Ariel Pablos-Mendez

Antimicrobial resistance in previously susceptible organisms occurs wherever antibiotics are used for the treatment of infectious diseases in humans and animals (Neu, 1992). Because of increasing antibiotic use and misuse over the past decades, resistance has emerged in all kinds of microorganisms—including Mycobacterium tuberculosis—posing new challenges for both clinical management and control programs (Dooley et al., 1992; Kochi et al., 1993).

Resistance of M. tuberculosis to antibiotics results from artificial amplification of spontaneous mutations in the genes of the tubercle bacilli. Treatment with a single drug—due to irregular drug supplies, inappropriate prescription practices, or poor adherence to treatment—suppresses the growth of susceptible strains to that drug but permits the multiplication of drug-resistant strains. This phenomenon is called acquired resistance. Subsequent transmission of such resistant strains from an infectious case to other persons leads to disease that is drug resistant from the outset, a phenomenon known as primary resistance.

Dramatic outbreaks of multidrug-resistant tuberculosis in HIV-infected patients in the United States (Centers for Disease Control, 1991; Edlin et al., 1992) and Europe (Monno et al., 1991; Herrera et al., 1996) recently focused international attention on the emergence of strains of M. tuberculosis that are resistant to antimycobacterial drugs (Frieden et al., 1993). Multidrug-resistant tuberculosis—defined as resistance to the two most important antimicrobial drugs, isoniazid and rifampin—is a signifi-

cant threat to tuberculosis control. Patients infected with strains resistant to multiple drugs are extremely difficult to cure, and the necessary treatment is much more toxic and expensive. Drug resistance is therefore a potential threat to the standard international method of tuberculosis control, the DOTS strategy, or “Directly Observed Treatment, Short-Course.”

In early 1994 the Global Tuberculosis Program of the World Health Organization (WHO) joined forces with the International Union Against Tuberculosis and Lung Disease and started the Global Project on Anti-Tuberculosis Drug Resistance Surveillance. The objectives of the project were to measure the prevalence of anti-tuberculosis drug resistance in several countries worldwide using standard methods and to study the correlation between the level of drug resistance and treatment policies in those countries.

The first step toward achieving the objectives was the development in 1994 of common definitions and guidelines. These focused on three major principles: (1) surveillance must be based on a sample of tuberculosis patients representative of all cases in the country, (2) primary and acquired drug resistance must be clearly distinguished in order to interpret the data correctly, and (3) proper laboratory performance must be assured.

The second step entailed establishment of a global network of Supranational Reference Laboratories (SRLs) to serve as reference centers for quality assurance of drug susceptibility testing. The network comprised 22 SRLs in 1997 and is expanding. The third step involved organizing a working group under the leadership of the WHO with representatives of national tuberculosis programs and research institutions from more than 50 countries to implement surveillance projects at the national level.

The first phase of the global project gathered results from 35 countries on five continents. Surveillance activities were conducted on approximately 50,000 tuberculosis cases in areas representing 20 percent of the world's population. Each study enrolled on average 1,200 tuberculosis patients (the range was 59 to 14,344). Testing for the presence of isoniazid and rifampin was conducted, and resistance to ethambutol and streptomycin was evaluated. Overall, agreement between the results of the SRL and the various national reference laboratories (NRLs) was 96 percent. All but three countries were able to distinguish between primary and acquired resistance.

Information was obtained from cases with effectively no previous treatment, thus reflecting the transmission of strains that were already resistant. The prevalence of resistance to any drug ranged from 2 percent (Czech Republic) to 40 percent (Dominican Republic), with a median value of 9.9 percent. Primary resistance to all four drugs tested was found in a median of 0.2 percent of the cases (range of 0 to 4.6 percent). Primary multidrug-resistant tuberculosis was found in every country surveyed except Kenya, with a median prevalence of 1.4 percent and a range of 0 (Kenya) to 14.4 percent (Latvia).

The presence of acquired drug resistance reflects more recent case mismanagement. The populations assessed for this part of the study were patients who were treated for a month or longer in the past. As expected, the prevalence of acquired drug resistance was much higher than that of primary drug resistance. The prevalence of acquired resistance to any drug ranged from 5.3 percent (New Zealand) to 100 percent (Ivanovo Oblast, Russia), with a median value of 36 percent. Resistance to all four drugs among previously treated patients was reported in a median of 4.4 percent of the cases (range 0 to 17 percent). The median prevalence of acquired multidrug-resistant tuberculosis was 13 percent, with a range of 0 percent (Kenya) to 54 percent (Latvia).

