Tuberculosis at Mid-20th Century

At mid-century, rates of death from tuberculosis had been decreasing steadily in the United States, from 194.4 per 100 000 persons in 1900 [1-3] to 40 per 100 000 persons in 1945 (Figure 1). Using mathematical modeling, Blower and colleagues [6] suggested that some of the decline in the rate of tuberculosis is due simply to the long period that is necessary for a tuberculosis epidemic to rise, fall, and reach a stable epidemic level. The epidemic of tuberculosis began in Europe in the early 1600s and peaked at the end of the 18th or the beginning of the 19th century. The major epidemic in North America began after the European epidemic, and both epidemics have been in decline since at least the beginning of the 20th century. Other factors must also have contributed to the decline, including improved socioeconomic conditions and public health interventions. The latter consisted of the widespread use of case-finding done by mass photofluorography and by tuberculin testing followed by chest radiography for persons with positive reactions. Patients with tuberculosis were isolated in sanitariums for treatment and to break the chain of person-to-person transmission.

View larger version (16K): [in this window] [in a new window]   Figure 1. Tuberculosis mortality rates per 100 000 persons in the United States. Top. 1930 to 1949. The arrow marks 1944, the year in which streptomycin was discovered. Mortality decreased steadily, from 70 per 100 000 persons in 1930 to 40 per 100 000 persons in 1945; this was a result of the natural course of the tuberculosis epidemic in the western world, improved socioeconomic conditions, the public health measures of case-finding and isolation, and sanitarium treatment. The rate of decline accelerated after 1944 [4]. Bottom. 1945 to 1992. The arrow marks 1954, the year in which isoniazid was introduced. The rate of decline in mortality that had been maintained between 1945 and 1954 slowed after 1954 but reached the low single digits by 1965 [4, 5].

In the United States in 1945, however, 63 000 persons died of tuberculosis and 115 000 new cases of tuberculosis emerged, for a rate of 87 per 100 000 persons (Figure 2). Tuberculin test surveys [1] suggested that more than 50% of the U.S. population was infected with M. tuberculosis [1]. Worldwide, it was estimated that about 5 million deaths could be attributed to tuberculosis and that 50 million persons had the disease [1].

View larger version (15K): [in this window] [in a new window]   Figure 2. Annual newly reported cases of tuberculosis per 100 000 persons in the United States, 1930 to 1992. The arrows mark 1944 and 1954, the years in which streptomycin and isoniazid, respectively, were introduced. Rates of new cases decreased steadily after 1930. The increased rates between 1941 and 1954 probably represent an increased burden of cases in armed forces personnel and immigrants as a result of World War II. Note the increase in the rate of new cases beginning in 1984 [3-5].

Tuberculosis killed many persons in the prime of life and permeated all aspects of society. In the fall of 1939, like all of the other students entering the University of Toronto, I was given a tuberculin test. My skin test result was positive, but, fortunately, my chest radiograph was negative. A less fortunate classmate with an abnormal chest film was immediately admitted to a sanitarium, where she died, that December, of tuberculous pneumonia.

In 1945, the United States had 450 tuberculosis hospitals with 79 000 beds (Figure 3). That there were only 86 000 admissions to those beds, despite the much larger number of new cases of disease, indicates the great length of sanitarium stays and the severe shortage of sanitarium beds in that era. In 1950, Chicago had about 1000 patients with tuberculosis who were awaiting admission to sanitariums; this city was typical of all large cities in the United States.

View larger version (17K): [in this window] [in a new window]   Figure 3. Tuberculosis hospitals and beds in the United States, 1946 to 1992. The increase in the number of hospitals from 1953 to 1956 probably represents building that was planned during World War II but was delayed until after the war. The rapid decline in the numbers of hospitals and beds that began in 1956 coincided with the recognition of the superiority of chemotherapy to bedrest for the treatment of tuberculosis and the rapid sputum conversion that occurred with isoniazid-containing regimens. Few tuberculosis hospitals remained after the late 1970s [4, 7].

Case fatality rates for cavitary tuberculosis were high, about 50% at 5 years [1]. Therapy depended on bedrest and collapse therapy, and its limited efficacy was legendary. Even after discharge from the sanitarium, the risk for "having a break" was ever present. The relapse could take the form of recurrence of clinical symptoms, a lung hemorrhage, a "bad x-ray," or a sputum sample positive for M. tuberculosis.

