Round Table Discussion

TB control and access to second-line drugs: better model needed

Marcos A. Espinal¹ & Richard Zaleski²

Pablos-Méndez et al. rightly point out that multidrug-resistant tuberculosis (MDR-TB) is not a major pandemic (see pp. 489 - 494). However, as drug-susceptible TB is a worldwide problem, the first priority for national TB programmes should be the implementation or expansion of the DOTS strategy. MDR-TB is in most cases a sign of poor programme performance, although there may be highly virulent strains spreading rapidly. A weak national programme can do more harm than good if its main focus is the widespread introduction of second-line drugs to manage this problem.

To tackle MDR-TB Pablos-Méndez et al. propose a matrix based on two variables: treatment success for new TB cases, and prevalence of primary MDR-TB. They propose that the use of second-line drugs should be limited to countries which belong in specified quadrants according to these two variables. We find it questionable that the management of a MDR-TB, or any other disease, should be based on only two variables. We live in a world in which the control of illness calls for modern multidisciplinary approaches (1). We will come back to this point.

The proposal of Pablos-Méndez et al. is difficult to accept for at least two reasons. Firstly, more than two variables are needed to decide if a country should treat MDR-TB. For instance, a country may score well on treatment success, have a low number of primary MDR-TB cases but still have a high number of treatment failure cases (a variable not taken into account in the proposed model), which are likely to have MDR (2). Such a country may need to implement management of MDR-TB as well as DOTS, regardless of its level of primary MDR and treatment success. Furthermore, treatment success could be a very misleading variable since it is the result of cure plus treatment completion. There are some examples of poor national TB programmes having high rates of success upon completion of treatment but low cure rates.

Secondly, do we really need cut-off points to manage a disease? On what basis can we choose 5% and not, say, 3% for MDR prevalence, or 70% and not, say, 60% for treatment success? No biological, statistical or epidemiological reason is given for choosing such cut-off points. A straightforward indication of the point at which to start management of MDR-TB could be helpful, but the issue is not that simple, and other matters need to be carefully looked at when taking such a decision. The assertion that DOTS can reduce MDR-TB has not been fully proved, although it is clear that short-course chemotherapy can prevent MDR-TB. Countries that have reduced MDR-TB have also used second-line drugs and it is not clear to what extent the use of both first-line and second-line drugs have contributed to reducing MDR-TB. It is also well known that short-course chemotherapy, one of the pillars of DOTS, only cures an unacceptably low fraction of MDR- TB (3).

The approach suggested by Pablos-Méndez et al. needs rethinking. First of all, any model for managing MDR-TB must recognize that such a decision has to be made by the countries concerned. It will depend on several national factors, including the resources available, the epidemiological profile, the status of TB control, and ethical and humanitarian issues. Certainly an economic threshold is likely to exist. A country choosing whether to manage MDR-TB may benefit from a comprehensive multidisciplinary assessment of its situation, in order to decide if such drugs are needed or not. If the decision to go ahead is made, the path to follow should be a strategy that includes - but is not limited to - DOTS to reduce transmission of MDR M. tuberculosis strains.

The international community needs to pursue a feasible and cost-effective strategy to manage MDR-TB, which enables countries to offer a cure to patients (4). Although current evidence is limited, there are indications that treatment of MDR in resource-limited setting with strong TB control programmes may be feasible and cost-effective (5). This information can benefit patients even in settings where MDR-TB rates are below or above the threshold proposed by Pablos-Méndez et al.

1. Porter J, Ogden J, Pronyk P. Infectious disease policy: towards the production of health. Health Policy and Planning 1999;14:322-8.

2. Sabogal I, Velarde N, Palhua M, Rodriguez F, Paredes M. Resistencia a farmacos antituberculosos en fracasos a un esquema de tratamiento primario. [Resistance to anti-TB drugs in primary treatment] In: Tuberculosis en el Peru. Informe 1997. Lima: Ministry of Health. p. 141-4.

3. Espinal MA, Kim SJ, Suarez PG, Kam KM, Khomenko AG, Migliori GB, et al. Standard short-course chemotherapy for drug resistant tuberculosis. JAMA 2000; 283:2537-45.