These findings are probably an underestimate of the magnitude of the problem worldwide because the countries surveyed had better than average tuberculosis control. Resistance to tuberculosis drugs is probably present everywhere in the world. Certainly, multidrug-resistant tuberculosis is present on all five continents; a third of the countries surveyed had levels of multidrug-resistant tuberculosis in more than 2 percent of new patients. In Latvia 30 percent of all patients presenting for treatment had multidrug-resistant tuberculosis. The region of Russia surveyed showed 5 percent of tuberculosis patients with multidrug-resistant tuberculosis. In the Dominican Republic 10 percent of tuberculosis patients had multidrug-resistant tuberculosis. In Africa the Ivory Coast has also witnessed the emergence of multidrug-resistant tuberculosis. Preliminary reports from Asia (India and China) show high levels of drug resistance

as well. In the state of Delhi, India, 13 percent of all tuberculosis patients had multidrug-resistant tuberculosis.

An important finding of the study was the higher prevalence of multidrug-resistant tuberculosis in countries categorized by the WHO as having poor control programs. Similarly, the higher the proportion of retreatment cases (the result of a poor program), the higher the levels of drug resistance among new patients. The use of standardized short-course chemotherapy regimens, on the other hand, was associated with lower levels of drug resistance.

First, drug resistance is ubiquitous. The Global Project found it in all countries surveyed. The levels of resistance to isoniazid are high, and continued failure to improve tuberculosis control will fuel multidrug resistance. Second, there are several hot spots around the world where multidrug-resistant tuberculosis prevalence is high and could threaten control programs. These include Latvia, Estonia, and Russia in the former Soviet Union, the Dominican Republic and Argentina in the Americas; and the Ivory Coast in Africa. Preliminary reports from Asia also show high levels of multidrug-resistant tuberculosis. Urgent intervention is needed in these areas.

Third, there is a strong correlation between both the overall quality of tuberculosis control and the use of standardized short-course chemotherapy and low levels of drug resistance. A high prevalence of multidrug-resistant tuberculosis is the result of therapeutic anarchy. Half the countries or regions with the worst tuberculosis control had primary MDR levels above 2 percent, compared with one-fifth of those with moderate control and none of the countries with the highest standard of tuberculosis control.

Finally, the level of multidrug-resistant tuberculosis is a useful indicator of national tuberculosis program performance. As shown by the global project and by previous experiences in Korea and New York, the prevalence of primary multidrug-resistant tuberculosis is a good “summary” indicator of the performance of tuberculosis control programs in recent years.

The network of SRLs of the global project is a model of international scientific collaboration in support of an important public health initiative. The benefits brought to NRLs by this network will extend beyond the global project itself. However, the results of the project thus far merely update the pessimistic view expressed by Fox (1977) about drug susceptibility testing (1977). Twenty years later, as drug resistance is emerging as a clinical and an epidemiologic concern, several laboratories in the world are finally capable of providing accurate and reliable drug susceptibility testing results, especially for isoniazid and rifampin.

With the implementation of a quality assurance program, internationally comparable results of drug susceptibility testing could be obtained. When future surveys are planned in a given country, an SRL should first conduct proficiency testing of the responsible NRL and together develop a scheme for quality control (World Health Organization/International Union Against Tuberculosis and Lung Disease, 1997). When an NRL does not exist or its drug susceptibility testing performance remains sub-optimal, it is preferable to have strains tested at the SRL itself. In addition to the accuracy of drug susceptibility testing, quality control needs to emphasize safety in the laboratory. For international transport of strains, adherence to international regulations is mandatory (World Health Organization, 1993).

One of the important findings of the global project is that all four laboratory methods for drug-susceptibility testing can achieve similar levels of accuracy (Canetti, 1965). Standard and economic variants of the proportion, the resistance ratio, the absolute concentration, and the radiometric BACTEC 460 methods have yielded similar drug susceptibility results, both between and within laboratories, for the four first-line anti-tuberculosis drugs evaluated. These findings justify the continued use of traditional drug susceptibility testing methods in institutions familiar with them or unable to afford more recent technology.

For efficient implementation of the global project, ideally, several regional networks should be established to conduct technical exchanges between the participating countries in the region. Geographical proximity and cultural similarity facilitate technical exchange and collaboration to implement local surveys. The regional offices of the WHO could play a critical role in the development of these networks by identifying target countries and experts in the region.

