Throughout history, tuberculosis has been a plague on mankind. Tuberculosis remains the most common cause of death from infectious agents in childbearing-age women (14 to 49 years) worldwide.¹ Approximately one third of the world's population manifests evidence of current or past infection with Mycobacterium tuberculosis as indicated by tuberculin skin testing.^2,3 Between 10% and 15% of these individuals have reactivation of their latent disease, and many die as a result of the disease. Overpopulation, human immunodeficiency virus (HIV) infection, increasing poverty, and the increasing incidence of antibiotic-resistant isolates are increasing the burden of tuberculosis on women and society.

Cases of tuberculosis in childbearing women and their infants, when compared across gender and age, suggest a disproportionate number of infected women. When the reported cases of tuberculosis (1985 through 1990) in the United States is plotted by age, there are peaks in the first year of life (about 900), age 35(2700), and a broader increase between the ages of 50 and 80 years (1500 to 1700 cases per year).⁴ This graph is remarkable for its peaks in infancy and childbearing age women. These peaks have been increasing over the past 10 years, in part because of the increasing incidence of HIV positivity and the increasing numbers of women at high risk for tuberculosis. Although the male to female ratio of new tuberculosis notifications is about 2:1 in developed countries and 1.2:1.7 in developing countries, there is some evidence that tuberculosis in women is underreported compared with the statistics for men.¹ This is especially true for developing countries.

Historically, there are conflicting opinions concerning the effect on the progression of tuberculosis during pregnancy. In the era before chemotherapy, pregnancy and the immediate postpartum period were associated with higher rates of active tuberculosis and faster progression of the disease. Most studies in the era before chemotherapy fail to provide adequate comparison populations. Selection bias may have resulted in the inclusion of higher-risk women in the study groups. In the era of effective chemotherapy, there appears to be no difference in the progression of the disease or the cure rates when the results are controlled for risk factors and compliance. However, tuberculosis remains a significant worldwide concern, obstetrically and economically. Among pregnant or postpartum women, tuberculosis annually kills more women worldwide than all other causes of maternal mortality combined.⁵ In developing countries, women contribute 40% to 60% of household income if home production of household products or export crops are counted. Children whose mothers die are 3 to 10 times more likely to die within 2 years than those with both parents alive.⁵ This impact is twice that if the father were to die.⁶ Tuberculosis kills or maims reproductive-age women, leaving orphans, impoverished families, and limited economic development.

Tuberculosis often seems far removed from the obstetric care provider in the United States or Western Europe. The high levels of socioeconomic development and public health commitment have reduced the incidence of symptomatic tuberculosis dramatically. In the mid-1990s, the rate of active tuberculosis in the United States was relatively constant at 10 cases per 100,000 persons.

Between 5% and 10% of reproductive-age women in the United States have a reactive tuberculin skin test.⁴ Most of the active tuberculosis affects specific high-risk populations: recent immigrants, institutionalized individuals, individuals whose lives are complicated by substance abuse, and HIV-infected individuals. One third of all persons with tuberculosis in the United States are middle- or upper-income individuals.

The outbreaks of multidrug-resistant tuberculosis (MDR-TB) among health care workers^7–11 should raise significantly the obstetrician's caution and index of suspicion. A review of case reports suggests that approximately one of three exposed health care workers has a tuberculin skin test that converts. Active multidrug-resistant tuberculosis (MDR-TB) has been reported in more than 25 health care workers. The therapy of MDR-TB is highly complex and is associated with major morbidity and mortality in more than 50% of immunocompetent health care workers. Among HIV-infected patients with active MDR-TB, the mortality is 80% to 90%, with a remarkably shortened life expectancy. As a primary care provider for women of childbearing age, a commonly affected group, the obstetrician and gynecologist have a mandate to identify women at risk for tuberculosis or those with previous exposure.¹ During pregnancy, the obstetrician has a major responsibility to predict and assess the effects of chemotherapy on the fetus. Throughout the course of her therapy, a team of an internist (infectious disease specialist), obstetrician, and public health nurse provide the support and education needed for compliance with her chemotherapy.

M. tuberculosis is the major human pathogen of the family Mycobacteriaceae, order Actinomycetales. Humans are the major reservoir for this species. Mycobacterium bovis (cattle) occasionally infects select populations (3% of tuberculosis in San Diego) and demonstrates uniform resistance to pyrazinamide.¹² Mycobacterium ulcerans, Mycobacterium microti, and Mycobacterium africanium are biologically similar to M. tuberculosis and M. bovis but are associated with disease in rodents or, rarely, humans. These five organisms are the members of the M. tuberculosis complex.

Other similar but distinct acid-fast Mycobacterium species can cause significant disease: Mycobacterium leprae (leprosy), Mycobacterium avium (M. avium complex), Mycobacterium intracellulare, and Mycobacterium scofuluceum. Many other Mycobacterium species exist in soil, water, and animal reservoirs and only rarely are associated with human disease. Mycobacterium species other than the M. tuberculosis complex and M. leprae have been classified into major groups: photochromogenic, scotochromogenic, nonchromogenic, and rapid-growing species.

M. tuberculosis is an aerobic, non-spore-forming, nonmotile bacillus with a high cell wall content of high-molecular-weight lipids. Its generation time is 15 to 20 hours, compared with less than 1 hour for most common bacterial pathogens. Visible colonies require 3 to 4 weeks and appear as serpentine cording.

The clinical impact of delay in identification is the decision whether to initiate chemotherapy and at what intensity. Unfortunately, the decision to treat with three to four powerful antibiotics for 6 to 9 months may be made without the benefit of positive cultures and rest on the epidemiology and clinical presentation.

In an effort to reduce the incidence of unnecessary treatment, an acid-fast stain is used to screen specimens. In the classic Ziehl-Neelsen stain, a fixed smear covered with carbol fuchsin is heated, rinsed, decolorized with acid-alcohol, and counterstained with methylene blue. The Kinyoun stain is similar but modified to make the heating unnecessary. Many laboratories use a fluorochrome stain with phenolic auramine or auramine-rhodamine in the initial staining, a slightly modified acid-alcohol decolonization step,¹³ and potassium permanganate counterstaining. The mycobacteria fluoresce bright orange-yellow against a dark background with a strong blue light source. Under the 100× oil-immersion objective, the mycobacteria are beaded, slightly bent rods 2 to 4 μm long and 0.2 to 0.5 μm wide. In sputum, they lie in parallel or adhere end to end to form a V shape. This smear has a sensitivity of 78%, with 11% false-positive results.

An estimated 10,000 organisms per milliliter of sputum are required for acid-fast stain positivity, and the identification of a single organism on the entire slide is highly suspicious. Paucity of organisms dictates the acquisition of specimens with a high concentration of organisms and thorough examination of the prepared slides. At least three early morning (before rising) specimens are recommended. Early morning gastric aspirates are effective if obtained before ambulation. Thin fluid (i.e. urine) should be examined after sedimentation with centrifugation.