4. Espinal MA, Dye C, Raviglione M, Kochi A. Rational 'DOTS-Plus' for the control of MDR-TB. International Journal of Tuberculosis and Lung Disease 1999;3:561-3.

5. Suaréz PG, Floyd K, Portocarrero J, Alarcón E, Rapiti E, Ramos G, et al. Feasibility and cost-effectiveness of second-line drug treatment for chronic tuberculosis patients: a national cohort study in Peru. Lancet. In press 2002.

¹ World Health Organization, STOP TB Department, 1211 Geneva 27, Switzerland (email: espinalm@who.int).

² World Health Organization, Copenhagen, Denmark.

First requirement for control of multidrug- resistant TB: realism

Pedro Guillermo Suárez¹

The primary cause of an uncontrolled and increasing TB epidemic worldwide is the neglect of TB control programmes. This neglect is evidenced by lack of political support, scarce financial resources, and little or no leadership. Successful programmes in both industrialized and developing countries indicate that a DOTS strategy prevents multidrug-resistant tuberculosis (MDR-TB). The timely and appropriate diagnosis and treatment of new and previously-treated TB patients is the focus in DOTS. As Pablos-Méndez et al. suggest (see pp. 489 - 494), a DOTS-Plus strategy is needed to control MDR-TB only after the DOTS programme has been established and is being adequately implemented.

However, besides the epidemiological and operative factors involved in implementing a DOTS-Plus strategy, there is an international debate about the ethics and humanism involved. To deny treatment to patients with MDR-TB is to violate their human rights. Experience in Peru indicates the need for sustained and long-term efforts in preventing the emergence of MDR-TB with a DOTS strategy. Only then does it become possible to treat MDR-TB by applying a DOTS-Plus strategy in the context of an efficient, sustainable and comprehensive TB control programme.

The major area of controversy in applying a DOTS-Plus strategy is about the use of standardized or individualized regimens to treat MDR-TB in countries with limited resources. In high-income countries with a low incidence of TB and sufficient financial, technical and human resources, MDR-TB treatment with individualized regimens is based on drug susceptibility tests. The feasibility of using this approach in low- or medium-income countries has not been assessed. The options for using standardized or individualized regimens for MDR-TB in these countries should be examined in pilot projects. These should obtain comparable data and have the following aims: to develop an evidence-based approach; to design and implement the most appropriate strategy according to the epidemiology and operational conditions in each country; and to be subject to rigorous evaluation by international standards. A DOTS-Plus strategy should also be based on national and international technical assistance to tackle MDR-TB.

The major components for the implementation of a DOTS-Plus strategy would then be: (i) an efficient, effective and integrated TB control programme; (ii) first-line and second-line anti-TB drugs provided free of charge to each patient with MDR-TB; (iii) drug susceptibility tests for first-line and second-line anti-TB drugs, not charged to the patients; (iv) appropriately designed regimens for MDR-TB, standardized or individualized; (v) a reporting system for data management, monitoring and evaluation of individual and aggregated data on MDR-TB cases; (vi) community-based strategies, with the participation of local governments in order to enhance adherence to the regimens; and (vii) the adequate training and organization of health professionals responsible for the care of MDR-TB patients.

Countrywide public health and political commitment to sustaining the DOTS strategy remains the most important element for implementing DOTS-Plus strategies. In reality it may be the only means of achieving MDR-TB control.

¹ Former Medical Director of the National Tuberculosis Control Programme, Av. Alameda los Horizontes Mz. D8-Lt.10, Urb. Los Cedros de Villa, Chorrillos, Lima, Peru (email: pgsuarez@terra.com.pe).

Multidrug-resistant TB - unexamined costs and complexities

Tim Cullinan²

Pablo-Méndez et al. (see pp. 489-494) touch on many problems that have to be considered carefully before even the first steps towards comprehensive control of multidrug-resistant tuberculosis (MDR-TB) can be contemplated. I will mention only three of these, and only very briefly.

The first is the administration of second-line drugs, once they have been obtained. Such slight experience as currently exists in cohort treatment of MDR-TB comes mainly from relatively controlled situations. In these places it has been possible to marshal the resources needed to ensure compliance over the long period of treatment and the management of its manifold side-effects. Even in an urban situation such as Lima, Peru, the cost of establishing the basis for ambulatory care was enormous, and those arrangements cannot yet be contemplated for the rural areas.