Persons who had been treated for tuberculosis were seriously stigmatized. Employers were reluctant to hire them. They were considered risky marriage partners-men, because they might not be effective breadwinners, and women, because they might be interdicted from bearing and raising children. Small wonder that patients often concealed their tuberculosis. Sheila Rothman [8] wrote of patients with tuberculosis:

"Defined as contagious by public health officials they incorporated these negative judgements into their own self-identity. They referred to themselves as "tbs" or "lungers," the diseased organ representing their personas. They were "a special species," morally as well as physically disabled."

Because of the high frequency of occupational tuberculosis, health care personnel were at least as frightened of the disease as the general public. They frequently shunned patients in general hospitals who were awaiting transfer to sanitariums. When I told fellow house officers in 1949 that, after completing my training in internal medicine, I was planning to go to Trudeau Sanitarium for training in pulmonary disease, one physician, who had himself been treated in a sanitarium for tuberculosis, tried to convince me not to go because he felt certain I would acquire the disease. However, my wife and I were tuberculin positive, a state believed to confer some immunity to reinfection. After reassurance by a staff thoracic surgeon, who had also had tuberculosis, my wife and I departed in late June 1949 for Saranac Lake.

Trudeau Sanitarium

In July 1949, the summer camp-like atmosphere of Trudeau Sanitarium, with its central infirmary building and surrounding fieldstone cottages set in the foothills of the Adirondack mountains, was in sharp contrast to Boston's inner-city hospitals, which I had just left. Recently admitted and sicker patients were housed in two hospital units, or infirmaries; recovering patients, who were receiving a program of progressive ambulation, lived in the cottages. Each cottage bedroom was furnished with a hospital bed, which could be wheeled through French doors onto a screened sleeping porch. The floors of the porches sloped downward so that if it rained, water that came through the screens would drain out through holes at the level of the floor.

During my first day at work, almost the first question asked of me was, "Where did you cure?" When I answered that I had not had tuberculosis, I was asked, "Well then, where did your wife cure?" I soon learned that because tuberculosis was a serious occupational hazard, a large proportion of the physicians, nurses, and other employees of the sanitarium had themselves had tuberculosis or had a spouse with the disease. With rare exceptions, almost all pulmonary physicians of that era had entered the field because they themselves had been treated for tuberculosis.

Sanitarium stays ranged from 1 to 2 or more years in length, and progress was measured in blocks of time 6 to 12 weeks long. Bedrest and the sanitarium regimen dominated life. The initial prescription for most patients was bedrest 24 hours a day, with graduation to shorter periods over a period of months; "rest hour" from 1 p.m. to 3 p.m. remained mandatory throughout the sanitarium stay. The official hours of work were 8 a.m. to 1 p.m. and 3 p.m. to 6 p.m. Conferences, ward rounds, and ambulatory patient clinics were never scheduled during rest hour, which was as important to staff as to patients.

As was true in many towns adjacent to sanitariums, many of the townspeople had come to cure and had stayed on to work or to start small businesses that were supported by the local health care institutions. There were many "cure-cottages" in Saranac Lake-boarding houses that were less restrictive or less expensive than the sanitarium or were places for patients to cure while waiting for a sanitarium bed to open up. The physicians in the town cared for these patients. Although patients were taught that they had a responsibility to protect others from the disease and were generally meticulous in covering their mouths with a tissue when they coughed, my wife and I cringed when we heard someone cough in the darkness of the only movie theater in Saranac Lake.

In the pre-television age, we made heavy use of the Trudeau Library, which was superb. Perhaps its extensive, up-to-date collection of fiction and nonfiction work was a reflection of the many intellectuals who had been patients in Saranac Lake-from Robert Louis Stevenson in 1887 to Walker Percy in 1942 [9]. Although the droplet theory of airborne transmission from person to person had been developed in the 1930s [10], the old thesis that fomites transmitted infectious organisms was still in the process of being discarded. Although the books in the library were also read by patients, the librarians assured us that they had been proven to be culture negative.