The Western Pacific Regional Office of the WHO established an organization of NRLs and SRLs in the Pacific region, modeled after the global SRL network. The Korean Institute of Tuberculosis carried out a quality assurance study on drug susceptibility testing in 1995 and 1996 in which

the NRLs of China, Hong Kong, Malaysia, Thailand, and Vietnam participated. The results of the first round of proficiency testing, implemented by the Research Institute of Tuberculosis in Tokyo and the Korean Institute of Tuberculosis in Seoul, showed a fairly acceptable concordance, except for rifampin susceptibility testing. Three NRLs showed an unacceptable concordance rate (<80 percent) with RMP tests on the first round of testing but improved considerably (93 to 100 percent) in the second round. All of the participating NRLs showed an acceptable concordance rate for isoniazid susceptibility testing in both rounds of proficiency testing. Additional improvement in drug-susceptibility testing proficiency is expected as quality assurance at the regional level improves and expands.

Finally, although the global project almost exclusively used conventional epidemiologic and laboratory methodology, recent developments in molecular technology and their applicability in drug susceptibility testing and epidemiologic studies of multidrug-resistant tuberculosis must be acknowledged. Current tests can rapidly and directly ascertain the drug susceptibility status of a given strain in clinical specimens without culture (Telenti et al., 1993). We now know that the same genetic mutations underlie primary and acquired drug resistance, whether in AIDS patients or not, inside lung cavities or in extrapulmonary sites, or in Europe or sub-Saharan Africa (Cole, 1997).

An exciting approach to rapid drug susceptibility testing is the polymerase chain reaction (PCR)-based amplification of specific genes involved in resistance to individual drugs. These tests are approved for use only in smear-positive pulmonary tuberculosis, and the results must be examined in clinical context. The current limitations and expense associated with this technology prevent its use in routine clinical practice. An easier approach is to detect growth of M. tuberculosis in the presence of a given drug using mycobacteriophage (Jacobs et al., 1993). While these methods are promising for rifampin drug susceptibility testing, they have not been tested in field situations, and standardization is required before they can be used outside research laboratories.

Molecular techniques are also increasing our understanding of the epidemiology of tuberculosis and drug resistance (World Health Organization, 1995). Specific strains of M. tuberculosis can be identified by restriction fragment length polymorphism DNA fingerprinting (Thierry et al., 1990) and new DNA microarrays. In addition, clustering of strains sharing a DNA fingerprint suggests an increased probability of recent M. tuberculosis transmission (Daley et al., 1992). These molecular tools have provided refined descriptions of point-source tuberculosis outbreaks (Alland et al., 1994) as well as the dissemination of specific drug-resistant strains (Bifani et al., 1996). Genetic epidemiologists are also reexamining the

century-old debate about recent transmission versus reactivation of latent infection in tuberculosis epidemiology (Friedman et al., 1995).

The global project has highlighted the need for action in the areas of surveillance, management, and research.

As part of the global project, surveillance reports from 40 additional countries were made available in the year 2000, and trends over time will be available in some of the initial 35 participating countries. The following strategy is being followed:

• The well-established network of SRLs is a model for standardized surveillance of drug resistance and should be maintained as a global resource. Additional SRLs have been added, and regional networks are being consolidated.

• There will be adequate assessment of the level of multidrug-resistant tuberculosis in large countries (e.g., China, India, Russia), with expansion of surveillance activities beyond the regions studied. Areas not adequately covered during the first phase of the Global Project are being targeted.

• Future surveys are to collect and analyze individual data on age, HIV coinfection, and country of birth and on the contribution of unregulated private sectors to drug resistance.

Countries without the DOTS tuberculosis control strategy need to implement it. This requirement is supported by the global project 's finding of an association of low resistance and high-quality tuberculosis control. Previous experience has also demonstrated decreases in resistance, even in multidrug-resistant tuberculosis, following the introduction of DOTS tuberculosis control.

The global project does not directly address the issue of treatment regimens. Based on previous experience and the generally low levels of drug resistance, however, no alterations are yet required for the standard regimens recommended by the WHO and the International Union Against Tuberculosis and Lung Disease in new patients with tuberculosis (World Health Organization, 1996). For the management of multidrug-resistant

tuberculosis, the reader is referred to “Guidelines for the Management of Drug-resistant Tuberculosis” (Crofton et al., 1997).