As the authors point out, case-holding and compliance are the major difficulties in first-line DOTS programmes and a potent cause of MDR-TB. Yet few of the situations which might qualify as MDR sites on the authors' matrix have anything like the support mechanisms in place to ensure an effective control programme. Also, control must, presumably, imply legislative regulation of the drug supply.

A second issue is how countries, until now supposedly not exposed to heavy burdens of MDR-TB, discover they have problems. Many countries, especially in sub-Saharan Africa, have managed to institute effective DOTS programmes on the basis of microscopy alone. Few have culture facilities for the majority of patients, let alone the quality control mechanisms to maintain their performance. In these countries, the outcome "treatment failure" tends to play a much smaller role than elsewhere, and it is rarely subjected to further investigation. To set up culture facilities in order to investigate and treat MDR-TB would entail costs that few of these countries could contemplate.

A third difficulty, common to all too many WHO papers, is to base statements about a whole country on findings from a small, often atypical, model area. Yet the difference between a special, usually heavily funded, trial area and the rest of a country is often greater than differences between countries. With suitable caveats, such matters are not of great importance in a summary paper except that they may inhibit discussion of the potential for disease control in isolated circumstances such as prisons, displaced persons' camps etc. Such sites may meet criteria for MDR-TB control which exist nowhere else in the country and they should be considered eligible, even in isolation.

These thoughts are proffered only as an addendum to what seems to me a very useful and well-written paper.

² Health Adviser, Merlin, 5 - 13 Trinity St. London SE1 1DB, England (email: Tim.Cullinan@merlin.org.uk).

DOTS-Plus in the Philippines, a high-burden country: funds needed

Thelma Tupasi¹

Multidrug-resistant tuberculosis (MDR-TB) is a global problem requiring a global solution. The article of Pablos- Mendez et al. (see pp. 489 - 494) provides a rational framework for finding such a solution.

The Philippines is ranked No.7 on the list of 22 high-burden countries for TB (1). DOTS was introduced at the public health centres only in 1992, and treatment success was 87%. (J. Lagahid on the DOTS strategy at the Department of Health, personal communication). MDR-TB is estimated to be present in 1.5% of new cases (2), although a precise assessment still has to be made. These rates would put the Philippines in the "f" category on the matrix of Pablos-Méndez et al. (see p. 492) for rationalizing the control of MDR-TB. Here MDR-TB treatment is not seen as a priority of the national TB control service, but to be allowed in specialized centres with appropriate laboratory support and help from nongovernmental organizations.

Only a third of the patients who seek medical care for TB in the Philippines are treated at a public health centre. The majority are treated by private practitioners (3). The Makati Medical Center, a tertiary referral private hospital, established a DOTS Clinic in 1999 in the spirit of private - public collaboration in TB control (4). Re-treatment cases in this clinic steadily increased, and now account for 44% of the patients enrolled. Treatment success in two cohorts analysed showed a decline from 85.3% in the first year to 68.2% in the second, with a corresponding increase in failure rates from 5.6% to 10.3%. All the failures were MDR-TB among the re-treatment cases. Treatment for these MDR-TB cases, in spite of logistical constraints, was called for for clinical, public health, and socioeconomic reasons.

Despite severely limited resources, outpatient treatment was made possible for 117 patients with MDR-TB through the DOTS-Plus pilot project. Support was provided by the national TB programme of the Department of Health, the Philippines Charity Sweepstakes, and the local government unit. Results so far have been encouraging, with a 75% preliminary estimate of cure and likely cure, and 9.1% failure or likely failure, and a default rate of 7.7%. Patients who are household heads and have responded to therapy have gone back to work and are now gainfully employed.

DOTS expansion in the country, which includes harnessing the private sector to the programme, remains the priority, as this should put a stop to the generation of more MDR-TB (6). However, with a large segment of the population very susceptible owing to malnutrition and crowding, treatment for MDR-TB should also be pursued, in a well-supervised fashion as practised at the Makati Medical Center's DOTS-Plus pilot project. An epidemiological study of the extent of MDR-TB in the country is urgently needed, to determine whether the Philippines should be regarded as a hot spot fuelling a global pandemic. If it is, funds for DOTS-Plus are urgently needed.