The librarians were two sisters whom I estimated to be in their mid-thirties. Imagine my surprise a couple of months after my arrival when, during one of their regular checkups, I learned that both women were in their early fifties! Men and women who looked younger than their ages were commonplace at Trudeau, perhaps because exposure to full sun, which we now know enhances photoaging of the skin, was discouraged by the medical staff because it was thought to be a potential cause of hemoptysis. The text on tuberculosis that I used [11] said, in a chapter titled "Sunshine and Artificial Radiation," that

"[t]he beneficial effects of these in extra-pulmonary tuberculosis is universally admitted and need not be discussed here. The position is different, however, in regard to pulmonary tuberculosis. They are said to act as stimulants. A large dose of a stimulant may be very harmful, and even in small doses its administration needs careful regulation. It follows that direct sunshine, the administration of which cannot be finely regulated, is generally contraindicated in the treatment of phthisis. It has been said to favor the occurrence of haemoptysis, but this is difficult to prove."

Direct Ziehl-Neelsen-stained smears were used to identify acid-fast bacilli in sputum; a smear after concentration of the sputum had to be specially ordered. Gastric aspirates were used in the absence of sputum production. Guinea pig inoculation was the standard for diagnosis or for determining sputum negativity [11]; cultures were just coming into widespread use. The commonly used medium was Lowenstein-Jensen medium, which consists primarily of an inspissated slant in a test tube of egg yolk mixed with malachite green to suppress the growth of nontuberculous bacteria. It took us a while to figure out that angel food cake was such a popular dessert at dinners given by Trudeau staff families because egg whites, the main ingredient in the cake, were a byproduct of Lowenstein-Jensen medium.

That radiography is a source of exposure to ionizing irradiation was not yet widely known. Beginning at Trudeau and continuing until the late 1950s, I fluoroscoped 25 or 30 patients twice each week to evaluate the state of their artificial pneumothoraces, to estimate the size of air refills, or just to supplement the more expensive chest radiography.

Streptomycin, which was discovered in 1944 [12], was soon in use at Trudeau, and the frequent and rapid emergence of resistant M. tuberculosis when the drug was used alone was well known. Para-aminosalicylic acid, discovered in 1946 [13], was our second antituberculous drug. The acid came as a powder and was made into the sodium salt through the addition of sodium bicarbonate. A light amber solution resulted and became progressively darker over the following week, its maximum keeping-time. The drug was given in a dose of 12 to 14 g per day; tasted vile; and produced nausea, diarrhea, and bloating. Patients hated it. The window screens of patients' rooms and the walls below them became white with discarded para-aminosalicylic acid. However, this agent represented a major advance in drug therapy. It prevented the emergence of organisms resistant to streptomycin and permitted courses of streptomycin to be given two or three times weekly for several months, which was long enough for many cavities to heal.

Artificial pneumothorax, diaphragmatic paralysis produced by crushing the phrenic nerve, the introduction of pneumoperitoneum, and thoracoplasty and its offshoots were variants of collapse therapy used to close cavities. Lung resection was just coming into use.

Antituberculous Drugs and Randomized Clinical Trials

The success of large trials of tuberculosis therapy at mid-century led to the establishment of the randomized, cooperative clinical trial as the norm for all new drug development, and for therapeutic interventions generally, in the latter half of the 20th century. In September 1945, the Veterans Administration was operating slightly more than 9000 tuberculosis beds, of which slightly fewer than half were occupied by veterans of World War I. As the 12 000 000 persons who had fought in World War II were demobilized, the total number of patients with tuberculosis under the care of the Veterans Administration increased to 15 000. The Veterans Administration successfully obtained funds from the U.S. Congress for research and, in company with the U.S. Army and the U.S. Navy, began cooperative studies of streptomycin [14]. Untreated controls were not used in the initial studies of streptomycin in the United States because supplies of streptomycin were ample and the investigating physicians thought it would be unethical to withhold chemotherapy; only historical controls were used. In 1946, however, streptomycin was in short supply in Great Britain and investigators saw no ethical problems with a placebo arm in a randomized trial. The 1948 report from the Medical Research Council of Great Britain [15] on streptomycin for treatment of tuberculosis was the first to incorporate all of the elements of the modern, randomized clinical trial [16].

Randomized studies of new chemotherapeutic agents, bedrest, and surgical and collapse therapy soon followed the studies of streptomycin done by the Veterans Administration and Armed Forces investigators. The U.S. Public Health Service and the British Medical Research Council joined in. New, high-quality, trustworthy information accumulated and changed the management of tuberculosis at a dizzying pace.

