Contribution of medical colleges to tuberculosis control in India under the Revised National Tuberculosis Control Programme (RNTCP): lessons learnt & challenges ahead.

Medical college faculty, who are academicians are seldom directly involved in the implementation of national public health programmes. More than a decade ago for the first time in the global history of tuberculosis (TB) control, medical colleges of India were involved in the Revised National TB Control Programme (RNTCP) of Government of India (GOI). This report documents the unique and extraordinary course of events that led to the involvement of medical colleges in the RNTCP of GOI. It also reports the contributions made by the medical colleges to TB control in India. For more than a decade, medical colleges have been providing diagnostic services (Designated Microscopy Centres), treatment [Directly Observed Treatment (DOT) Centres] referral for treatment, recording and reporting data, carrying out advocacy for RNTCP and conducting operational research relevant to RNTCP. Medical colleges are contributing to diagnosis and treatment of human immunodeficiency virus (HIV)-TB co-infection and development of laboratory infrastructure for early diagnosis of multidrug-resistant and/or extensively drug-resistant TB (M/XDR-TB) and DOTS-Plus sites for treatment of MDR-TB cases. Overall, at a national level, medical colleges have contributed to 25 per cent of TB suspects referred for diagnosis; 23 per cent of 'new smear-positives' diagnosed; 7 per cent of DOT provision within medical college; and 86 per cent treatment success rate among new smear-positive patients. As the Programme widens its scope, future challenges include sustenance of this contribution and facilitating universal access to quality TB care; greater involvement in operational research relevant to the Programme needs; and better co-ordination mechanisms between district, state, zonal and national level to encourage their involvement.

Lessons from tuberculosis control for public health.

Tuberculosis (TB) control in many ways exemplifies evidence-based public health practice, rigorously implemented, with appropriate emphasis on the central importance of political support for success. With more than 30 million patients treated in the past decade, TB control has important implications for managing both communicable and non-communicable diseases. Simple diagnostic tests, meticulously proven standardized treatment regimens with assured drug supply, supportive case management and a superb information system that tracks the progress of every patient, all facilitate effective program implementation. TB control shows that public health programs, including those that require long-term treatment in the primary care system, can be effective in poor countries; however, TB rates are heavily influenced by the social, environmental and epidemiologic context, emphasizing that treatment is not enough and that socio-economic factors may be more important determinants of epidemiologic trends than treatment programs. TB control is effective when it combines two essential components: a practical, implementable, proven technical package, and political commitment. Political commitment is also essential to implement other interventions that can improve health, including healthier air, water and food, as well as programs to prevent or reduce tobacco smoking, cardiovascular disease, cancer, obesity and other growing public health problems. By implementing evidence-based practices, ensuring operational excellence, using information systems that facilitate accountability and evaluation, and obtaining and maintaining political support, we can address the public health challenges of the twenty-first century.

Rapid decline in prevalence of pulmonary tuberculosis after DOTS implementation in a rural area of South India.

SETTING: Tiruvallur District in Tamilnadu, South India, where the World Health Organization-recommended DOTS strategy was implemented as a tuberculosis (TB) control measure in 1999. OBJECTIVE: To assess the epidemiological impact of the DOTS strategy on the prevalence of pulmonary tuberculosis (PTB). DESIGN: Surveys of PTB were undertaken on representative population samples aged > or =15 years (n = 83000-90000), before and at 2.5 and 5 years after the implementation of the DOTS strategy. The prevalence of PTB (smear-positive/culture-positive) was estimated. RESULTS: TB prevalence declined by about 50% in 5 years, from 609 to 311 per 100000 population for culture-positive TB and from 326 to 169/100000 for smear-positive TB. The annual rate of decline was 12.6% (95%CI 11.2-14.0) for culture-positive TB and 12.3% (95%CI 8.6-15.8) for smear-positive TB. The decline was similar at all ages and for both sexes. CONCLUSION: With an efficient case detection programme and the DOTS strategy, it is feasible to bring about a substantial reduction in the burden of TB in the community.