1. Dye C, Sckheele S, Dolin P, Pathania V, Raviglione MC. The WHO Global Surveillance and Monitoring Project. Global burden of tuberculosis: estimated incidence, prevalence and mortality by country. JAMA 1999;282:677-86.

2. Tupasi TE, Sistla R, Co VM, Villa MLA, Quelapio MID, Mangubat NV, et al. Bacillary disease and health-seeking behavior among Filipinos with symptoms of tuberculosis: implications for control. International Journal of Tuberculosis and Lung Diseases 2000;4:1126-32.

3. Rivera AB, Tupasi TE, Balagtas ED, Cardano RC, Baello BQ, Quelapio MID, et al. Drug resistant tuberculosis in the Philippines. The International Journal of Tuberculosis and Lung Disease 1999;3:639.

4. Quelapio MID, Mira NRC, Abeleda MR, Rivera AB, Tupasi TE. Directly observed therapy - short-course (DOTS) at the Makati Medical Center. Philippine Journal of Microbiology and Infectious Diseases 2000;29:80-6.

5. Schaaf H, Vermeulen H, Gie R, Beyers N, Donal P. Evaluation of young children in household contact with adult multidrug-resistant pulmonary tuberculosis cases. Pediatric Infectious Disease Journal 1999;18:494-500

6. Dye C, Williams BG, Espinal MA, Raviglione MC. Erasing the world's slow stain: strategies to beat multidrug-resistant tuberculosis. Science 2002;295: 2042- 6.

¹ Senior Scientist, Tropical Disease Foundation, Makati Medical Center, Manila, the Philippines (email: drcramos@info.com.ph).

Examining assumptions about multidrug-resistant TB control

Jim Yong Kim¹, Carole D. Mitnick¹, Jaime Bayona^{1, 2}, Ross Blank¹, Edward A. Nardell¹, Joia S. Mukherjee¹, Michael Rich¹, Paul Farmer^{1, 3}, Mercedes C. Becerra,¹ & Megan Murray⁴

Pablos-Méndez, Gowda, and Frieden aim to establish a "rational" base for the control of multidrug-resistant tuberculosis (MDR-TB) in resource-poor settings (see pp. 489 - 494). To assess the rationality of their framework, we must examine its core assumptions.

The first assumption - that strains of MDR-TB "are, on average, less infectious" - is not supported by the literature. The cluster studies referenced found that drug-susceptible TB was more likely to occur in clusters than drug-resistant TB. Nonetheless, other such studies have identified MDR-TB as a risk factor for clustering (1). Similarly, multiple observational reports describe the widespread dissemination of MDR clones from the W-strain family (2, 3). Other groups have also raised questions about the interpretability of cluster studies for inference (4 - 6). Finally, longitudinal epidemiological studies of TB transmission among household contacts have failed to support this finding (7, 8).

A second assumption - that DOTS "can reduce MDR-TB once it has occurred" - is supported only by examples of falling MDR-TB rates: no evidence has causally linked the use of DOTS alone to observed declines. In fact, the claim that DOTS can reduce rates of MDR-TB is not supported by the growing body of evidence that patients with MDR-TB have poor outcomes with short-course chemotherapy (9 - 14). New York City is cited as an example of successful control, yet this success was achieved through comprehensive interventions that included massive investments in infrastructure, infection control, improved case detection, and treatment of active cases of MDR-TB (2).

The third assumption - that "the need to introduce second-line therapy should be determined by countrywide statistics on treatment success rates and the proportion of all cases caused by MDR strains" - may result in dangerous policy decisions. The authors suggest an arbitrary threshold for their "rational" framework; however, they note that "specific cut-points for ... good clinical outcomes and high levels of MDR-TB have not been empirically validated." Because MDR-TB outbreaks are focal, countrywide averages may underestimate the seriousness of the problem (15). Additionally, focal outbreaks present an opportunity to develop a control strategy before the MDR-TB rates compromise the efficacy of DOTS programmes.