Early in 1952, reports began appearing in the New York and Washington, D.C., papers of a wonder drug for the treatment of tuberculosis that was being tested at the Seaview Hospital on Staten Island. The results of that clinical study were published in April 1952 [17]. Reports of additional clinical studies and of the pharmacology, toxicology, and in vitro and experimental effects of isoniazid dominated the Annual Meeting of the American Trudeau Society held in May 1952. You can imagine the excitement of physicians at the prospect of having a relatively nontoxic drug to replace either para-aminosalicylic acid and its unpleasant side effects or intramuscular injections of streptomycin. By the early 1970s, ethambutol, rifampin, and pyrazinamide rounded out the list of first-line drugs for tuberculosis [18].

Effects of Chemotherapy on Sanitarium Care

The length of sanitarium stays began to shorten as sputum conversion began to occur within a few months in response to chemotherapy. Artificial pneumothorax had been abandoned as a form of collapse therapy by the mid-1950s. Controlled studies showed that bedrest and sanitarium care did not add to the benefits produced by chemotherapy [19-21], and ambulatory care became the standard of care. By the late 1970s, most tuberculosis sanitariums in the United States were closed or converted to other purposes (Figure 3). The political battles to close sanitariums were sometimes fierce. In most places, unfortunately, an effective network of outpatient clinics to manage tuberculosis was not substituted for the sanitarium.

The effect of the closure of sanitarium beds on the employment of physicians specializing in tuberculosis was similar to the threat posed to specialists in the current era of hospital restructuring. Older physicians retired, took jobs in tuberculosis clinics, or shifted to the practice of general medicine. Younger physicians learned pulmonary function testing, intensive care, and fiberoptic bronchoscopy and became skilled in the diagnosis and management of nontuberculous lung disease.

Implications of Chemotherapy for the Control of Tuberculosis

Bacille Calmette–Guérin (BCG) vaccine was never widely used to control tuberculosis in the United States, but in many countries, 80% or more of the population was vaccinated [22]. Controversies flourished about the effectiveness of BCG vaccine. In a country with a relatively low risk for tuberculous infection, such as the United States, the vaccine was seen as having little effect on the incidence of new cases; we put our faith in chemoprevention. In 1959, confident that tuberculosis could be controlled with preventive chemotherapy, the Arden House Conference made recommendations for the eradication of tuberculosis in the United States [23]. Drug treatment, the cornerstone of the plan, was to be applied vigorously to persons with disease and as preventive therapy to those who were only infected. The report of this conference was the basis of the tuberculosis control program in the United States for the next two decades. Those decades saw the widespread use of preventive chemotherapy for tuberculous infection and the development of various 6- to 9-month drug regimens. Six months was still too long for good compliance, and two- to three-times-weekly regimens and fixed-dose combinations of antituberculous drugs were developed [24]. Even more important in assuring compliance and preventing the emergence of resistant organisms was directly observed therapy-having patients come to the clinic two or more times weekly and take their antituberculous medication under direct observation [25-28].

Tuberculosis at the End of the 20th Century

By 1984, the mortality rate in the United States had decreased to 0.7 per 100 000 persons (Figure 1) and the rate of new cases had decreased to 9.4 per 100 000 persons (Figure 2 and Figure 4). As the statistics improved, federal funding for tuberculosis control decreased. New public health needs resulted in increased competition by health departments for program dollars, and many state and city governments downgraded their tuberculosis control programs and supervision of chemotherapy. After 1984, however, the previously downward-trending curve of new case rates against time leveled off and then began to increase. Between 1985 and 1992, 51 700 more new cases than expected were seen (Figure 4) [30]. A tide of immigrants from countries with high prevalences of tuberculosis, the increase in homelessness and drug abuse, and the epidemic of HIV infection and AIDS in the United States were all believed to be playing a part in the resurgence of tuberculosis [30]. Poor compliance with therapy by persons with social problems and the importation of drug-resistant organisms from countries in which isoniazid is incorporated into cough medicines resulted in expanding pockets of multidrug-resistant tuberculosis in many U.S. cities. The emergence of multidrug resistance and the number of excess cases of tuberculosis seen was alarming and shook the complacency of physicians, the authorities, and the public.

