cure the temporary detention . . . until they are non-infectious is an ineffective public health strategy.” Under the newly enacted Section 11.41 of the city's Health Code, the Commissioner of Health could confine an individual for whom there was a “substantial likelihood, based on such person's past and present behavior, that he or she cannot be relied upon” to complete treatment. Under the amended code, a court review had to be accorded to a confined patient within 5 business days, even if the individual did not request release. The Department of Health had to seek judicial review within the first 60 days of detention and subsequently at 90-day intervals. Upon requesting release, the confined individual was entitled to a lawyer and, if too poor to afford counsel, had to be provided with one at public expense. Maryland, too, extended the authority to quarantine tuberculosis patients in the “non-communicable stage” who “refused to receive sufficient therapy” (Rothenberg and Lovoy, 1994). California extended quarantine authority to non-infectious patients in 1993 and incorporated this into the California Health and Safety Code 121365 in 1994.
Among the controversial elements of the revised New York City code was the way in which the requirements of the least restrictive alternative standard were to be applied. During the hearing process that preceded the adoption of the amended code, civil liberties advocates claimed that the proposed new regulations were unconstitutional because they did not require the city to “exhaust” all less restrictive alternatives before seeking the involuntary confinement of a persistently noncompliant tuberculosis patient.
This broad interpretation of the least restrictive alternative doctrine formed the basis for a constitutional challenge brought against New York City's Health Department in early 1994 (In re Application of the City of New York v. Doe No. 40770/94 (1994)). In that challenge, it was argued that a distinction should be made between the latitude that should be available
to the Commissioner of Health when confronting a patient with infectious tuberculosis and a noncompliant patient who was not contagious. The distinction between contagious and noncontagious patients was precisely the distinction that the public health authorities concerned about the rise in drug resistance had sought to erase because of the importance of “treatment to cure,” regardless of current infectiousness.
Both the extension of procedural protections and the extension of the scope of tuberculosis control to those who are noninfectious have important implications for a policy that is directed at tuberculosis elimination, a centerpiece of which will be identification and treatment of individuals with latent infection. The courts could hold that mandatory screening and treatment represent an unconstitutional extension of the police powers not justified by the level of risk posed by individuals who might decline therapy for latent tuberculosis infection. On the other hand, they could find compelling a public health finding that the elimination of tuberculosis justified the compulsory treatment of latent infection just as the threat of multidrug-resistant tuberculosis justified the imposition of the requirement that those with tuberculosis complete their therapy even though they no longer posed an immediate public health threat.
In commenting on this tension, Lawrence Gostin has noted
While traditional public health law inquiries focus principally on present infectiousness, there is no reason to limit the direct threat doctrine in this way. . . . Direct threats therefore ought to include significant risks that are reasonably foreseeable. . . . After all, a health department's duty to protect citizens from the risk of foreseeable harm is as strong as its duty to protect citizens from the more imminent risk of infectious transmission. . . . If the state can demonstrate through objective data that the person is likely to develop or reactivate clinically infectious tuberculosis then there is no reason why the state cannot intervene to prevent the future risk to the public. For example, the development of clinically infectious tuberculosis in a person dually infected with HIV and MTB [M. tuberculosis] presents a significant risk to fellow residents in a congregate setting. The risk justifies requiring the completion of a course of isoniazid preventive treatment. (Gostin, 1995)
The screening of high-risk populations for latent tuberculosis infection and the provision of treatment of latent infection have been recognized for more than a decade as being central to the goal of tuberculosis elimination. First priority in such efforts must be given to those at highest risk—those recently exposed to tuberculosis, contacts of individuals with suspected or confirmed cases of tuberculosis, and patients with or at risk for HIV infection.
THE REALITY OF ELIMINATION
“Elimination,” as applied to infectious diseases, has been defined in many different ways, including reduction of the incidence, prevalence, morbidity, mortality, or transmission of a disease or of the pathogen that causes it. Often, especially since the successful eradication of smallpox, the word “elimination” has been used deliberately to invoke the cachet of eradication in circumstances in which the goal of true eradication is known to be unattainable. Such use of the term “elimination” is thus designed to mislead by implying one thing, that is, eradication, while really denoting something else, namely, a goal that is different from eradication. Even the potential distinctions between “elimination” and “eradication” in English are lost in some other languages that use only one word that means eradication and that is equivalent in meaning to both English words. Inappropriate, indiscriminate, or promiscuous use of “elimination” and “eradication” causes confusion and risks devaluing the legitimate concept of disease eradication among the general public and lay policy makers.