Lastly, it is assumed that to prioritize DOTS expansion "most national TB programmes do not need to introduce second-line anti-TB therapy". Poor outcomes of DOTS re-treatment regimens documented in patients who fail their initial round of short-course chemotherapy, however, raise an important ethical challenge to this assertion (14). As a result, WHO must now make a decision about whether or not to recommend second-line drugs as part of standard re-treatment regimens for all DOTS programmes.

DOTS expansion is the first priority in global TB control, but it is short-sighted to conclude that the latter can be achieved without effective strategies to treat and control MDR-TB. Rather than pit DOTS expansion against MDR-TB therapy, the task at hand is to obtain dramatically increased funds for comprehensive global TB control. With the new Global Fund to Fight AIDS, TB and Malaria (16), these resources may soon be forthcoming. We would argue that, now, to deny access to effective treatment for those already sick on the grounds that resources are scarce would be irrational or worse.

1. Alland D, Kalkut GE, Moss AR, McAdam RA, Hahn JA, Bosworth W, et al. Transmission of tuberculosis in New York City. An analysis by DNA fingerprinting and conventional epidemiologic methods. New England Journal of Medicine 1994;330:1710-6.

2. Frieden TR, Sherman LF, Maw KL, Fujiwara PI, Crawford JT, Nivin B, et al. A multi-institutional outbreak of highly drug-resistant tuberculosis: epidemiology and clinical outcomes. JAMA 1996;276:1229-35.

3. Bifani PJ, Mathema B, Kurepina NE and Kreiswirth BN. Global dissemination of the Mycobacterium tuberculosis W-Beijing family strains. Trends in Microbiology 2002;10:45-51.

4. Glynn JR, Vynnycky E, Fine PEM. Influence of sampling on estimates of clustering and recent transmission of Mycobacterium tuberculosis derived from DNA fingerprinting techniques. American Journal of Epidemiology 1999;149:366-71.

5. Murray M, Alland D. Methodological problems in the molecular epidemiology of tuberculosis. American Journal of Epidemiology 2002;155:565-71.

6. Murray M. Determinants of cluster distribution in the molecular epidemiology of tuberculosis. Proceedings of the National Academy of Sciences USA 2002;99:1538-43.

7. Snider DE Jr, Kelly GD, Cauthen GM, Thompson NJ, Kilburn JO. Infection and disease among contacts of tuberculosis cases with drug-resistant and drug- susceptible bacilli. American Review of Respiratory Disease 1985;132:125-32.

8. Teixeira L, Perkins MD, Johnson JL, Keller R, Palaci M, do Valle Dettoni V, et al. Infection and disease among household contacts of patients with multidrug- resistant tuberculosis. International Journal of Tuberculosis and Lung Disease 2001;5:321-8.

9. Mitchison DA, Nunn AJ. Influence of initial drug resistance on the response to short-course chemotherapy of pulmonary tuberculosis. American Review of Respiratory Disease 1986;133:423-30.

10. Manalo F, Tan F, Sbarbaro JA, Iseman MD. Community-based short-course treatment of pulmonary tuberculosis in a developing nation. Initial report of an eight-month, largely intermittent regimen in a population with a high prevalence of drug resistance. American Review of Respiratory Disease 1990;142:1301-5.

11. Kimerling ME, Kluge H, Vezhnina N, Iacovazzi T, Demeulenaere T, Portaels F, et al. Inadequacy of the current WHO re-treatment regimen in a central Siberian prison: treatment failure and MDR-TB. International Journal of Tuberculosis and Lung Disease 1999;3:451-3.

12. Centers for Disease Control and Prevention. Primary multidrug-resistant tuberculosis - Ivanovo Oblast, Russia, 1999. Morbidity and Mortality Weekly Report 1999;48:661-3.

13. Coninx R, Mathieu C, Debacker M, Mirzoev F, Ismaelov A, de Haller R, et al. First-line tuberculosis therapy and drug-resistant Mycobacterium tuberculosis in prisons. Lancet 1999;353:969-73.

14. Espinal MA, Kim SJ, Suarez PG, Kam KM, Khomenko AG, Migliori GB, et al. Standard short-course chemotherapy for drug resistant tuberculosis: treatment outcomes in six countries. JAMA 2000;283:2537-45.