Active community surveillance of the impact of different tuberculosis control measures, Tiruvallur, South India, 1968-2001.

BACKGROUND: Tuberculosis is curable, but community surveys documenting epidemiological impact of the WHO-recommended DOTS strategy on tuberculosis prevalence have not been published. We used active community surveillance to compare the impact of DOTS with earlier programmes. METHODS: We conducted tuberculosis disease surveys using random cluster sampling of a rural population in South India approximately every 2.5 years from 1968 to 1986, using radiography as a screening tool for sputum examination. In 1999, DOTS was implemented in the area. Prevalence surveys using radiography and symptom screening were conducted at the start of DOTS implementation and after 2.5 years. RESULTS: From 1968 to 1999, culture-positive and smear-positive tuberculosis declined by 2.3 and 2.5% per annum compared with 11.9 and 5.6% after DOTS implementation. The 2.5 year period of DOTS implementation accounted for one-fourth of the decline in prevalence of culture-positive tuberculosis over 33 years. Multivariate analysis showed that prevalence of culture-positive tuberculosis decreased substantially (10.0% per annum, 95% CI: 2.8-16.6%) owing to DOTS after only slight declines related to temporal trends (2.1% annual decline, 95% CI: 1.1-3.2%) and short-course chemotherapy (1.5% annual decline, 95% CI: -9.7% to 11.5%). Under DOTS, the proportion of total cases identified through clinical care increased from 81 to 92%. CONCLUSIONS: Following DOTS implementation, prevalence of culture-positive tuberculosis decreased rapidly following a gradual decline for the previous 30 years. In the absence of a large HIV epidemic and with relatively low levels of rifampicin resistance, DOTS was associated with rapid reduction of tuberculosis prevalence.

Value of dual testing for identifying tuberculous infection.

SETTING: A rural community in Chingleput district in Tamil Nadu state in south India. OBJECTIVE: To determine the value of dual testing with PPD-S and PPD-B for identifying subjects with a tuberculous infection. DESIGN: About 240,000 subjects in rural south India, all of whom were tested initially with PPD-S and PPD-B, were followed up for 15 years, mainly by total population survey once in every 212 years. The incidence of culture-positive tuberculosis was estimated using life-table technique. RESULTS: Among 17,530 subjects with an intermediate reaction (8-11 mm) to PPD-S at intake, 285 with an induration to PPD-S exceeding the induration to PPD-B by at least 2mm, had a significantly higher incidence of culture-positive tuberculosis than the remaining (154 and 93 per 100,000), and similarly 481 who had an induration of <10mm to PPD-B compared to those with >or=10 mm (131 and 93 per 100,000). These subjects may be regarded as having a tuberculous infection. Infection with non-tuberculous mycobacteria conferred protection of about 30% against the development of tuberculosis over a 15-year period. CONCLUSION: In subjects with an intermediate reaction (8-11 mm) to PPD-S, dual testing with PPD-B enabled identification of those with a tuberculous infection. Most of the reactions were due to non-tuberculous mycobacteria.

Ensuring accountability: the contribution of the cohort review method to tuberculosis control in New York City.

SETTING: In 1993, the New York City (NYC) Bureau of Tuberculosis Control developed the cohort review process as a quality assurance method to track and improve patient outcomes. METHODS: The Bureau Director reviews every tuberculosis (TB) case quarterly in a multi-disciplinary staff meeting. In 2004 we also began collecting details on issues identified at cohort review to quantify how this process directly impacts TB control efforts. RESULTS: From 1992 to 2004, NYC TB cases decreased by 72.7% and treatment success rates significantly increased by 26.7%. Implementing the cohort review was key to improving case management, thus leading to these results. For the 1039 patients in 2004, 596 issues were identified among 424 patients; 55.0% were incorrect, unclear or unknown patient information, 13.8% were treatment issues, 12.4% were case management issues and 10.6% were incomplete contact investigations. Most (76.5%) issues were addressed within 30 days of the cohort reviews. CONCLUSION: A systematic review of every TB case improves the quality of patient information, enhances patient treatment and ensures accountability at all levels of the TB control program.