The results of the global project have brought into focus several important questions. First, efforts must be made to assess the transmissibility and clinical virulence of multidrug-resistant tuberculosis compared to disease caused by drug susceptible strains. In vitro studies in the 1950s (Cohn et al., 1954) and recent epidemiologic evidence suggest that multidrug-resistant strains of M. tuberculosis may be less virulent or transmissible than drug-susceptible organisms. Modeling exercises could then put in perspective the impact of multidrug-resistant tuberculosis in global and regional trends.

Second, the global project has highlighted the need to define the impact of multidrug-resistant tuberculosis on treatment outcomes under program conditions in developing countries. In most settings with relatively low numbers of multidrug-resistant tuberculosis, current practice (standard regimens without drug-susceptibility testing) may still achieve target cure rates. However, in settings with an unusually high prevalence of multidrug-resistant tuberculosis, concerns have been raised about the risk of worsening the problem with the use standardized regimens vis-à-vis individualized regimens (based on drug-susceptibility testing). Beyond anecdotal evidence and small case series, evidence in support of that concern is lacking. Most treatment failures are due to reasons other than multidrug-resistant tuberculosis. While in individual patients resistance can only increase with additional waves of treatment, experience around the world shows that resistance levels in the community can be reduced by implementing sound control policies with standardized regimens (Kim and Hong, 1992; Chaulet, 1993). Theoretically, however, there must be a threshold of MDR levels above which standardized regimens will not work and that may indeed promote additional drug resistance. In 1999 the WHO formed a DOTS-Plus Working Group to address these issues.

Finally, pharmaceutical companies are urged to develop new anti-tuberculosis drugs. Simple replacement of older products may not be a wise investment. The prime need for such drugs is to make DOTS strategies more efficient and to shorten the duration of treatment, thus making resistance less likely to emerge in the first place.

Multidrug resistance may or may not become an insurmountable barrier to well-implemented tuberculosis control programs. On the other hand, patients with clinically active disease caused by multidrug-resistant tuberculosis face uncertain prospects of successful treatment (Goble et al., 1993; Mitchison and Nunn, 1986), side effects from medication (when available), and the associated expense of the medication (Mahmoudi and Iseman, 1993). Confronted with a potentially fatal disease, clinicians are ethically compelled to offer every available treatment. However, antibiotic misuse should not be justified by the (genuine) concerns of clinicians, a phenomenon that underlies the emergence of multidrug-resistant tuberculosis. In a program context, futile efforts made for a few incurable cases may only drain resources that could be used to cure many patients with drug susceptible tuberculosis and prevent multidrug-resistant tuberculosis in the first place.

Routine drug-susceptibility testing for clinical management of all tuberculosis patients is simply out of the question in most countries where the disease is concentrated. Even when available, such results do not affect the regimen used to treat over 95 percent of tuberculosis patients. Patients whose treatment fails after two courses of the standardized regimen are likely to harbor multidrug-resistant tuberculosis (50 percent or more; Kritski et al., 1997; Mazouni et al., 1992) and should be referred for expert management. A specialized unit may be regarded as an expensive luxury in many countries, but second-line drugs should be available only to such centers.

Some experts argue that in countries with a high tuberculosis burden and few resources, drug susceptibility testing should be done for surveillance purposes only, not to guide therapeutic decisions in individual patients. In such countries, drug susceptibility testing may distract from the essential duty to perform smear microscopy. Feedback of drug susceptibility testing results to individual physicians may be futile where second-line agents are not available or are unaffordable. The results of drug susceptibility testing in laboratories with a low volume of tests (<300 per year) may not even be accurate (Nitta et al., 1996). More importantly, drug susceptibility testing results may cause confusion and prompt inappropriate retailoring of therapeutic regimens without improving their efficacy (i.e., changing standardized regimens in patients with, for example, isoniazid monoresistance; Mahmoudi and Iseman, 1993). Poor implementation of recommended regimens may then lead to the therapeutic chaos in which multidrug-resistant tuberculosis thrives.

Policy makers and responsible clinicians should keep in mind that the

cheapest multidrug-resistant tuberculosis treatment regimen is 100 times more expensive than the best first-line regimen (Crofton et al., 1997). Difficult choices must be made, since less than a third of patients with tuberculosis currently have access to adequate management. Countries that have secured these basic strategies may then decide to devote additional resources to fight multidrug-resistant tuberculosis. With qualified laboratories and available second-line drugs, drug susceptibility testing has been recommended for all new cases to help tailor the best possible therapeutic regimens under expert supervision (Centers for Disease Control and Prevention, 1993).