In 1997, a group of scientists who met in Dahlem (Berlin), Germany, for a week to consider aspects of the eradication of infectious diseases proposed that elimination of disease be defined as “reduction to zero of the incidence of a specified disease in a defined geographic area as a result of deliberate efforts; continued intervention measures are required” (because the same disease might still exist somewhere else). Elimination of infection was defined as “reduction to zero of the incidence of infection caused by a specific agent in a defined geographic area as a result of deliberate efforts; continued measures to prevent the reestablishment of transmission are required.” Eradication was defined to indicate “permanent reduction to zero of the worldwide incidence of infection caused by a specific agent as a result of deliberate efforts,” whereas control referred to “reduction of disease incidence, prevalence, morbidity or mortality to a locally acceptable level as a result of deliberate efforts” (Dowdle and Hopkins, 1998). The scientists at Dahlem did not evaluate various infectious diseases as to their eradicability or whether they could reasonably be targeted for eradication, elimination or control.
In 1989, the CDC and the ACET published a strategic plan that articulated a national goal to reduce the incidence of tuberculosis to 3.5 cases per 100,000 population by the year 2000 and to less than 1 case per 1 million population by 2010. This goal was called “elimination of tuberculosis” (in the United States). Although the stated goal qualifies use of “elimination” by specifying a quantified target level of disease incidence, that stated goal does not meet the criteria for elimination suggested by the Dahlem Conference. By the Dahlem Conference criteria, the CDC-ACET target is a goal of tuberculosis control.
Ironically, the CDC-ACET goal was promulgated during a period when case rates were not declining, and in the following year an unexpected upsurge in the incidence of tuberculosis ensued in the United States. Instead of the steady decline in the number of cases by about 7 percent per year that had occurred before 1985, the number of new cases of tuberculosis in the United States increased from 22,201 in 1985 to 26,673 in 1992. Since 1992, however, tuberculosis incidence in the United States has resumed its decline, again at a rate of about 6 percent per year, to 17,528 new cases in 1999, or 6.4 cases per 100,000 population. Thus, the 1999 U.S. incidence (6.4 per 100,000 population) is almost twice the stated goal for 2000 (3.5 per 100,000 population). At the current rate of decline, the U.S.'s case rate of tuberculosis would reach the stated goal for 2010 (1 per 1 million population) only after about 60 more years. If the current rate of reduction in annual incidence were doubled, which is imaginable by use of current tuberculosis control tools, it would still take about 27 years to reach the target of “elimination.” Achievement of the current target of 1 case per 1 million population in the United States by the target year of 2010 would require an average annual rate of decline of 32 percent.
The current Institute of Medicine committee endorses the general concept of “elimination” of tuberculosis in the United States as rapidly as feasible. However, the committee stresses the urgent need to accelerate the annual rate of reduction in the incidence of the disease by better use of the tools that are already available and by development of more effective tools to reduce the risk of transmission of the infection and to reduce the risk of development of the disease by those already infected in the United States. Accelerating the annual rate of decline of tuberculosis in the United States also requires that more attention be given to assisting with tuberculosis control efforts in a few countries that are the sources of the most cases imported into the United States. Any relaxation of current efforts against tuberculosis in the United States risks another resurgence in the disease and the potential loss of hard-won gains. Acceleration of the rate of reduction in the incidence of new cases is important for the following reasons:
It would reduce the suffering of individuals with tuberculosis.
It would reduce the costs of medical care for such patients.
It would reduce the risk of resurgence of the disease.
It would reduce the risk of multidrug-resistant cases.
It would reduce the risk of spread of cases to tuberculosis-free areas of the United States.
It would reduce the complications caused by tuberculosis in HIV-infected individuals.
It would lead and contribute to reduction in the global burden of tuberculosis.
It would hasten the eventual elimination of tuberculosis from the United States.