Improved tuberculosis case detection through public-private partnership and laboratory-based surveillance, Kannur District, Kerala, India, 2001-2002.

BACKGROUND: Efforts to intensify global tuberculosis (TB) control are limited by difficulties in coordinating with private doctors. More than half of Indian TB patients may initially consult a private provider, but many are neither diagnosed accurately nor treated effectively. We established and evaluated a public-private partnership based on surveillance of TB detected in private laboratories and use of standardised directly observed treatment regimens. METHODS: In one district, the governmental TB control programme offered training in microscopy to all large private sector laboratories, and educated private physicians on the importance of microscopy for TB diagnosis. We reviewed records from participating private laboratories and all publicly diagnosed patients. RESULTS: Of 2328 pulmonary TB patients registered from July 2001 to December 2002, 404 (17%) were detected in the private sector. The annual new AFB-positive case notification rate increased by 21%, from 27.8/100,000 in 2000 to 33.5/100,000 in 2002. Surveillance at private laboratories found an additional 260 nonregistered AFB-positive patients. CONCLUSIONS: This public-private partnership substantially increased TB case detection and established a sustainable framework for private sector involvement in TB control. In the setting of a strong public sector programme, the combination of active surveillance of private laboratories along with physician sensitisation is a promising approach to improve TB case detection.

Comparison of cough of 2 and 3 weeks to improve detection of smear-positive tuberculosis cases among out-patients in India.

SETTING: Governmental health facilities in six districts of India. OBJECTIVE: To estimate the prevalence of cough and to compare the detection of smear-positive tuberculosis (TB) among out-patients with cough of > or =2 or > or =3 weeks. DESIGN: Trained health workers questioned each out-patient for presence of cough. Those with cough > or =2 weeks underwent sputum microscopy. RESULTS: Of 55561 out-patients interviewed, 2210 (4%) had cough > or =2 weeks, of whom 267 had sputum-positive TB, compared to 182/1370 with cough > or =3 weeks. The 31% who did not spontaneously complain of cough were less likely to be sputum-positive than those who did (45/680 [7%] vs. 222/1530 [15%], P < 0.001), but they accounted for 45/267 smear-positive cases. Using cough > or =2 weeks as the screening criterion, the estimated number of smears performed per day at each primary and secondary health care facility was respectively 8 and 19, compared to 5 and 12 using cough > or =3 weeks. CONCLUSION: The detection of smear-positive TB cases can be substantially improved by actively eliciting history of cough from all out-patients, and by changing the screening criterion for performing sputum microscopy among out-patients from cough > or =3 weeks to > or =2 weeks. Before implementing this change nationally, its programmatic feasibility should be assessed.

Lung abscess with group A beta-hemolytic Streptococcus. Case report and review.

Invasive infections with group A beta-hemolytic streptococci became less common in the early 20th century prior to the widespread use of antibiotics. From the early 1960s until the mid-1980s, reports of invasive infections continued to decline. In the past 5 years, there has been a resurgence of invasive infections and, possibly, also of postinfectious sequelae from this organism. We describe a patient with lung abscess from group A beta-hemolytic Streptococcus. Lung abscess from hemolytic streptococci was not uncommon in Osler's day, but it was not reported in the English-language literature for 20 years until recently. Clinicians should be aware of the broad and growing spectrum of infections with this pathogen.

Typhoid fever in New York City, 1980 through 1990.