Before media inoculation, sputum and other contaminated specimens require liquidization and decontamination with N-acetyl-L-cysteine in 1% sodium hydroxide solution. The sample is then neutralized and centrifuged, and the sediment is inoculated into the media. Uncontaminated specimens (i.e. surgical tissue or cerebrospinal fluid) should not be decontaminated, because doing so reduces the growth of Mycobacterium organisms. Solid media is of two types, agar-based or egg-based, with suppressive additives for bacteria other than mycobacteria. Many clinical laboratories use the BACTEC radiometric system (Johnston Laboratories, Towson, MD, USA). Radioactive palmitate is used as the sole carbon source in this liquid culture system. Within 9 to 16 days, metabolism is detected if M. tuberculosis is present.

M. tuberculosis can be differentiated from other mycobacteria by its slow growth, lack of pigment, production of niacin and small quantities of heat-sensitive catalase, reduction of nitrates, and isoniazid sensitivity. Antibiotic sensitivity is determined by comparison of growth from appropriately diluted inocula on antibiotic-containing media (solid or liquid) to growth from on a antibiotic-free media. Antibiotic resistant is determined when growth on the antibiotic-containing media exceeds 1% of the growth on antibiotic-free media.

The use of polymerase chain reaction (PCR) in the diagnosis is rapidly changing the specific identification of M. tuberculosis. This technique can identify as few as 10 organisms in clinical specimens, compared with the 10,000 organisms necessary for AFB smear positivity, possibly within a day. Although the PCR-based diagnosis is not widely available, in the near future, treatment without diagnosis will be an issue of the past. Traditional culture and sensitivity is still necessary, because PCR cannot detect antibiotic sensitivity.¹⁴

Worldwide, almost 1.75 billion people have been infected with tuberculosis, and 3 million die each year as a consequence. Increasingly crowded living conditions and exposure to naive populations help explain these statistics.¹⁵ In the past 100 years, the attack rate and death rates in the United States and Western Europe have dropped 8% to 10% per year until 1985. In developing countries, the flight from rural areas to urban centers where inadequate infrastructure, including housing, jobs, and medical facilities, create urban ghettos. As tuberculosis is spread by airborne particles, these areas are a driving force in the epidemic.

There is an element of increased susceptibility in the populations of the developing world. When the industrial revolution in Europe and the United States brought farmers into the overcrowded cities, one third of all adult deaths were associated with tuberculosis. This selective pressure eliminated those who had the least resistance. During the 19th century, each new infectious case, those with primary, acute tuberculosis or cavitary lesions, infected 20 or more new individuals. In developed countries today, each case infects 13 or less new individuals. These observations can be explained by changes in host resistance, better public health services, or decreased virulence of the organisms. The latter is less likely, given the epidemic seen in developing countries.

In the past 150 years, the epidemic has spread to those countries whose industrialization and urbanization is in process. The epidemic is fueled by crowded living and a large pool of naive humans. In 1985, the rate of tuberculosis cases in the United States leveled out with an excess of 39,000 cases occurring between 1985 and 1991. This change reflects the reduction in public support of tuberculosis control programs, the increases in at-risk populations (i.e. intravenous drug users, HIV-infected individuals, and homeless families). Changing this alarming trend remains a critical challenge for the U.S. health care system.

Most tuberculosis in the United States is found in high-risk populations. These high-risk populations include health care workers, the urban poor, alcoholics, the homeless, intravenous drug abusers, migrant farm workers, and individuals who are chronically institutionalized (e.g. prisoners, state mental institution, nursing homes). The worldwide, new case rates per 100,000 persons are depicted in Table 1. In the United States, two thirds of all new cases are minority individuals.

TABLE 1. Estimated Worldwide Active Tuberculosis Rates Per 100,000 Population

The rates of active tuberculosis among U.S. populations reflects three factors: the prevalence of exposure, the prevalence of remote infection (positive tuberculin test), and the ability of the host to resist systemic infection. In the general population of developed countries, the most likely source of active tuberculosis is reactivation of remote disease in the elderly population. However, within high-risk U.S. populations, active tuberculosis is found among children and young adults where exposure is high in crowded conditions and stress, drugs, or HIV weaken their immune systems. The dramatic effect of HIV positivity is illustrated by the fact that 46% of tuberculosis patients in New York City¹⁷ and 0.3% in Honolulu were HIV positive. Among HIV-positive patients in this country, 10% harbor the tubercle bacilli, and of these, 60% have active disease. Relatively asymptomatic HIV-infected patients can be infectious for tuberculosis without cavitary lesions.

Although tuberculosis can be transmitted by infected raw cow's milk (M. bovis) or skin inoculation from contamination of an abrasion or laceration, almost all tuberculosis infections are caused by inhalation of infectious particles aerosolized by coughing, sneezing, or talking. The droplets are small enough to allow them to dry while airborne and become particles that remain suspended for long duration. A cough, sneezing, or talking for 5 minutes can produce 3000 infectious particles. The number of infectious particles required to produce infection is unknown, but it is likely that the risk of infection is inversely correlated with host defenses. For immunocompetent individuals, prolonged exposure to multiple aerosol inocula are needed for infection. In the United States, 27% of household contacts of AFB smear-positive index cases become infected.¹⁸ Pregnant women and children bear the brunt of new infections. When the index case is AFB smear negative but culture positive, the risk of infection in household contacts is 5% to 10%. When the air environment is closed (e.g. a submarine), the attack rate may be as high as 50% to 80%.¹¹ In an nursing home epidemic, one case infected 31% of the tuberculin-negative patients but 79% of patients residing on the same wing.¹⁶ When a HIV-sensitized young patient was admitted three times for persistent pneumonia and was found to have active tuberculosis, 10 of 34 tuberculin-negative health care workers had converted tuberculin test results. When the exposure occurs on a hospital acquired immunodeficiency syndrome (AIDS) ward, the risk of infection is 40% to 80%, depending on the degree of immunocompromise (i.e. CD4 count) of the patients.¹⁶

Three to four percent of infected persons develop active tuberculosis during the first year after tuberculin conversion, and 5% to 15% will do so during the remainder of their life.¹⁶ They are at greatest risk when their host immunity is decreased: aging, concurrent disease, pregnancy, and primary compromise of their immune system (i.e. HIV infection, immunosuppressive drug therapy, steroids, and cancer chemotherapy). The Centers for Disease Control and Prevention (CDC) reported that tuberculin-positive, HIV-positive intravenous drug abusers develop active tuberculosis at a rate of 8% per year. Isoniazid prophylaxis seems to prevent the progression.¹⁹

Aerosolized tuberculosis particles need to be small enough to be carried to the terminal air spaces, heavier droplets come into earlier contact with the bronchial mucociliary surface and are expectorated before infection occurs. High-airflow areas, mid-lung regions (i.e. lower upper lobes or upper lower lobes), favor the initiation of infection in the subpleural spaces. In 75% of cases, there is just one area of infection, and in 25%, there are multiple areas. Alveolar macrophages ingest the bacilli, and multiplication of the bacteria continues unimpeded. Blood-borne lymphocytes and macrophages begin to ingest the bacilli. The infected macrophages migrate to the regional lymph nodes (usually hilar or mediastinal) and form granulomatous lesions. During this phase, the bacilli disseminate to the corpus at an amount proportional to regional blood flow. Certain areas favor the continued multiplication of the bacilli: lymph nodes, posterior-apical lungs, meningeal areas, vertebral bodies, kidneys, and epiphysis of the long bones.