Successful eradication programs have used a few key indicators to monitor progress toward their ultimate objective. The committee recommends the use of a few such key indicators to monitor and to regularly report on progress related to defined, quantitative targets and interim benchmarks for the tuberculosis elimination program. The development of these indicators will require a better understanding of the epidemiology as it approaches elimination.
THE ETHICS OF TUBERCULOSIS ELIMINATION
With the incidence and prevalence of tuberculosis declining to the lowest levels in history in the United States and with the burden of tuberculosis continuing to exact an enormous toll in less developed nations, the ethical challenges posed by the goal of tuberculosis elimination are neither simple nor straightforward. In the United States, federal, state, and local governments have the moral duty to fund tuberculosis control programs so that every case of tuberculosis is promptly identified and every patient is treated until he or she is cured. The experience of the past decade has demonstrated the price to be paid in morbidity and mortality when such efforts are less than adequate. There is no justification for repeating the cycle of concern and neglect. The past decade has also demonstrated the necessity of using a range of strategies to ensure that treatment completion rates of 90 percent and better are sustained. Both ethical principles and constitutional standards require that the least restrictive measures that can be effective be used to achieve that end. Social supports and incentives for those patients are essential. Recourse to liberty-depriving measures may be necessary but should be viewed as a last resort and should be relied upon only after a hearing before a tribunal where the rights of the individual are given due accord.
The goal of tuberculosis elimination, however, will require more than case finding and treatment. It is now clear that an increasing focus on the identification and treatment of individuals with latent infections will be essential. It is in this regard that the ethical challenges will be new and difficult. How should individuals with latent tuberculosis infection be identified? From an epidemiological perspective, it is clear that targeted screening efforts are more efficient than those that cast a net that includes individuals at very low risk. However, can targeting efforts be undertaken without the taint of stigmatization? This is a question that is bound
to arise since those at greatest risk are also poor, vulnerable, and marginalized.
More difficult still is the question of whether such screening should be voluntary, undertaken with the cooperation of various community-backed organizations that can effectively encourage those at risk to be tested for latent infection, or should be compulsory.
There is little question that in selected settings screening for active tuberculosis can be undertaken on a compulsory basis. The constitutional authority granted under the police powers of the state has been used to justify such efforts. From an ethical perspective, the concept of the harm principle justifies state intervention when an individual places others at risk. Are these doctrines and principles, however, sufficiently robust to justify mandatory screening for latent infection that poses only the prospect of a public health threat? Are the constitutional doctrines that, since early in the 20th century, have recognized the authority of the state to impose vaccination requirements to prevent disease sufficiently robust to justify mandatory screening for those at risk for disease?
Finally, there is the question of whether those who are identified to have latent infection should be encouraged to undergo treatment or should be required to do so. The answer to this question very much depends on the risks involved in both the decision to treat and the decision not to treat and the benefits to both the individual and the public health from the initiation and completion of treatment of latent infection. The consistency with which treatment of latent infection is now recommended makes it clear that, with the exception of those in whom treatment is clinically contraindicated, the weight of medical authority now supports treatment. Even when treatment is clearly beneficial, however, the traditions of informed consent derived from constitutional doctrines and the ethical principle of autonomy typically dictate that competent adults be free to choose whether or not to undergo therapy. Even in the context of those committed to a hospital on psychiatric grounds, the courts have held that the decision to impose therapy against the wishes of the patient requires a special hearing.
On the other hand, the courts have long recognized the right of the state to require treatment for infectious diseases such as tuberculosis that pose a public health threat. Furthermore, in the early 1990s, a number of jurisdictions recognized that the authority to treat infectious tuberculosis was sufficiently robust to justify the requirement that patients in whom tuberculosis was no longer infectious undergo treatment until cure. Does the extension of the scope of the state's public health authority to individuals who pose only a potential threat provide the justification for mandatory therapy for individuals with latent infection? Do different contexts defined in terms of a heightened risk of transmission, were actual disease
to develop, help to identify those settings within which mandatory treatment for latent infection would be justifiable? Do, for example, the unique features of homeless shelters or prisons justify mandatory treatment that would otherwise raise constitutional and ethical questions? As discussed further in Chapter 4 , the conclusion of this committee is that the answer to all of these questions is yes. However, the committee was also cognizant of the social vulnerability of these populations and the need to protect against unnecessary coercion of the powerless.