View larger version (22K): [in this window] [in a new window]   Figure 4. Newly reported cases of tuberculosis per 100 000 persons by year in the United States, 1980 to 1995 [2, 3, 5, 29]. The rate of decline of about 0.85 new cases per 100 000 persons per year seen between 1975 and 1984 is continued to 1995. After 1984, however, the previously downward-trending curve of new cases against time leveled off and then began to increase from 9.3 per 100 000 persons in 1985 to 10.3 per 100 000 persons in 1990, when it again leveled off. The downward trend resumed in 1993 and decreased to 8.7 per 100 000 persons in 1995; this was the lowest rate of new cases of tuberculosis to be seen since surveillance began in 1953. The shaded area represents the excess of new cases above the number expected between 1985 and 1995. (Figure adapted from [30].).

The case rate of tuberculosis in the United States resumed its decline in 1993 and was 8.7 per 100 000 persons in 1995 (Figure 4); this was the lowest rate for tuberculosis cases reported since 1953, when national surveillance began. It is likely that this resumption of the decrease in the rate of new cases was largely due to a reinfusion of federal funds, the expansion of inner-city clinics, and the institution of directly observed therapy. Although the number of cases reported in the United States from 1994 to 1995 decreased by 6.4%, the percentage of reported cases occurring foreign-born persons increased from 32% in 1994 to 36% in 1995 [29].

The Global Perspective

From a global perspective, the magnitude of the tuberculosis problem has not diminished over the last 50 years. It was estimated that in 1990, there were 7.6 million new cases in developing countries and 400 000 new cases in industrialized countries, for a worldwide total of 8 million new cases [31]. About 5 million new cases are estimated to occur annually in Asia, but the highest estimated case rates are in sub-Saharan Africa: The current rate there is 229 per 100 000 persons [32]. It is estimated that one third of the world's population is infected with M. tuberculosis; 75% to 80% of adults in developing countries have M. tuberculosis infection. It has also been estimated that tuberculosis caused 2.9 million deaths worldwide in 1990, and all but 40 000 of these occurred in developing countries [22, 31].

The pandemic of HIV infection has had a profound effect on the global problem of tuberculosis. It is estimated that 9 to 11 million adults and 1 million children worldwide have HIV infection, and about 85% of this burden is in developing countries. In these countries, the combination of a high frequency of HIV infection with a high frequency of M. tuberculosis infection results in a high rate of tuberculous disease as a complication of AIDS: 20% to 40% in Africa; 18% in Haiti; and as high as 25% in Brazil, Mexico, and Argentina [29]. It is evident that efforts to control tuberculosis will depend on the success of efforts to control the spread of HIV. The picture is grim.

Epilogue

As Susan Sontag has made clear [33], diseases that attain epidemic proportions before their causes are fully understood and their treatments become effective develop complex metaphoric meanings. Syphilis, in the 19th century, implied a moral judgment about "off-limits" sex. Tuberculosis was romanticized in the literature and opera of the 19th century; it was thought to be apt to strike a person who was hypersensitive or talented or had passionate feelings for a loved one, feelings that were usually thwarted. In our own time, cancer and AIDS are often treated as evil, invincible predators rather than as diseases susceptible to treatment. The association of HIV infection with male homosexuality has invoked the judgment by many that AIDS is a just punishment for "unnatural" acts. The stigma is extended to persons who have acquired the disease through transfusions of blood products or heterosexual transmission. The denial of death, so rampant in our culture, contributes to the shunning of such patients by their friends and acquaintances, who may have irrational fears of acquiring HIV through ordinary social contact.

The last 50 years have seen dramatic advances in our power to control and cure tuberculosis. But no easy solution has surfaced-no highly efficacious vaccine, no antibiotic combination that cures the disease in a week or two. Although tuberculosis has come under control in industrialized nations, it is still, as Dubos described it in 1952 [34], "The Captain of All the Men of Death" in developing countries, which have neither the resources nor the social organization to duplicate the feats achieved in the developed nations.

Even in industrialized countries such as the United States, we must remain vigilant. Clinicians must be mindful of the disease if they are to diagnose it. Social conditions such as housing and nutrition continue to be important in the development and spread of tuberculosis. We must make certain that our health care agencies are funded so they can play their part in the old-fashioned, painstaking methods of case-finding, isolation, case-holding, and successful drug treatment. Physicians should explore, together with the pharmaceutical industry, financially viable ways to develop new antituberculous agents [35]. We have accomplished much in the last 50 years, but tuberculosis is still very much with us. Its control requires a major international collaborative effort by physicians, public health workers, researchers, industry, and governments.