15. Becerra MC, Bayona J, Freeman J, Farmer PE, Kim JY. Redefining MDR-TB transmission "hot spots". International Journal of Tuberculosis and Lung Disease 2000;4:387-94.

16. The Global Fund to Fight AIDS, Tuberculosis and Malaria. Available from: URL: http://www.globalfundatm.org

¹ Partners In Health and Program in Infectious Disease and Social Change, Department of Social Medicine, Harvard Medical School, 641 Huntington Avenue, Boston, MA 02115, USA. Correspondence should be addressed to Jim Yong Kim (email: jimkimpih@pih.org).

² Socios En Salud - Sucursal Perú, Lima, Peru.

³ Zanmi Lasante, Cange, Haiti.

⁴ Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA.

Sustainable TB control: the questions that have to be answered

Hans L. Rieder¹ & Catharina S.B. Lambregts-van Weezenbeek²

Multidrug-resistant tuberculosis (MDR-TB) is virtually never amenable to successful treatment with the six essential anti-TB medications. Patients who have it remain potential transmitters until spontaneous recovery or the more frequent fatal outcome. Several questions need thus to be addressed that are raised in the paper by Pablos-Méndez and collaborators in this issue of the Bulletin (see pp. 489 - 494)

The magnitude of the problem. A patient continuing to excrete sufficiently large numbers of bacilli at five months or later in treatment in an order of magnitude that they can be seen on a microscopic examination is defined as treatment failure. If the patient has been on a rifampicin-throughout regimen, it is thus tempting to assume that the strain is multidrug resistant. However, while there is a strong correlation between culture and microscopy results in patients on a regimen not containing rifampicin (1), the correlation is poor in patients who are on such a regimen (2). Furthermore, if the bacilli are indeed viable, failure is most often attributable to non-adherence rather than to drug resistance.

Technically demanding tests are required to demonstrate drug resistance (3). Pablos-Méndez and colleagues show that the failure to account for the operating characteristics of susceptibility testing often grossly overestimates the level of resistance. The published reports have not adjusted for this, and have identified "hot spots" which may not be ones (4, 5).

Conversely, the global map of resistance is very incomplete, and the information has usually been obtained from countries which have made efforts for national TB control, allowing representative sampling required for inclusion in the surveys. There is a critical need to expand coverage of the global surveillance system significantly and quickly.

Natural history. Untreated and untreatable sputum smear-positive TB has a very high case fatality (6 - 8). However, patients may survive and disseminate M. tuberculosis for years before succumbing. Drug-resistant organisms are therefore expected to have a comparative advantage. If the risk of secondary disease in the case of infection is the same, resistant organisms will ultimately gain the upper hand. Curiously, this has not always been the case. Pablos-Méndez and colleagues cite the example of New York City. When appropriate TB control measures were introduced for all patients, the number of multidrug-resistant cases dropped much faster than the number of fully drug-susceptible cases. Multidrug-resistant strains may thus actually have a disadvantage. This hypothesis fits with data from countries with a solid TB control programme such as Benin, where primary multidrug resistance remained barely measurable even after over a decade of rifampicin-containing chemotherapy (9). It is also congruent with experimental data that demonstrate inferior virulence of the subset of isoniazid-resistant strains whose resistance is attributable to katG gene deletion (10).

HIV infection may change the pattern of circulation of strains importantly. If virulence of Mycobacterium tuberculosis matters for non-compromised hosts, any selective advantage of less virulent strains becomes almost certainly irrelevant in the compromised host. On the other hand, the high death rate in HIV-infected TB patients may importantly curtail their ability to transmit. How this affects the epidemiology from drug resistance is far from clear.

Required public health action. Pablos-Méndez and colleagues propose a framework for how national programmes could prioritize their actions concerning DOTS and DOTS-Plus. Chemotherapy as recommended for DOTS programmes is highly cost-effective (11). An unbiased analysis of the ranking of a so-called DOTS-Plus strategy (12), providing more costly treatment to patients with MDR-TB has never been carried out.