BACKGROUND: Although typhoid fever incidence decreased in the 1960s and 1970s in New York City and elsewhere, it did not disappear. In this article, trends associated with various modes of transmission of Salmonella typhi in New York City patients are described. METHODS: Typhoid fever surveillance reports from 1980 to 1990 were reviewed for clinical, demographic, and epidemiologic characteristics. Cases of typhoid fever were classified as travel related or domestically acquired. RESULTS: In all, 479 typhoid cases were identified, of which 67% were travel related. The age groups most frequently affected were children and adolescents. Cases more than doubled in the decade, and the ratio of travel-related cases to domestically acquired cases increased steadily from 63% to 80%. Travelers to Southeast Asia were at three times higher risk than those visiting South America and eight times higher than those visiting the Caribbean. The case-fatality proportion was 1.5%. CONCLUSION: The trends of S typhi infection in New York City followed the trends observed in the United States since 1978, which demonstrates the importance of international travel. Although food and water precautions may be effective for short-term travelers, selective use of oral antityphoid vaccines for New York City travelers to high-risk endemic countries should be encouraged.

Isoniazid preventive therapy in areas of high isoniazid resistance.

BACKGROUND: Previous decision analyses of isoniazid preventive therapy for low-risk tuberculin reactors aged 20 to 34 years have not accounted for the recently increased isoniazid resistance rate. Drug resistance trends could also affect the decision to use isoniazid preventive therapy for patients with recent conversion of tuberculin skin tests who are seronegative for human immunodeficiency virus. METHODS: A decision analysis was performed with a Markov simulation to assess the difference in life expectancy between those who receive isoniazid preventive therapy and those who do not. Probability estimates were determined from a review of the literature. RESULTS: For tuberculin reactors aged 20 to 34 years living in areas with 26% isoniazid resistance, isoniazid preventive therapy increases life expectancy by 2 days. Withholding isoniazid is clearly favored if the isoniazid hepatitis rate is 1.1% and the hepatitis fatality rate exceeds 2.8%. For recent tuberculin converters, isoniazid preventive therapy increases life expectancy by 14 to 17 days, depending on patient age. Withholding isoniazid from converters is favored only if the isoniazid resistance rate exceeds 90% to 98%, according to patient age. Two-way sensitivity analysis of isoniazid-associated hepatitis and hepatitis-related fatality rate did not affect the decision to use isoniazid for recent converters. CONCLUSIONS: For tuberculin reactors aged 20 to 34 years who are seronegative for human immunodeficiency virus and living in areas with high isoniazid resistance, there is minimal net benefit of isoniazid preventive therapy. The current recommendation to provide isoniazid preventive therapy to this patient population should be reexamined. For recent tuberculin converters aged 20 to 64 years who are seronegative for human immunodeficiency virus, isoniazid preventive therapy provides a small increase in life expectancy. Withholding isoniazid preventive therapy for human immunodeficiency virus-seronegative skin test converters at high risk for isoniazid-induced hepatitis may be considered; preventive therapy is advisable for all other recent converters.

A continuing survey of drug-resistant tuberculosis, New York City, April 1994.

BACKGROUND: A 1991 survey showed high levels of drug resistance among tuberculosis patients in New York, NY. As a result, the tuberculosis control program was strengthened, including expanded use of directly observed therapy and improved infection control. METHODS: We collected isolates from every patient in New York City with a positive culture for Mycobacterium tuberculosis during April 1994; results were compared with those in the April 1991 survey. RESULTS: From 1991 to 1994, the number of patients decreased from 466 to 332 patients. The percentage with isolates resistant to 1 or more antituberculosis drugs decreased from 33% to 24% (P < .01); with isolates resistant to at least isoniazid decreased from 26% to 18% (P < .05); and with isolates resistant to both isoniazid and rifampin decreased from 19% to 13% (P < .05). The number of patients with isolates resistant to both isoniazid and rifampin decreased by more than 50%. Among never previously treated patients, the percentage with resistance to 1 or more drugs decreased from 22% in 1991 to 13% in 1994 (P < .05). The number of patients with consistently positive culture results for more than 4 months decreased from 130 to 44. A history of antituberculosis treatment was the strongest predictor of drug resistance (odds ratio = 3.1; P < .001). Human immunodeficiency virus infection was associated with drug resistance among patients who never had been treated for tuberculosis. CONCLUSIONS: Drug-resistant tuberculosis declined significantly in New York City from 1991 to 1994. Measures to control and prevent tuberculosis were associated with a 29% decrease in the proportion of drug resistance and a 52% decrease in the number of patients with multidrug-resistant tuberculosis.