The issues of screening and treatment for latent infection compel society to address the questions of when, if ever, it is appropriate to use compulsory public health powers and how to balance the collective well-being against the right of the individual to be free of intrusions when the threat that he or she poses to others is only statistical. These issues take on special significance in the context of the goal of tuberculosis elimination.
The ethical challenges posed by the continued burden of tuberculosis in less developed nations are very different. It is possible to argue that, on the grounds of narrow self-interest, the United States and other developed nations should be concerned with tuberculosis abroad because it is not possible to erect a protective cordon sanitaire. This is especially the case with tuberculosis at the border, as is the case with Mexico and the United States. To so frame the issue, however, is to unduly restrict the U.S. moral vision. Eight million cases of tuberculosis annually worldwide and 2 million to 3 million deaths a year demand attention, regardless of their ultimate impact on American well-being. The United States shares with other developed nations the obligation to shoulder the task of fostering the development of a tuberculosis vaccine as well as new diagnostic tests and drugs. Only the developed nations have the scientific, technological, and financial resources necessary to make possible the long-term effort that vaccine development will especially require. It is that capacity that imposes the moral duty on developed nations to act to save the lives of millions who would otherwise die.
The ethical duty to rescue those who might otherwise fall victim to tuberculosis requires that the developed nations provide the resource commitment that can effectively be used. Ongoing scientific analysis of the extent to which resource constraints are hindering efforts at vaccine development are critical. Although there is no justification for expending resources that cannot be reasonably used, neither is there a justification for spending less than can productively be invested.
Addington v. Texas 441 U.S. 418 , 425–426 (1979) .
Alland D , Kalkut GE , Moss AR , et al. 1994 . Transmission of tuberculosis in New York City: An analysis by DNA fingerprinting and conventional epidemiologic methods . N Engl J Med 330(24) : 1710–1716 .
Ball CA , and Barnes M. 1994 . Public health and individual rights: Tuberculosis control and detention procedures in New York City . Yale Law Pol Rev 12 : 38–67 .
Bayer, R. , Dupuis, L. , 1995 . Tuberculosis, Public Health, and Civil Liberties . Annual Review of Public Health 16 : 307–26 .
Bloom BR , and Murray CJ. 1992 . Tuberculosis: Commentary on a reemergent killer . Science 257(5073) : 1055–1064 .
Brown RE , Miller B , Taylor WR , et al. 1995 . Health-care expenditures for tuberculosis in the United States . Arch Intern Med 155(15) : 1595–1600 .
Centers for Disease Control and Prevention . 1978 . Extrapulmonary Tuberculosis in the United States (Pub. No. USPHS/CDC 78-8360) . Atlanta : CDC .
Centers for Disease Control and Prevention . 1992 . National action plan to combat multidrug-resistant tuberculosis . MMWR 41(RR-11) : 5–48 .
Centers for Disease Control and Prevention . 1993 . Tuberculosis control laws—United States, 1993 . Recommendations of the Advisory Council for the Elimination of Tuberculosis . MMWR 42(Suppl. RR-15) : 1–28 .
Centers for Disease Control and Prevention 2000 . Reported Tuberculosis in the United States, 1998 . Atlanta : CDC .
Curtis, AB , Ridzon R , Vogel R , et al. 1999 . Extensive transmission of Mycobacterium tuberculosis from a child . N Engl J Med 341 : 1491–1495 .
Daley CL , Small PM , Schecter GF , et al. 1992 . An outbreak of tuberculosis with accelerated progression among persons infected with the human immunodeficiency virus: An analysis using restriction-fragment-length polymorphisms . N Engl J Med 326(4) : 231–235 .
Dalovisio JR , Stetter M , and Mikota-Wells S. 1992 Rhinoceros Rhinorrhea: Cause of an outbreak of infection due to airborne Mycobacterium bovis in zookeepers . Clin Infect Dis 15 : 598–600 .
Danker WM , Waecker NJ , Essey MA , Moser K , Thompson M , and Davis CE. 1993 . Mycobacterium bovis infections in San Diego: A clinicoepidemiologic study of 73 patients and a historical review of a forgotten pathogen . Medicine 77 : 11–37 .
Dowdle WR , and Hopkins DR (eds.). 1998 . The Eradication of Infectious Diseases . New York : John Wiley & Sons .