The proposed framework considers essentially the prevalence of primary multidrug resistance, but not the total burden in a country. In some high-prevalence countries, such as South Africa, the proportion of re-treatment cases is considerable and related to a high burden of multidrug resistance, despite primary drug resistance rates that are still low. For this reason, we think that any framework should take into account the prevalence of combined resistance and an "acquired multidrug resistance index" in order to make a realistic estimate of the overall problem in the community (4).

In reality many countries already have second-line drugs at their disposal. It is fairly straightforward to develop a policy when these drugs are not available, but when they are avaiable they are used, and not always wisely. Therefore, we are in favour of a framework that works for both settings - with and without second-line drugs. Misuse of second-line drugs is a public health emergency in itself!

DOTS-Plus programmes must be sustainable if disaster is to be prevented. The WHO-coordinated Green Light Committee that reviews applications for preferentially priced second-line drugs plays a major role in ensuring access to high-quality drugs and at the same time ensures that these drugs are adequately used within the context of an integrated DOTS/DOTS-Plus strategy. In addition, it has a critical role in developing an evidence base to make more rational decisions on when and how to implement a comprehensive TB control policy that includes treatment of all patients, both those with susceptible and those with resistant
strains.

1. Mount FW, Ferebee SH. United States Public Health Service Cooperative Investigation of antimicrobial therapy of tuberculosis. V. Report of thirty-two-week observations on combinations of isoniazid, streptomycin, and para- aminosalicylic acid. American Review of Tuberculosis 1954;70:521-6.

2. Newman R, Doster B, Murray FJ, Woolpert SF. Rifampin in initial treatment of pulmonary tuberculosis. A U.S. Public Health Service tuberculosis therapy trial. American Review of Respiratory Diseases 1974;109:216-32.

3. Laszlo A, Rahman M, Raviglione M, Bustreo F. Quality assurance programme for drug susceptibility testing of Mycobcterium tuberculosis in the WHO/ IUATLD supranational laboratory network: first round of proficiency testing. International Journal of Tuberculosis and Lung Disease 1997;1:231-8.

4. Anti-tuberculosis drug resistance in the world. The WHO / IUATLD Global Project on Anti-Tuberculosis Drug Resistance Surveillance. Geneva: World Health Organization 1997. Unpublished document WHO/TB/ 97.229.

5. Anti-tuberculosis drug resistance in the world. Report No. 2. Geneva: World Health Organization; 2000. Unpublished document WHO/CDS/TB/ 2000.278.

6. Berg G. The prognosis of open pulmonary tuberculosis. A clinical-statistical analysis. Lund, Sweden: Håkan Ohlson; 1939.

7. Thompson BC. Survival rates in pulmonary tuberculosis. British Medical Journal 1943;2:721.

8. Buhl K, Nyboe J. Epidemiological basis of tuberculosis eradication. 9. Changes in mortality of Danish tuberculosis patients since 1925. Bulletin of the World Health Organization 1967;37:907-25.

9. Trébucq A, Anagonou S, Gninafon M, Lambregts K, Boulahbal F. Prevalence of primary and acquired resistance of Mycobacterium tuberculosis to antituberculosis drugs in Benin after 12 years of short-course chemotherapy. International Journal of Tuberculosis and Lung Disease 1999;3:466-70.

10. Li Z, Kelley C, Collins F, Rouse D, Morris S. Expression of katG in Mycobacterium tuberculosis is associated with its growth and persistence in mice and guinea pigs. Journal of Infectious Diseases 1998;177:1030-5.

11. Murray CJL, De Jonghe E, Chum HJ, Nyangulu DS, Salomao A, Styblo K. Cost effectiveness of chemotherapy for pulmonary tuberculosis in three sub- Saharan African countries. Lancet 1991;338:1305-8.

12. Guidelines for establishing DOTS-Plus projects for the managment of multidrug- resistant tuberculosis (MDR-TB). Geneva: World Health Organization; 2000. Unpublished document WHO document WHO/CDS/TB/2000.279:1-87.

¹ International Union Against Tuberculosis and Lung Disease, Jetzikofenstrasse 12, 3038 Kirchlindach, Switzerland (email: TBRieder@cs.com). Correspondence should be addressed to this author.

² Royal Netherlands Tuberculosis Association, The Hague, The Netherlands.