The WHO estimates the global incidence of tuberculosis at 8 million cases per year, with most occurring in India and sub-Saharan Africa (World Health Organization, 1999). Extrapolating results from the global project, 300,000 of these new cases are due to multidrug-resistant tuberculosis. New diagnostic technologies that simplify the tasks of surveillance programs and clinicians in less developed countries would be most welcome. Such methods should be acceptable in terms of cost, accuracy, speed, and complexity (direct based as opposed to culture based). In particular, new methods to detect microbial resistance to rifampin (practically equivalent to MDR) may revolutionize current recommendations for global surveillance and facilitate the development of therapeutic regimens appropriate to each country or region.

Alland, D., G. E. Kalkut, A. R. Moss, R. A. McAdam, J. A. Hahn, W. Bosworth, E. Drucker, and B. R. Bloom. 1994. Transmission of tuberculosis in New York City: An analysis by DNA fingerprinting and conventional epidemiological methods. New England Journal of Medicine, 330:1710-1716.

Bifani, P. J., B. B. Plikaytis, V. Kapur, K. Stockbauer, X. Pan, M. L. Lutfey, S. L Moghazeh, W. Eisner, T. M. Daniel, M. H. Kaplan, J. T. Crawford, J. M. Musser, and B. N. Kreiswirth. 1996. Origin and interstate spread of a New York City multidrug-resistant Mycobacterium tuberculosis clone family. Journal of the American Medical Association, 275:452-457.

Canetti, G. 1965. Present aspects of bacterial resistance in tuberculosis. American Review of Respiratory Disease, 92:687-703.

Centers for Disease Control and Prevention. 1991. Nosocomial transmission of multidrug-resistant tuberculosis among HIV-infected persons: Florida and New York, 1988-1991. Morbidity and Mortality Weekly Report, 40:585-591.

Centers for Disease Control and Prevention. 1993. Tuberculosis control laws—United States, 1993: Recommendations of the advisory council for the elimination of tuberculosis. Morbidity and Mortality Weekly Report, 42(RR-15):1-13.

Chaulet, P. 1993. Tuberculose et transition épidemiologique: le cas de l'Algerie. Annals of the Institut Pasteur, 4:181-187.

Cohn, M. L., C. Kovitz, U. Oda, and G. Middlebrook. 1954. Studies on isoniazid and tubercle bacilli: The growth requirements, catalase activities, and pathogenic properties of isoniazid-resistant mutants. American Review of Tuberculosis, 70:641-664.

Cole, S. T. 1997. Molecular basis of drug resistance in M. tuberculosis. The Wellcome Trust Meeting on Antibiotic Resistance, United Kingdom, May 18-21.

Crofton, J., P. Chaulet, D. Maher, et al. 1997. Guidelines for the Management of Drug-Resistant Tuberculosis. WHO/Gtuberculosis/96.210. Geneva: World Health Organization.

Daley, C. L., P. Small, G. Schecter, G. K. Schoolnik, R. A. McAdam, W. R. Jacobs, Jr., and P. C. Hopewell. 1992. An outbreak of tuberculosis with accelerated progression among persons infected with human immunodeficiency virus. New England Journal of Medicine, 326(4):231-235.

Dooley, S. W., W. R. Jarvis, W. J. Marlone, and D. E. Snider. 1992. Multidrug-resistant tuberculosis. Annals of Internal Medicine, 117:257-259.

Edlin, B. R., J. L. Tokars, M. H. Grieco, J. T. Crawford, J. Williams, E. M. Sordillo, K. R. Ong, J. O. Kilburn, S. W. Dooley, K. G. Castrio, et al. 1992. An outbreak of multidrug-resistant tuberculosis among hospitalized patients with the acquired immunodeficiency syndrome. New England Journal of Medicine, 326:1514-1521.

Frieden, T., T. Sterling, A. Pablos-Méndez, J. O. Kilburn, G. M. Cauthen, and S. W. Dooley. 1993. The emergence of drug-resistant tuberculosis in New York City. New England Journal of Medicine, 328:521-26.

Friedman, C., M. Y. Stoekle, B. N. Kreiswirth, W. D. Johnson, Jr., S. M. Manoach, J. Berger, K. Sathianathan, A. Hafner, and L. W. Riley. 1995. Transmission of multidrug-resistant tuberculosis in a large urban setting. American Journal of Respiratory and Critical Care Medicine, 152:355-359.

Fox, W. 1977. The modern management and therapy of pulmonary tuberculosis. Proceedings of the Royal Society of Medicine, 70:4.