Until the development of tuberculin hypersensitivity, bacterial growth and tissue destruction continue. The speed at which the host develops the cellular hypersensitivity determines the survival of the bacilli (and patient). In the immunocompromised patient, the widely disseminated bacilli begin to multiply and destroy the local tissue. The incubation period for active tuberculosis has been reported to be as short as 20 days, and in most cases, it is 1 to 3 months. The most common manifestation is cavitary destruction of lung tissue in the apices. The most extreme form is described as miliary and is associated with difficult and complex chemotherapy and high death rates.

In most cases, the hypersensitivity reaction (i.e. tuberculin test positivity) occurs in 3 to 8 weeks after infection. Rapid destruction of the bacteria leaves only the positive tuberculin test as evidence of infection. In a minority of cases, in which the primary infection created more destruction, calcification of the primary focus (Ghon focus) and regional lymph nodes produces what is called the Ranke complex. Although most of these calcifications are sterile, some retain viable bacilli that provide a nidus for active tuberculosis many years later when the patient's immune system becomes compromised by age, illness (e.g. HIV), or medication (e.g. chemotherapy, steroids).

Early tuberculosis is usually asymptomatic, only becoming symptomatic with advanced bacillary load. Table 2 describes the presenting signs in nonpregnant women versus pregnant individuals versus HIV-seropositive individuals.

TABLE 2. Symptoms of Active Tuberculosis in Women

The early symptoms are notoriously nonspecific—anorexia, fatigue, weight loss, chilly sensations, and night sweats—and are often well tolerated by the patient. Local symptoms indicate advancing disease. A productive cough is usually present as bronchial involvement advances. The mucopurulent sputum is nonspecific in character, and the mild coughing and sputum are easily ignored by patients with a history of bronchitis (i.e. smokers). As the areas of infection and necrosis grow, surrounding structures become involved. Bronchial artery or vein involvement results in hemoptysis. Meningeal involvement results in neurologic symptoms and signs. Invasion of the vertebrae causes back pain and vertebral collapse (Pott's disease). Consolidation, pleural involvement, and effusion mimics pneumonia. Genitourinary disease produces symptoms suggesting chronic urinary infections, chronic pelvic inflammatory disease, cervical erosions, or infertility.

Because most cases of symptomatic disease present with pulmonary symptoms, the chest radiograph is the centerpiece of the clinical diagnosis.³⁰ During the primary phase of the disease, radiographic abnormalities include pulmonary consolidation (50%), cavitation (29%), hilar or mediastinal lymphadenopathy (35%), disseminated miliary disease (6%), and a normal chest radiograph (15%). During the postprimary phase of the disease, common radiographic abnormalities include exudative or fibroproductive parenchymal densities (100%), predominantly in the apical and posterior segments of the upper lobe (91%), cavitation (45%) with bronchogenic spread of the disease(21%), marked fibrotic response in the lungs (29%), pleural effusion(18%), and and emphysema (4%).

Fiberoptic bronchoscopy can play a critical role in the diagnosis of active tuberculosis. Patients who have signs and symptom or radiographic evidence of active tuberculosis without a positive AFB smear should have a fiberoptic bronchoscopy. At Walter Reed Hospital, 22 of 25 patients with confirmed M. tuberculosis infection had at least three prebronchoscopy AFB smears that were negative for tuberculosis. Many of these false-negative AFB smears occur in patients with anergy. Immunocompromised patients with pulmonary symptoms should be considered for bronchoscopic diagnosis and culture.¹⁶

The mortality rate for untreated active tuberculosis is 50% to 60% within 5 years. The interim period is marked by increasing disability and symptoms related to destructive, caseating granulomas in various organs, severe weight loss and fatigue, increasing respiratory symptoms, and superimposed infectious disease (e.g. pneumonia). About 15% develop miliary tuberculosis. This complication, if untreated, is rapidly fatal. With modern chemotherapy and adequate patient compliance, remission or cure of active tuberculosis occurs in more than 95% of patients (80% of patients with HIV coinfection or miliary dissemination).

Active or previous infection with tuberculosis can be detected by a skin challenge, the tuberculin skin test. These tests monitor the reaction to an intracutaneous injection of sterile antigenic components of M. tuberculosis called tuberculins. Two preparations of tuberculins are licensed in the United States: old tuberculin (OT), which is used in the multipuncture technique (Tine Test), and purified protein derivative (PPD), which is applied by intracutaneous injection (Mantoux test). All tuberculins have been standardized to the activity of a large batch of PPD produced by Seibert in 1939, lot number 49608 (PPD-S). The multipuncture techniques deliver the tuberculin by solid needles coated with the tuberculin. Although the actual dose of tuberculin is variable, the multipuncture technique is easier to store and administer to large populations where medical services are absent or poor.

The Mantoux test is an intradermal injection test and is the standard of diagnosis in the United States.^31–33 One tenth of a milliliter of PPD (5 tuberculin units [TU]) is injected intradermally (27-gauge needle with the bevel up) on the volar or dorsal forearm. In selected cases, a smaller dose (1 TU) or larger dose (250 TU) is applied. A discrete, pale elevation of the skin (wheal) 6 to 10 mm in diameter should be produced. If the first injection does not appear as described, a second site is selected a few centimeters from the first.

Tests should be read 48 to 72 hours after the inoculation. The cellular hypersensitivity reaction persists 4 to 5 days in most cases. The interpretation is based on the presence and size of the induration (greatest diameter) at the injection site. Multiple-puncture techniques may have one or more areas of induration; the largest is measured. If they coalesce, the longest diameter is measured. A vesicular reaction on a multipuncture technique is a positive test. Otherwise, induration (<15 mm) on multiple-puncture test should be followed by a Mantoux test within a week. The median size of Mantoux skin tests in immunocompetent women is 15 to 18 mm. A 2- to 3-week delay in repeat testing may result in a false-positive result because of the mild antigenic response of the first test (<5 mm); this is called a booster effect. Serial testing (yearly) with PPD does not sensitize patient to the tuberculin protein. The booster effect is seen primarily in women who have been vaccinated with bacille Calmette-Guérin (BCG) or have been infected with other Mycobacterium species. In patients, who are younger than 35 years old, with a previously positive Mantoux test within the prior 2 years, an increase in reaction size by 10 mm or more should be considered to have a new tuberculosis infection.³¹ Pregnancy does not seem to affect the reactivity in incidence or size of the reaction to tuberculin skin testing.³⁴ The interpretation of a positive Mantoux test (5 TU) is shown in Table 3.