Purified protein derivative testing and tuberculosis preventive therapy for HIV-infected patients in New York City.

OBJECTIVE: To determine whether Centers for Disease Control and Prevention recommendations for purified protein derivative (PPD) testing and tuberculosis (TB) preventive therapy for PPD-positive patients are implemented in HIV clinics. DESIGN: Retrospective medical chart review. SETTING: Ten hospital-based HIV clinics in New York City. PARTICIPANTS: A total of 2397 patients with a first clinic visit in 1995. OUTCOME MEASURES: PPD testing of eligible patients, and recommendation of preventive therapy and completion of regimen in PPD-positive patients. METHOD: Outpatient medical records were abstracted for TB history, PPD testing, TB preventive therapy, and patient demographic, social and clinical characteristics. Multivariate analyses were performed using logistic regression. RESULTS: Of 1342 patients with an indication for a PPD test, 865 (64%) were PPD tested in the clinic and 757 (88%) returned to have it read. Factors strongly associated with PPD testing in the clinic were number of visits, same sex behavior with men, and CD4+ lymphocyte count above 200 x 10(6)/l. Preventive therapy was recommended for 80% of newly identified PPD-positive patients and 22% of previously identified PPD-positive patients. Of 119 patients on preventive therapy in the clinic, 49 (41%) completed the regimen, 50 (42%) were lost to follow-up, and 20 (17%) discontinued therapy or their status could not be determined. CONCLUSION: A significant number of missed opportunities to implement TB prevention practices were identified in HIV clinics. Focused attention in HIV clinics, and increased collaboration between HIV clinics and TB control programs may be needed to increase adherence to prevention guidelines.

Nonadherence in tuberculosis treatment: predictors and consequences in New York City.

BACKGROUND: Poor adherence to antituberculosis treatment is the most important obstacle to tuberculosis control. PURPOSE: To identify and analyze predictors and consequences of nonadherence to antituberculosis treatment. PATIENTS AND METHODS: Retrospective study of a citywide cohort of 184 patients with tuberculosis in New York City, newly diagnosed by culture in April 1991-before the strengthening of its control program-and followed up through 1994. Follow-up information was collected through the New York City tuberculosis registry. Nonadherence was defined as treatment default for at least 2 months. RESULTS: Eighty-eight of the 184 (48%) patients were nonadherent. Greater nonadherence was noted among blacks (unadjusted relative risk [RR] 3.0, 95% confidence interval [CI] 1.1 to 8.6, compared with whites), injection drug users (RR 1.5, 95% CI 1.1 to 2.0), homeless (RR 1.4, 95% CI 1.0 to 1.8), alcoholics (RR 1.4, 95% CI 1.0 to 1.9), and HIV-infected patients (RR 1.4, 95% CI 1.1 to 1.9); also, census-derived estimates of household income were lower among nonadherent patients (P = 0.018). In multivariate analysis, only injection drug use and homelessness predicted nonadherence, yet 46 (39%) of 117 patients who were neither homeless nor drug users were nonadherent. Nonadherent patients took longer to convert to negative culture (254 versus 64 days, P < 0.001), were more likely to acquire drug resistance (RR 5.6, 95% CI 0.7 to 44.2), required longer treatment regimens (560 versus 324 days, P < 0.0001), and were less likely to complete treatment (RR 0.5, 95% CI 0.4 to 0.7). There was no association between treatment adherence and all-cause mortality. CONCLUSIONS: In the absence of public health intervention, half the patients defaulted treatment for 2 months or longer. Although common among the homeless and injection drug users, the problem occurred frequently and unpredictably in other patients. Nonadherence may contribute to the spread of tuberculosis and the emergence of drug resistance, and may increase the cost of treatment. These data lend support to directly observed therapy in tuberculosis.