Goble, M., M. D. Iseman, L. A. Madsen, D. Waite, L. Ackerson, and C. R. Horsburgh, Jr. 1993. Treatment of 171 patients with pulmonary tuberculosis resistant to isoniazid and rifampin. New England Journal of Medicine, 328:527-532.

Herrera, D., R. Cano, and P. M. K. Godoy. 1996. Multidrug-resistant tuberculosis outbreak on an HIV ward: Madrid, Spain, 1991-1995. CDC Morbidity and Mortality Weekly Report, 45:330-333.

Kim, S. J., and Y. P. Hong. 1992. Drug resistance of Mycobacterium tuberculosis in Korea. Tubercle and Lung Disease, 73:219-224.

Kochi, A., B. Vareldzis, and K. Styblo. 1993. Multidrug resistant tuberculosis and its control. Research in Microbiology, 144:104-110.

Kritski, A., L. S. Rodriguez de Jesus, M. K. Andrade, E. Werneck-Barroso, M. A. Vieira, A. Haffner, and L. W. Riley. 1997. Retreatment tuberculosis cases: Factors associated with drug resistance and adverse outcomes. Chest, 111:1162-1167.

Jacobs, W. R., R. G. Barletta, R. Udani, J. Chan, G. Kalkut, G. Sosne, T. Kieser, G. J. Sarkis, G. F. Hatfull, and B. R. Bloom. 1993. Rapid assessment of drug susceptibilities of Mycobacterium tuberculosis by means of luciferase reporter phages. Science, 260:819-822.

Mahmoudi, A., and M. D. Iseman. 1993. Pitfalls in the care of patients with tuberculosis. Journal of the American Medical Association, 270:65-68.

Mazouni, L., N. Zidouni, F. Boulahbal, and P. Chaulet. 1992. Treatment of Failure and Relapse Cases of Pulmonary Tuberculosis Within a National Programme Based on Short Course Chemotherapy: Preliminary Report. Tuberculosis Surveillance Research Unit of the International Union Against Tuberculosis and Lung Disease, Progress Report 1992 2:36-42.

Mitchison, D. A., and A. J. Nunn. 1986. Influence of initial drug resistance on the response to short-course chemotherapy of pulmonary tuberculosis. American Review of Respiratory Disease, 133:423-430.

Monno, L., G. Angarano, S. Carbonara, S. Coppola, D. Costa, M. Quarto, and G. Pastore. 1991. Emergence of drug-resistant Mycobacterium tuberculosis in HIV-infected patients. Lancet, 337:852.

Neu, H. C. 1992. The crisis of antibiotic resistance. Science, 257:1064-1073.

Nitta, A. T., P. T. Davidson, M. L. de Koning, and R. J. Kilman. 1996. Misdiagnosis of multidrug-resistant tuberculosis possibly due to laboratory-related errors. Journal of the American Medical Association, 276:1980-1983.

Telenti, A., P. Imboden, F. Marchesi, D. Lowrie, S. Cole, M. J. Colston, L. Matter, K. Schopfer, and T. Bodmer. 1993. Detection of rifampin-resistance mutations in Mycobacterium tuberculosis. Lancet, 341:647-650.

Thierry, D., M. D. Cave, K. D. Eisenach, J. T. Crawford, J. H. Bates, B. Gicquel, and J. L. Guesdon. 1990. IS6110, an IS-like element of Mycobacterium tuberculosis complex. Nucleic Acids Research, 18:188.

World Health Organization. 1993. Safe shipment of specimens and infectious materials. Pp. 48-54 in Laboratory Biosafety Manual, 2nd ed. Geneva: WHO.

World Health Organization. 1996. Treatment of Tuberculosis: Guidelines for National Programmes. WHO/tuberculosis/96.199. Geneva: WHO.

World Health Organization Global Tuberculosis Programme. 1995. The Use of Restriction Fragment Length Polymorphism (RFLP) Analysis for Epidemiological Studies of Tuberculosis in Developing Countries Report of a WHO Review and Planning Meeting. WHO/tuberculosis/95.194. Geneva: WHO.

World Health Organization Global Tuberculosis Program. 1999 Global Tuberculosis Control: WHO Report. WHO/tuberculosis/99.250. Geneva: WHO (www.who.int/gtb/publications/globrep99/index.html).

World Health Organization/International Union Against Tuberculosis and Lung Disease. Global Working Group on Antituberculosis Drug Resistance Surveillance 1997. Guidelines for Surveillance of Drug Resistance in Tuberculosis. WHO/tuberculosis/96.216. Geneva: WHO.