TABLE 3. Interpretation of a Reactive Tuberculin Test

The clinical efficacy of the Mantoux skin test depends on the prevalence of active or remote tuberculosis in the population most similar to the patient. Among extreme-risk patients, the prevalence may be as high as 50%; the positive predictive value for a tuberculin test (assuming a greater than 95% specificity) is 95%. Among high-risk patients with a prevalence of 25%, the positive predictive value is 86%. Among low-risk populations with a prevalence of 5%, the predictive value of a positive tuberculin test is 50%; one of two positive tuberculin skin tests is a false-positive result. These false-positive results are the result of exposure to nonpathogenic Mycobacterium sp. (in the southern United States³²), infection with M. avium, or vaccination with BCG.

The probability that the induration from a positive tuberculin skin test in a patient that has been previously vaccinated with BCG is a true M. tuberculosis infection is related to the size of the induration, risk strata of the patient, and the interval between vaccination and skin testing.^35–38 Many (30% to 50%) infants from countries where the incidence of active tuberculosis is high have received BCG vaccination. Because vaccination-induced reactivity wanes with time, the reactive effect of BCG is minimal after 10 years. The current recommendation is to interpret the tuberculin reaction in women who were vaccinated during infancy no differently than other patients (Table 3).

False-negative results are of concern, especially in high-risk populations. The incidence of false-negative tests is related to the immunocompetence of the tested individual and testing or interpretation errors.³⁹ Revision of a positive PPD result occurs for 8.1% of responders, and an age less than 19 at first positivity is most common (22%).⁴⁰ Only 50% to 60% patients with symptomatic tuberculosis, regardless of HIV status, have a reactive tuberculin skin test result. Patients with chronic pulmonary or systemic symptoms need bacteriologic workup, regardless of the tuberculin reactivity status. These observations probably result from decreased immunocompetence of the patient. Similar observations are true with other immunocompromised states. In a study of HIV-seronegative postpartum women in Uganda,⁴¹ 82% had reactions to tuberculin skin testing of more than 2 mm. Similar HIV-seropositive postpartum women had only a 48% incidence of positive tuberculin skin testing. The medium induration sizes were 10.6 mm (HIV seronegative) and 7.5 mm (HIV seropositive, statistically significant). Among HIV-positive patients with active tuberculosis and whose CD4 count is more than 200 cells/mm3, only 60% have a reaction to tuberculin skin testing that is more than 9 mm. For a CD4 count of less than 200 cells/mm3, the incidence of a positive skin test is only 35%. Similar increases in false-negative rates would be expected among populations with other immunosuppressive disorders or medications.

For immunocompromised patients (e.g. HIV infections, opportunistic infections, suspected symptomatic tuberculosis, transplant patients on immunosuppressive chemotherapy, chronic steroid therapy, cancer chemotherapy), anergy testing should be performed.³⁹ Intracutaneous injection of 0.1 ml of antigen is followed in a similar fashion to tuberculin skin testing. Commonly used antigens include coccidioidin, histoplasmin, trichophytin, tetanus toxoid (1:10 dilution), Candida albicans, and mumps. At least two antigens in addition to tuberculin should be placed simultaneously. Induration of more than 2 mm in 48 to 72 hours is considered positive for an adequate cellular hypersensitivity response; the tuberculin test result may be considered valid. Among patients with intravenous drug abuse, HIV-seronegative patients had an 11% incidence of anergy; 44% of HIV-seropositive patients were anergic. For HIV-infected patients, the CD4 count predicts anergy.⁴² In seropositive patients with CD4 counts higher than 500 cells/mm³, the incidence of anergy is similar to seronegative patients (about 10%). When the CD4 count falls below 200 cells/mm³, anergy occurs in two thirds of seropositive patients. When the CD4 count is below 50 cells/mm³, the incidence of anergy is 80%.

Cellular hypersensitivity to tuberculins does not imply complete immunity to new infection. DNA studies in a epidemic associated with a population of homeless individuals documented the same isolate as the cause of active tuberculosis despite the presence positive tuberculin testing in several individuals. In patients with a previously positive Mantoux test result, if the induration size increases more than 10 mm, a new infection must be considered. Although unproved, some reactivation of old tuberculosis may actually be a second primary infection.

Patients who have a positive tuberculin skin test should undergo screening for active disease. A detailed review of signs and symptoms and a chest radiograph are central to screening. If the patient is symptomatic or the chest radiograph is consistent with old or new tuberculosis, three morning sputum samples for AFB and culture are recommended. Those with recent conversion of the tuberculin skin test (<2 years), immunocompromised patients, and household contacts with active tuberculosis can benefit from three early morning sputum AFB smears as well. Approximately 20% of women who are diagnosed with active tuberculosis result from screening programs. Most of these women are asymptomatic but infectious. About 5% of men with active tuberculosis are asymptomatic and are identified through screening programs.^43,44

Prophylactic Isoniazid Therapy

When taken as prescribed, isoniazid preventive therapy (10 mg/kg/day for children and 300 mg/day for adults for at least 6 continuous months; immunocompromised patients [HIV seropositive] require 12 months) is highly effective in preventing latent tuberculosis from progressing to clinically apparent disease. Multiple trials using isoniazid preventive therapy have reduced the incidence of active tuberculosis by 50% to 95% among patients at high risk for tuberculosis.¹⁶ The major reason for the wide variance in success is patient compliance for the extended therapy.⁴⁵ Despite its clear efficacy, health care providers in the United States are poor at identifying and treating eligible individuals. Although 75% of patients with active tuberculosis and often with risk factors had contact with a health care provider within 5 years of the diagnosis, only one third had a tuberculin skin test. Only 5% of patients with a positive tuberculin skin test result and risk factors were offered prophylactic isoniazid therapy.⁴⁶

Table 4 outlines the 1990 CDC recommendations for prophylactic isoniazid to limit the progression latent or asymptomatic infection to active tuberculosis or to prevent a primary tuberculosis infection. The age stratification at 35 years of age is based on the association between isoniazid-induced injury and age. Based on data from a CDC study,⁴⁷ the rates (per 1000 treated individuals) of isoniazid hepatitis is 1 case for those younger than 20 years old, 10 cases for those between 20 and 34, 23 cases for those between 35 and 49, 42 cases for those between 50 and 64, and 26 cases for those older than 64. Between 1972 and 1988 in the United States, 1,084,760 persons were started on isoniazid preventive therapy, and 655,867 (60%) finished their therapy. The death rate per 100,000 was 14 by intention-to-treat standards and 23.2 for those who completed treatment. Sixty-one percent of patients who die of isoniazid toxicity are older than age 35.⁴⁸ In a study of pregnant or postpartum (>1 year) Hispanic women in southern California,⁴⁹ the isoniazid-related clinical hepatitis case rate was 2.6 per 1000 women-years for those between the ages of 15 and 34 and 22.7 for women between the ages of 35 and 44. In this population of 3681 women undergoing preventive isoniazid therapy, two deaths (death rate of 54 per 100,000) occurred that were felt to be related to isoniazid toxicity. The deaths occurred after 4 and 7 months of therapy and at 3 and 4 months postpartum, respectively. This apparent increase in death rate (14 to 54 per 100,000) has increased reluctance on the part of internists and medical infectious disease specialists to start preventive isoniazid during pregnancy.