Tuberculosis in children younger than five years old: New York City.

We examined medical and health department records for children < 5 years of age with suspected or confirmed tuberculosis reported to the New York City Health Department from January, 1992, through June, 1992, in order to describe the epidemiology of tuberculosis in young children and identify prevention strategies. Forty-seven children were treated for suspected or confirmed tuberculosis. Sixty-two percent (21 of 34) were foreign-born (n = 11) or had foreign-born caretakers (n = 10). A source case was found for 10 of 47 (21%) children; for 8 the adult source was diagnosed before the child. One child was human immunodeficiency virus-seropositive, however, 83% of children and 70% of adult source cases did not have human immunodeficiency virus test results available. Health care providers should test children at high risk for tuberculosis infection as recommended by the American Academy of Pediatrics and improve contact tracing to identify children exposed to adults with tuberculosis. Because most cases of tuberculosis in children are diagnosed clinically rather than by isolating Mycobacterium tuberculosis, identification of the source case is important for selecting appropriate treatment.

The status and prospects of tuberculosis control in India.

SETTING: India, where much of the global strategy for tuberculosis control was established, but where, every year, there are an estimated 2 million cases of tuberculosis. OBJECTIVE: To describe the policies, initial results, and lessons learned from implementation of a Revised National Tuberculosis Control Programme using the principles of DOTS (Directly Observed Treatment, Short-course). DESIGN: A Revised National Tuberculosis Control Programme (RNTCP) was designed and implemented starting in 1993. With funding from the Government of India, State Governments, the World Bank and bilateral donors, regular supply of drugs and logistics was ensured. Persons with chest symptoms who attend health facilities are referred to microscopy centres for diagnosis. Diagnosed cases are categorized as per World Health Organization guidelines, and treatment is given by direct observation. Systematic recording and cohort reporting is done. RESULTS: From October 1993 through mid-1999, 146,012 patients were put on treatment in the programme. The quality of diagnosis was improved, with the ratio of smear-positive to smear-negative patients being maintained at 1:1. Case detection rates varied greatly between project sites and correlated with the percentage of patients who were smear-positive among those examined for diagnosis, suggesting heterogeneous disease rates. Treatment success was achieved in 81% of new smear-positive patients, 82% of new smear-negative patients, 89% of patients with extra-pulmonary tuberculosis, and 70% of re-treatment patients. CONCLUSION: The RNTCP has successfully treated approximately 80% of patients in 20 districts of 15 states of India. Treatment success rates are more than double and death rates are less than a seventh those of the previous programme. Starting in late 1998, the programme began to scale up and now covers more than 130 million people. Maintaining the quality of implementation during the expansion phase is the next challenge.

Impact of national consultants on successful expansion of effective tuberculosis control in India.

SETTING: India, during a period of rapid expansion of DOTS services. DOTS expansion has been slow in many countries. OBJECTIVE: To document use of consultants to expand DOTS effectively. DESIGN: Staff were contracted to monitor DOTS expansion and implementation. To estimate the impact of these staff, we compared areas with and without consultants, and individual areas before and after consultants were assigned. Consultants were preferentially assigned to the more difficult areas; the temporary absence of consultants reflected non-availability of candidates. RESULTS: Areas with consultants met pre-defined criteria and began DOTS service delivery faster (median 9 vs. 18 months of preparation) than areas without consultants. Rates of sputum conversion (87% vs. 83%, P < 0.001) and treatment success (83% vs. 78%, P < 0.001) were significantly higher in areas with consultants present. CONCLUSION: Assignment of consultants resulted in much more rapid implementation of the DOTS strategy, and better quality performance. Continued effective performance in these areas will rely on many factors, but the need for consultants appears to be decreasing, suggesting that they have provided sustainable improvements. The effectiveness of local consultants may have important implications for efforts to scale up public health interventions for tuberculosis, malaria, AIDS and other diseases in developing countries.