Dubos R , and Dubos J. 1952 . Tuberculosis, Man and Society: The White Plague . New Brunswick, NJ : Rutgers University Press .
Edlin BR , Tokars JI , Grieco MH , et al. 1992 . An outbreak of multidrug-resistant tuberculosis among hospitalized patients with the acquired immunodeficiency syndrome . N Engl J Med 326(23) : 1514–1521 .
Fanning A , and Edwards S. 1991 . Mycobacterium bovis infection in human beings in contact with elk (Cervus elaphus) in Alberta, Canada . The Lancet 338 : 1253–1255 .
Frieden TR , Fujiwara PI , Washko RM , and Hamburg MA. 1995 . Tuberculosis in New York City—Turning the tide . N Engl J Med 333(4) : 229–233 .
Good RC. 1980 . Isolation of nontuberculous mycobacteria in the United States, 1979 . J Infect Dis 142(5) : 779–783 .
Gostin, L. 1993 . Controlling the resurgent tuberculosis epidemic: A 50-state survey of TB statutes and proposals from reform . J Am Med Assoc 269 : 255–261 .
Gostin L. 1995 . The resurgent tuberculosis epidemic in the era of AIDS: Reflections on public health, law and society . Maryland Law Rev 54 : 1–131 .
Greene v. Edwards 263 S.E. 2d 661 (W. Va. 1980) .
Hopewell PC , and Small PM. 1996 . Applications of Molecular Epidemiology to the Prevention, Control, and Study of Tuberculosis . In: Tuberculosis (Eds. Rom WM , and Garay S ). Boston : Little, Brown, and Co. , pp. 113–127 .
In re Application of the City of New York v. Doe (No. 40770/94) (1994) .
Keers RY. 1978 . Pulmonary Tuberculosis: A Journey Down the Centuries . London : Balliere-Tindall .
Koch R. 1882 . Die Ätiologie der Tuberkulose . Berliner Klinischen Wchenschrift 15 : 221–230 .
Magnus K. 1966 . Epidemiological basis of tuberculosis eradication . 3. Risk of pulmonary tuberculosis after human and bovine infection . Bull World Health Org 35 : 483–508 .
McKenna MT , McCray E , Jones JL , Onorato IM , and Castro KG. 1998 . The fall after the rise: Tuberculosis in the United States, 1991 through 1994 . Am J Pub Health 88(7) : 1059–1063 .
Moda G , Daborn CJ , Grange JM , and Cosivi O. 1996 . The zoonotic importance of Mycobacterium bovis . Tubercl Lung Dis 77 : 103–108 .
Nation PN , Fanning EA , Hopf HB , and Church TL. 1999 . Observations on animal and human health during the outbreak of Mycobacterium bovis in game farm Wapiti in Alberta . Can Vet J 40 : 113–117 .
Rothenberg KH , and Lovoy EC. 1994 . Something old, something new: The challenge of tuberculosis control in the age of AIDS . Buffalo Law Rev 42 : 715–760 .
Small PM , Hopewell PC , Singh SP , et al. 1994 . The epidemiology of tuberculosis in San Francisco: A population-based study using conventional and molecular methods . N Engl J Med 330(24) : 1703–1709 .
Shelton v. Tucker 364 U.S. 479, 488 (1960) .
Thompson et al. 1993 . Seals, seal trainers, and mycobacterial infectionl . Am Rev Respir Dis 147 : 164–167 .
U.S. Congress, Office of Technology Assessment (OTA) . 1993 . The continuing challenge of tuberculosis . (Pub. No. OTA-H-574) . Washington, DC : OTA .
Valway SE , Sanchez MPC , Shinnick TF , et al. 1998 . An outbreak involving extensive transmission of a virulent strain of Mycobacteium tuberculosis . N Engl J Med 338(10) : 633–639 .
Walsh JA , and Warren KS. 1979 . Selective primary health care: An interim strategy for disease control in developing countries . N Engl J Med 301(18) : 967–974 .
World Health Organization/International Union Against Tuberculosis . 2000 Global Project on Anti-Tuberculosis Drug Resistance Surveillance (Rep. No. 2) . Geneva, Switzerland : Author .
Young T. 1815 . A Practical and Historical Treatise on Consumptive Disease . London .