TABLE 4. 1990 Centers for Disease Control and Prevention Recommendations for the Use of Preventive Therapy for Tuberculosis in the United States

Delay may not be wise choice for a variety of reasons. First, the risk of active tuberculosis is much greater than the risk of isoniazid-related hepatitis or death. The risk of active tuberculosis within high-risk or high-prevalence populations, who do not receive preventive chemotherapy, is 2% in the subsequent year. Even with aggressive chemotherapy, 5% die (case death rate of 1000 per 100,000). With isoniazid therapy, the incidence of clinical hepatitis and death in pregnant and postpartum women are 260 and 54 per 100,000, respectively. Second, the efficacy of isoniazid is clear. Among household contacts of index cases of active tuberculosis the incidence of active tuberculosis fell 61% (1550 to 610 cases per 100,000).⁵⁰ Of the tuberculin-positive women who enter nursing homes, 2.3% develop active tuberculosis without preventive chemotherapy; with isoniazid preventive chemotherapy, the incidence is less than 0.3%. Among women whose tuberculin test converted, 7.6% develop active tuberculosis without preventive chemotherapy and only 0.2% with isoniazid preventive chemotherapy.^48,50,51 Third, isoniazid-related hepatotoxicity is usually not an acute process, taking weeks to develop. Often, there is a component of patient or physician delay in the identification of symptoms. Biochemical monitoring of liver function tests may reduce the mortality considerably. Fourth, the recommendation is based on one study⁴⁹ with 28% of subjects lost to follow up, small numbers (two deaths among treated group and one death in nonpregnant group). These deaths occurred months postpartum, and the recommendation to delay prophylactic isoniazid therapy until 3 to 6 months postpartum may increase the risk of active tuberculosis and not change the risk of hepatitis or death. The study also had an inappropriate comparison group with very different demographics, and it failed to control for the concomitant other hepatotoxic agents such as alcohol or acetaminophen.

Pregnant or postpartum women should receive isoniazid preventive therapy (300 mg/day for 6 continuous months and 12 months for immunocompromised individuals), regardless of gestational age, in high-risk or high-prevalence populations when the tuberculin test result is positive. In low-risk populations, isoniazid should be started after the first trimester.

Liver function tests should be obtained at baseline and every month. The patient is educated about compliance, hepatotoxic drugs and medications, and the signs and symptoms of hepatitis. Isoniazid blocks the conversion of substrate to pyridoxal phosphate and increases the urinary excretion of vitamin B₆. Deficiency of vitamin B₆ is associated with developmental defects in laboratory animals and a pyridoxine responsive-peripheral neuropathy. Because pregnancy is associated with low pyridoxine levels, 50 mg/day of vitamin B₆ is recommended during isoniazid therapy.⁵² Prenatal vitamins contain 2 to 10 mg of pyridoxine. Isoniazid has been linked to vitamin K deficiency in the newborn. Supplemental vitamin K (10 mg) is recommended in the last 2 weeks of pregnancy and in the newborn.

Bacille Calmette-Guérin Vaccination

In 1921, Dr. Weil Halle administered a live-attenuated strain of M. bovis (i.e. bacille Calmette-Guérin) to a child at risk for acquired tuberculosis. The child experienced no untoward effects. By 1945, BCG vaccination was well established in Europe, although less so in the United States. Starting in 1935, controlled trials were initiated to determine the efficacy of BCG vaccination in reducing tuberculosis morbidity.³⁵ The protective effect ranged from a risk of increased morbidity in the treated group of 56% to an 80% reduction in tuberculosis morbidity. Clemens and coworkers³⁵ reviewed in detail the biases and methodologic errors in the trials. He measured allocation safeguards, equidistribution of tuberculosis risk, surveillance bias, diagnostic testing bias, diagnostic interpretation bias, and adequate statistical precision. The best designed studies had similar beneficial effects, BCG reduced tuberculosis by 75% to 80%. Currently, BCG is not given routinely to any population in the United States. It may be considered in neonates on a case-by-case basis in unavoidable tuberculosis exposure especially MDR-TB. In developing countries where active tuberculosis is common, prolonged chemotherapy is problematic, and the delivery of health care is limited, BCG vaccination remains a preventive modality with important public health benefits.

The issue for most American practitioners is among foreign-born immigrants who may or may not have had BCG vaccine. Many developing countries have public health policies that recommend routine vaccination of infants and sometimes of older children. However, the BCG vaccination coverage is only partial. Only 10% to 40% of eligible infants are vaccinated in Central and South America, with the exception of Cuba, Chile, and Costa Rica.³⁷ If the patient describes a vaccination for tuberculosis, the interpretation of the tuberculin skin test depends on the age at vaccination and the interval to skin testing. In Quebec,³⁶ BCG vaccination increased the incidence of positive tests (>10 mm) from 2.6% in unvaccinated individuals to 12.6% in vaccinated individuals. During tuberculin skin testing as an adult, 4.1% of unvaccinated individuals had a positive skin test, 10.3% if they were vaccinated as infants, and 25.5% if they were vaccinated when older. When the children were tuberculin skin tested in the sixth grade, the incidence of positive testing was 1.5% for unvaccinated, 4.9% for vaccinated as an infant, or 12.5% for vaccinated when older. Pregnancy does not seem to affect the reaction to tuberculin skin testing in BCG-vaccinated pregnant women.³⁸

Therapeutic Management of Tuberculosis

All patients with microbiologically confirmed tuberculosis (i.e. positive acid-fast smear or positive culture), chest radiography suggesting active disease or cavitary lesions, or extrapulmonary lesions are considered to have active tuberculosis. All patients require multiagent therapy and intense follow-up for progression of disease and to confirm compliance.⁵³Table 5 describes the antibiotics and toxicities used in the treatment of active tuberculosis.