Washing of new microscopic glass slides in dichromate solution does not influence sputum AFB smear results.

SETTING: Microscopy centres in Tiruvallur District, Tamil Nadu, India, implementing DOTS. OBJECTIVE: To know whether washing new glass slides in dichromate solution is essential for effective sputum acid-fast bacilli microscopy. METHODS: Two direct smears were prepared from each of 1750 sputum samples. One was made on dichromate solution-cleaned new glass slides and the other was made on unwashed new glass slides. The smears were blinded and examined. RESULTS: Of the 1750 specimens, 13.5% and 13.08% were positive for AFB using washed and unwashed slides, respectively (P = 0.12). The concordance between these two (including one grade above and one below) was 98.7%. CONCLUSION: Washing of new glass slides in dichromate solution is not essential for AFB microscopy.

A survey of tuberculosis hospitals in India.

SETTING: Hospitals with beds for tuberculosis (TB) in India. OBJECTIVES: To assess diagnostic and treatment practices at institutions offering secondary or tertiary level care for TB patients, and to determine the resources being used at these institutions. DESIGN: Countrywide cross-sectional survey of TB hospitals using a mailed semi-structured questionnaire sent to all 105 hospitals with 100 or more beds and to all State Directorate Health Services. RESULTS: The 94 hospitals that returned the questionnaire had 15773 TB beds, one third of the total TB beds in the country. Nearly 1 million patients sought treatment in the TB hospitals and one third were diagnosed with TB; the ratio of smear-positive to smear-negative patients was 1:2.7. Sixty-four per cent of hospitals prescribed unobserved rifampicin in the continuation phase, and 56% of sputum smear-positive patients were hospitalised. The annual expenditure for the TB hospitals was more than the total annual budget for the TB control programme of the country. CONCLUSIONS: In view of the high number of patients seen and the suboptimal practices observed, urgent steps should be taken to ensure implementation of correct diagnostic and treatment policies in hospitals with TB beds.

Rates and risk factors for discontinuation of rifampicin.

SETTING: All patients with culture-confirmed, rifampin-susceptible Mycobacterium tuberculosis diagnosed during a 20-month period in New York City, who were started on a rifampin-containing regimen and received > or =60 days of treatment. OBJECTIVE: To identify rates of and reasons for rifampin discontinuation. DESIGN: Retrospective case-control study using surveillance data and medical record reviews. Discontinuation due to thrombocytopenia, creatinine >2.0 mg/dl, bilirubin >2.0 mg/dl or severe reactions (generalized rash, persistent drug fever, or severe interference with methadone metabolism) were defined as appropriate for discontinuation of rifampin. All other reactions were classified as inappropriate. RESULTS: Of 3,520 patients, rifampin was discontinued in 68 (1.9%); of these, 57% had rifampin discontinued unnecessarily. Treatment by an inexperienced provider (adjusted odds ratio [ORadj] 4.0; 95% confidence interval [CI] 1.9-8.5), race (ORadj 3.1; 95%CI 1.4-6.9), history of previous treatment (ORadj 4.8; 95%CI 1.9-12.5), and history of methadone drug treatment (ORadj 12.6; 95%CI 5.3-29.9) were all associated with inappropriate rifampin discontinuation. CONCLUSION: True intolerance was rare, even among those patients infected with the human immunodeficiency virus. Most patients with minor reactions can successfully complete treatment with rifampin, particularly if managed by a physician experienced in the treatment of tuberculosis.