TABLE 5. Standard Therapy and Toxicities For Active Tuberculosis

The initial treatment of active tuberculosis in the nonpregnant woman should include four drugs (i.e. isoniazid, rifampin, pyrazinamide, and ethambutol or streptomycin) and continued for 6 months. Immunocompromised patients (e.g. HIV seropositive) require 12 months of four-agent therapy. Pregnant women should receive triple therapy (i.e. isoniazid, rifampin, and ethambutol) for 9 months. For populations in which the likelihood of resistance to any antibiotic is more than 4%, pyrazinamide should be added. Breast-feeding mothers should receive four-agent therapy. When a woman with active tuberculosis delivers without therapy for tuberculosis, both the mother and infant are treated. The issue of whether to separate the mother from her neonate is controversial. Many physicians advocate at least a 1 week of maternal therapy before returning the infant to its mother. This is very difficult for the breast-feeding mother. If the mother and the infant are receiving chemotherapy, there does not seem to be reasonable to separate them. Household contacts should be screened for tuberculosis before discharge. Table 6 describes the fetal and breast-feeding effects of chemotherapy for active tuberculosis.

TABLE 6. Chemotherapy for Tuberculosis During Pregnancy and Breast-feeding

Isoniazid, ethambutol, and rifampin are considered safe to use in pregnancy because the risk of defects in human birth appears to be no greater than baseline incidences. Although the risk of congenital ototoxicity is small, streptomycin should not be used in pregnant women unless required by drug sensitivities. In breast-feeding women, the optimal drug combinations should not be changed. The infant should be managed as if he or she were not breast-feeding; the recommended doses for neonates should not be altered.

Patients under therapy need close follow-up. Liver function tests should be obtained monthly and the patient should be educated concerning hepatitis. At 2-week intervals after the initiation of therapy, three early morning sputum samples are obtained for AFB smears. This is repeated at 1, 2, 3, and 6 months after the start of therapy. The AFB smear should be negative by 2 months; if not negative by 4 months, drug resistance must be suspected. Two drugs for which the isolate is sensitive are substituted in the regimen. A chest radiograph is obtained every 2 months until cure is recognized.

A major cause of drug-resistant tuberculosis and subsequent treatment failure is the patients noncompliance with the antibiotics. Treatment failure and multidrug resistance are life-threatening events. Noncompliance leads to death and more new infections in the community. Directly observed therapy (DOT) is an effective method of increasing compliance. One study in New York City reported a compliance rate of only 11% with standard follow-up, but the rate of compliance rose to 98% after the initiation of DOT.⁵³ The State of North Carolina recommends that all patients under therapy for tuberculosis have DOT.

Environmental Control of Tuberculosis

The environmental control of tuberculosis refers to building-associated interventions to prevent institutional airborne transmission from unsuspected cases or diagnosed cases before they are treated. The airborne infectious nuclei of tuberculosis (1 to 5 μm) allow wide dispersion through airflow patterns in the hospital. In the late 1950s, a classic experiment in hospital infection control was conducted at the Baltimore Veterans Administration Hospital by R.L. Riley.^57,58 All the air from a chronic tuberculosis ward was pumped through a guinea pig colony in another room. The guinea pigs were skin tested monthly, and bacilli recovered from infected guinea pigs were matched by drug sensitivity patterns of the bacilli recovered from the patients. An intervention was conducted in the second 2 years of the study; the airflow was split, and one half was disinfected with ultraviolet light before exposure of a colony of guinea pigs. These experiments established that tuberculosis is an airborne disease; the infectiousness of a patient varies widely (1 to 250 infectious nuclei per hour) and is not well correlated with symptoms; an average concentration of one infectious nuclei per 11,000 cf in the well-ventilated room was sufficient to causes an tuberculosis rate similar to human epidemics. Ultraviolet light is an effective intervention.

The most frequent interventions in clinics or wards include increased ventilation, air filtration, and ultraviolet irradiation.⁵⁹ Ventilation is the centerpiece of environmental control of tuberculosis. Engineering designs control dispersion of the organism (negative pressure gradient), dilute the concentration of tuberculosis per unit area (air exchanges), direct the contaminated air to the outside, or in-line ultraviolet irradiation.

Directional airflow is the creation of a room pressure that is less than the air pressure of adjacent rooms or hallways. Negative pressure is achieved by exhausting isolation room air faster than it can enter. This limits the dispersion of tuberculosis outside of the isolation room. The challenges to the maintenance of negative pressure include unrecognized air leaks and opening and closing doors and windows. The isolation room needs to be sealed except for a portal entry. A small anteroom with intermediate pressures can buffer the entry exchange of outside air. Entry and exit to the isolation room is strictly limited.

Direct exhaust airflow directs the contaminated air to outside dilution, filtration, or germicidal irradiation. Many jurisdictions require that contaminated air be exhausted at least 25 feet from the nearest entrance to the building (e.g. door, window) to prevent reentry of tuberculosis. As this requirement is hard to accomplish with reconstruction but easier in new construction; the airflow is often directed to filters or germicidal ultraviolet irradiation.

Negative pressure and direct ventilation are critical for limiting the spread of tuberculosis to the facility. In-room health care providers and visitors are not protected by the latter precautions. Dilution ventilation is a major technique used to reduce the risk to the individual who enters the room. Dilution ventilation is measured by air exchanges per hour. One air exchange per hour (ACH) can reduce the concentration of infectious particles by 63%, 3 ACHs by 95%, 6 ACHs by 99.75%, and 10 ACHs by more than 99.99%. The current standard for tuberculosis isolation rooms is 6 ACHs, with at least 2 ACHs being fresh outdoor air.

The two primary supplemental environmental control methods for tuberculosis are air filtration and ultraviolet germicidal irradiation. Mechanical filtration has been employed to control particulate dust, aerosols, and microorganisms for many years. Particles with an aerodynamic diameter of less than 0.3 μm are most difficult to remove by filtration. Because of its small size (0.5 to 5 μm), control of tuberculosis requires a highly efficient filter (i.e. HEPA filter). These filters have a removal efficiency of 99.97% at 0.3-μm particle size. HEPA filters are used in exhaust ventilation ducts, in modular air filtration units, for enhancing existing room ventilation, or in portable booths or canopy units where the need is temporary. Because of their efficiency and predilection to become clogged and block air flow, HEPA filters require intense and regular maintenance.

Ultraviolet germicidal irradiation has a lethal effect on a variety of microorganisms, including tuberculosis. The germicidal range for ultraviolet radiation is in the range of 220–300 nm with a maximum at 265 nm. Health care workers, visitors, and patients are at risk with direct exposure. Prolonged exposure to ultraviolet radiation among health care workers can result in erythema (sunburn), prokeratitis, and conjunctivitis.

There are three different applications of germicidal radiation to control tuberculosis within facilities: upper room irradiation, in-duct lamps, and recirculating room air cleaners. Normal in-room air movement carries the infectious particles to the ultraviolet light installed to irradiate the upper room near the 9-foot ceilings. The cleaning capacity of upper room ultraviolet is approximately 1.5 to 2.0 air exchange hour equivalents. With in-line duct ultraviolet, the cleaning capacity is equivalent to the air exchange rate. The efficiency of the cleaning is proportional to the flow rate; the greater the ACH, the less efficient is the cleaning power of ultraviolet radiation. Regardless of the lessened efficiency at high rates of ACH, there is significant increase in cleaning power when both methods are used. Ultraviolet irradiation is often combined with HEPA filters in modular units. Essentially, 100% of the tuberculosis particles are sterilized during each pass. The cleaning efficiency of ultraviolet irradiation is proportional to the ACH of the unit.

Effective environmental control of tuberculosis requires much more that placing equipment.⁶⁰ An entire program with administrative support, technical support, monitoring systems, quality assurance, and staff and patient education is necessary for successful environmental control of tuberculosis. The advent of MDR-TB with its attendant 50% morbidity or mortality rate demands an appropriate system response to protect patients, visitors, and health care workers.

Tuberculosis is an increasingly greater challenge for the obstetric care provider. Missing an active case of tuberculosis can be a lethal mistake for the care provider and the patient if the organism is resistant to several antibiotics. The obstetrician should have a low threshold for preforming a tuberculosis skin test (Mantoux Test) on pregnant women, especially those with household tuberculosis contacts, respiratory symptoms, or members of high-risk populations. Induration of more than 4 mm in HIV-seropositive populations, more than 9 mm in high-risk populations, and more than 14 mm in low-risk populations are considered evidence of current or past tuberculosis infection. Three morning sputum samples with a chest radiograph are essential parts of the workup for active disease. When this workup is negative, most pregnant women should be treated with isoniazid (300 mg/day for 6 months or 12 months for HIV-seropositive patients). Liver function tests should be obtained at baseline and each month thereafter. Pyridoxine (50 mg/day) and vitamin K (10 mg/day) are prescribed in the last 2 weeks of pregnancy. Regardless of the results of the tuberculin skin test, any pregnant women with signs or symptoms of tuberculosis should have three morning sputum examinations for AFB smear and culture and a chest radiograph. Bronchoscopy with bronchial brushings can help in the diagnosis of tuberculosis in these symptomatic patients.

Tuberculosis during pregnancy is treated with isoniazid, rifampin and ethambutol for 6 months. DOT is recommended to improve compliance. A team approach (i.e. obstetrician, infectious disease specialist, and public health nurse) maximizes the likelihood of successful therapy. As long as the infant is receiving prophylactic isoniazid and the mother has been under therapy for more than 6 days, the infant may breast-feed while the mother is receiving her four-agent antibiotic therapy. The mother and infant should be monitored for hepatitis and pro-vided with prophylactic Pyridoxine and vitamin K therapy.

1. Connolly M, Nunn P: Women and tuberculosis. World Health Stat Q 49: 115– 119, 1996

2. Dolin PJ, Rauzglione MC, Kochi A: Global tuberculosis incidence and mortality during 1990-2000. Bull World Health Organ 72: 213– 220, 1994

3. Kochi A: Tuberculosis: Distribution, risk factors, mortality. Immunobiology 191: 325– 336, 1994

4. Jereb JA, Kelly GD, Dooley SW et al: Tuberculosis morbidity in the United States: Final Data, 1990. MMWR Morb Mortal Wkly Rep 40(SS-3): 23– 27, 1990

5. Strong MA: The health of adults in the developing world: The view from Bangladesh. Health Transit Rev 2: 215– 224, 1992

6. Over M: The consequences of adult ill health. In Feachem RGA (ed): The Health of Adults in the Developing World. New York: Oxford University Press, 1992

7. Beck-Sague C, Dooley SW, Hutton MD et al: Hospital outbreak of multidrug-resistant Mycobacterium tuberculosis infections. JAMA 268: 1280– 1286, 1992

8. Catanzaro A: Nosocomial tuberculosis. Am Rev Respir Dis 125: 559– 562, 1982

9. Pearson ML, Jereb JA, Frieden TR et al: Nosocomial transmission of multidrug-resistant Mycobacterium tuberculosis. Ann Intern Med 117: 191– 196, 1992

10. Frieden TR, Sterling T, Pablos-Mendez A et al: The emergence of drug-resistant tuberculosis in New York City. N Engl J Med 328: 521– 526, 1993

11. Houk VN, Kent DC, Baker JH et al: The Byrd study. Arch Environ Health 16: 4– 6, 1968

12. Dankner WM, Waecker NJ, Essey MA et al: Mycobacterium bovis infections in San Diego: A clinicoepidemiologic study of 73 patients and a historical review of a forgotten pathogen. Medicine (Baltimore) 72: 11– 37, 1993

13. Strumpf J, Tsang AY, Schork MA, Weg JG: The reliability of gastric smears by auramine-rhodamine staining technique for the diagnosis of tuberculosis. Am Rev Respir Dis 114: 971– 976, 1976

14. Ellner JJ, Hinman AR, Dooley SM et al: Tuberculosis symposium: Emerging problems and promise. J Infect Dis 168: 537– 551, 1993

15. Bates JH, Stead WW: The history of tuberculosis as a global epidemic. Med Clin North Am 77: 1205– 1217, 1993

16. Haas DW, Des Prez RM: Mycobacterium tuberculosis: Principles and Practice of Infectious Diseases, pp 2215–2216. 4th ed. New York: Churchill Livingstone, 1995

17. Onorato IM, McCray E, Field Services Branch: Prevalence of human immunodeficiency virus infection among patients attending tuberculosis clinics in the United States. J Infect Dis 165:87–92, 1992

18. Styblo K: Recent advances in epidemiological research in tuberculosis. Adv Tuberc Res 20: 1– 63, 1980

19. Selwyn PS, Hartel D, Lewis VA et al: A prospective study of the risk of tuberculosis among intravenous drug users with human immunodeficiency virus infection. N Engl J Med 320: 545– 555, 1989

20. Carter EJ, Mates S: Tuberculosis during pregnancy: The Rhode Island experience, 1987 to 1991. Chest 106: 1466– 1470, 1994

21. Margono F, Mroueh J, Garely A et al: Resurgence of active tuberculosis among pregnant women. Obstet Gynecol 83: 911– 914, 1994

22. Vallejo JG, Starke JR: Tuberculosis and pregnancy. Clin Chest Med 13: 693– 707, 1992

23. Good JT, Iseman MD, Davidson PT et al: Tuberculosis in association with pregnancy. Am J Obstet Gynecol 140: 492– 498, 1981

24. Snider DE, Layde PM, Johnson MW, Lyle MA: Treatment of tuberculosis during pregnancy. Am Rev Respir Dis 122: 65– 79, 1980

25. Schaefer G, Zervoudakis IA, Fuchs FF, David S: Pregnancy and pulmonary tuberculosis. Obstet Gynecol 46: 706– 715, 1975

26. Centers for Disease Control and Prevention: Epidemiologic notes and reports: Tuberculosis among pregnant women—New York City, 1985-1992. MMWR Morb Mortal Wkly Rep 42:605–614, 1993

27. De March P: Tuberculosis and pregnancy: Five- to ten-year review of 215 patients in their fertile age. Chest 68:800–804; 1975

28. Mofenson LM, Rodriguez EM, Hershow R et al: Mycobacterium tuberculosis infection in pregnant and non-pregnant women infected with HIV in the Women and Infants Transmission Study. Arch Intern Med 155: 1066– 1072, 1995

29. Margono F, Mroueh J, Garely A et al: Resurgence of active tuberculosis among pregnant women. Obstet Gynecol 83:911–914; 1994

30. Woodring JH, Vandiviere HM, Fried AM et al: Update: The radiographic features of pulmonary tuberculosis. AJR Am J Roentgenol 146: 497– 506, 1986

31. Centers for Disease Control and Prevention: American Thoracic Society: Diagnostic standards and classification of tuberculosis. Am Rev Respir Dis 142:725–735, 1990

32. Snider DE Jr: The tuberculin skin test. Am Rev Respir Dis 125(Pt 2): 108– 118, 1982

33. Huebner RE, Schein MF, Bass JB: The tuberculin skin test. Clin Infect Dis 17: 968– 975, 1993

34. Present PA, Comstock GW: Tuberculin sensitivity in pregnancy. Am Rev Respir Dis 112: 413– 416, 1975

35. Clemens JD, Chuong JJH, Feinstein AR: The BCG controversy: A methodological and statistical reappraisal. JAMA 249: 2362– 2369, 1983

36. Menzies R, Vissandjee B: Effect of bacille Calmette-Guérin vaccination on tuberculin reactivity. Am Rev Respir Dis 145: 621– 625, 1992

37. Luelmo F: BCG vaccination. Am Rev Respir Dis 125: 70– 72, 1982

38. Sepulveda RL, Gonzalez B, Gerszencveig R et al: The influence of BCG immunization on tuberculin reactivity in healthy Chilean women in the third trimester of pregnancy. Tuberc Lung Dis 76: 28– 34, 1995

39. Centers for Disease Control and Prevention: Purified protein derivative (PPD)—Tuberculin anergy and HIV infection. Guidelines for anergy testing and management of anergic persons at risk of tuberculosis. MMWR Morb Mortal Wkly Rep 40(RR-5):27–32, 1991

40. Grzybowski S, Allen EA: The challenge of tuberculosis in decline: A study based on the epidemiology of tuberculosis in Ontario, Canada. Am Rev Respir Dis 90: 707– 20, 1964

41. Centers for Disease Control and Prevention: Tuberculin reactions in apparently healthy HIV seropositive and HIV-seropositive women: Uganda. MMWR Morb Mortal Wkly Rep 39:638–646, 1990

42. Chin DP, Osmond D, Page-Shafer K et al: Reliability of anergy skin testing in persons with HIV infection: The pulmonary complications of HIV infection study group. Am J Respir Crit Care Med 153(6 Pt 1): 1982– 198, 1996

43. Centers for disease Control: American Thoracic Society, American Academy of Pediatrics, Infectious Disease Society of America: Control of tuberculosis in the United States. Am Rev Respir Dis 146:1623–1633, 1992

44. Kvale PA, Hansen NI, Markowitz N et al: Routine analysis of induced sputum is not an effective strategy for screening persons infected with human immunodeficiency virus for Mycobacterium tuberculosis or Pneumocystis carinii: Pulmonary Complications of HIV Infection Study Group. Clin Infect Dis 19: 410– 416, 1994

45. Centers for Disease Control and Prevention: The use of preventive therapy for tuberculous infection in the United States: Recommendations of the advisory committee for elimination of tuberculosis. MMWR Morb Mortal Wkly Rep 39(RR-8):9–12, 1990

46. Glassroth J, Bailey WC, Hopewell PC et al: Why tuberculosis is not prevented. Am Rev Respir Dis 141 (5 Pt 1): 1236– 1240, 1990

47. Rose DN, Schechter CB, Silver AL: The age threshold for isoniazid chemoprophylaxis. JAMA 256: 2709– 2713, 1986

48. Steele MA, Burk RF, DesPrez RM: Toxic hepatitis with isoniazid and rifampin: A meta-analysis. Chest 99: 465– 471, 1991

49. Franks AL, Binkin NJ, Snider DE et al: Isoniazid hepatitis among pregnant and postpartum Hispanic patients. Publ Health Rep 104: 151– 155, 1989

50. Sbuder DE Jr, Caras GJ: Isoniazid-associated hepatitis deaths: A review of available information. Am Rev Respir Dis 145: 494– 497, 1992

51. Centers for Disease Control and Prevention: The use of preventive therapy for tuberculous infection in the United States: Recommendations of the Advisory Committee for the Elimination of Tuberculosis. MMWR Morb Mortal Wkly Rep 39(RR-8):9–12, 1990

52. Atkins JN: Maternal plasma concentration of pyridoxal phosphate during pregnancy: Adequacy of vitamin B₆ supplementation during isoniazid therapy. Am Rev Respir Dis 126: 714– 716, 1982

53. Centers for disease Control and Prevention: Initial therapy for tuberculosis in the era of multidrug resistance recommendations of the advisory council for the elimination of tuberculosis. MMWR Morb Mortal Wkly Rep 42(RR-7):1–8, 1993

54. Snider DE, Powell KE: Should women taking anti tuberculosis drugs breast-feed? Arch Intern Med 144: 589– 590, 1984

55. Briggs GG, Freeman RK, Yaffe SJ: Drugs in Pregnancy and Lactation. 4th ed. Baltimore: Williams & Wilkins, 1994

56. Lawrence RA: Breastfeeding: A Guide for the Medical Profession. 4th ed. St. Louis: CV Mosby, 1994

57. Riley RL, Mills CC, Nyka W et al: Aerial dissemination of pulmonary tuberculosis: A two year study of contagion in a tuberculosis ward. Am J Hygiene 70: 185– 196, 1959

58. Riley RL, Nardell EA: Clearing the air: The theory and application of ultraviolet air disinfection. Am Rev Respir Dis 139: 1286– 1294, 1989

59. Nagin D, Pavelchak N, London M et al: Control of tuberculosis in the workplace: engineering controls. Occup Med 9: 609– 630, 1994

60. Watson LH, Rosen JD: Educating workers about tuberculosis. Occup Med 9: 681– 694, 1994