Published ahead of print on June 7, 2004, doi:10.1164/rccm.200401-095OC
© 2004 American Thoracic Society doi: 10.1164/rccm.200401-095OC
Tuberculosis Treatment OutcomesDirectly Observed Therapy Compared with Self-Administered TherapyDivision of Pulmonary and Critical Care Medicine, San Francisco General Hospital; Department of Medicine, and Department of Epidemiology and Biostatistics, University of California, San Francisco; Francis J. Curry National Tuberculosis Center; and Tuberculosis Control Section, San Francisco Department of Public Health, San Francisco, California Correspondence and requests should be addressed to Robert M. Jasmer, M.D., Division of Pulmonary and Critical Care Medicine, San Francisco General Hospital, Room 5K-1, 1001 Potrero Avenue, San Francisco, CA 94110. E-mail: rjasmer{at}itsa.ucsf.edu
Effective treatment of tuberculosis requires adherence to a minimum of 6 months treatment with multiple drugs. To improve adherence and cure rates, directly observed therapy is recommended for the treatment of pulmonary tuberculosis. We compared treatment outcomes among all culture-positive patients treated for active pulmonary tuberculosis (n = 372) in San Francisco County, California from 1998 through 2000. Patients treated by directly observed therapy at the start of therapy (n = 149) had a significantly higher cure rate compared with patients treated by self-administered therapy (n = 223) (the sum of bacteriologic cure and completion of treatment, 97.8% versus 88.6%, p < 0.002), and decreased tuberculosis-related mortality (0% vs. 5.5%, p = 0.002). Rates of treatment failure, relapse, and acquired drug resistance were similar between the two groups. Forty-four percent of patients who received self-administered therapy had risk factors for nonadherence and should have been assigned to directly observed therapy. We conclude that treatment plans that emphasize directly observed therapy from the start of therapy have the greatest success in improving tuberculosis treatment outcomes.
Key Words: directly observed therapy • epidemiology • pulmonary tuberculosis The main goals of tuberculosis treatment are to cure the individual with the disease and minimize the transmission of Mycobacterium tuberculosis to others in the community. Treatment can be challenging for both patients and providers because it requires taking multiple medications for a minimum of 6 months (1). The consequences of nonadherence to treatment include increased rates of treatment failure, relapse, acquired drug resistance, and prolonged infectiousness of patients (2–3). To improve adherence to treatment, directly observed therapy (DOT), in which patients are observed ingesting each dose of antituberculosis medications, is recommended as the standard of care for treatment of pulmonary tuberculosis by several leading international organizations (4–5). DOT is the cornerstone of a patient-centered approach to treatment to maximize the likelihood of completion of therapy (6). Most studies evaluating DOT have been observational (7–9), and some of the randomized trials (10–13) have lacked an appropriate comparison group. These studies, summarized in a consensus statement led by Chaulk and Kazandjian (14), showed an average of 86% completion for patients treated by DOT. In San Francisco, we have been using DOT in a substantial number of patients since 1979, particularly in patients with drug-resistant disease, HIV infection, substance abuse, psychiatric illness, children and teenagers, and others felt by clinicians to be at risk for treatment nonadherence. In this study, we assessed rates of completion, cure, failure, relapse, acquired drug resistance, and death due to tuberculosis among patients with tuberculosis reported in San Francisco over a 3-year period in an effort to determine any differences between patients treated by DOT compared with those treated by self-administered therapy. Some of the results of this study have been previously reported in abstract form (15).
Study Patients In this study, we included all patients with culture-positive tuberculosis who were reported to the San Francisco Tuberculosis Control Section of the Department of Public Health from January 1, 1998 through December 31, 2000. We excluded patients with extrapulmonary tuberculosis and those with culture-negative tuberculosis because we wanted to focus on patients in whom follow-up cultures would be readily available and in whom our outcomes could be readily assessed. DOT was defined as ingestion of antituberculosis medications that was directly supervised by a healthcare worker. DOT took place at either the tuberculosis clinic, or the patient's residence or workplace. Patients whose treatment began in the hospital were classified as receiving DOT. Patients treated by DOT received enhanced DOT in which enablers and incentives (e.g., travel vouchers, food, housing assistance) were used to promote adherence. Self-administered therapy (SAT) was defined as unsupervised administration of antituberculosis medications by patients as prescribed by their provider. DOT is recommended as the optimal treatment strategy by the San Francisco Tuberculosis Control Section for patients who have positive sputum smears, have a history of homelessness, injection or noninjection drug use, HIV infection, alcohol abuse, incarceration, are less than 21 years old, or are too infirm to self-manage therapy. If a patient had any of these characteristics and was initially treated by SAT, the assignment was deemed inappropriate. Patients treated by DOT received treatment daily for the first 2 months, and then, for the continuation phase, a case management team determined daily or twice-weekly DOT or SAT. Doses on weekends were self-administered. Medications taken by patients treated by SAT are paid for by the Department of Public Health (or by patients' insurers) and dispensed at the Tuberculosis Clinic or as prescribed by private providers. Patients treated by SAT are monitored at least once per month for adherence, drug toxicity, symptoms, and sputum conversion. Private patients who have DOT criteria are strongly encouraged to be treated through the Department of Public Health. Charts were reviewed for the following data: dates of treatment, whether treatment was initiated with DOT or SAT and any changes during treatment, reason for treatment by DOT, age, sex, country of birth, smear and culture results, presence of pulmonary or extrapulmonary disease, and presence of cavitation on a chest radiograph. The Institutional Review Board of the University of California, San Francisco approved the study protocol.
Outcomes
Acquired drug resistance.
Bacteriologic cure.
Completion of treatment.
Cure rate.
Death due to other causes.
Death due to tuberculosis.
Lost to follow-up.
Lost to post-treatment evaluation.
Moved.
Post-treatment evaluation.
Relapse.
Treatment failure.
Statistical Analysis
Study Population In the 3-year period from 1998–2000, 612 incident cases of tuberculosis were reported in San Francisco. Of these, 240 were excluded from the study population because of negative cultures (n = 110), absence of pulmonary disease (n = 76), initial resistance to isoniazid, rifampin, or pyrazinamide (n = 46), or lack of any treatment (n = 8). The remaining 372 patients comprised the study population (Figure 1).
Demographic and Clinical Characteristics Overall, 372 patients were treated for culture-positive pulmonary tuberculosis in San Francisco in 1998 through 2000, of whom 149 (40.0%) were treated by DOT at the start of treatment. As shown in Table 1, patients treated by DOT from the beginning of therapy were significantly more likely than patients treated with SAT to be male, younger by an average of 10 years, more likely to be born in the United States, and less likely to have a private medical provider. Patients treated by DOT were on DOT for a median of 28.4 weeks, due to missed doses and/or interruptions for drug toxicity.
Patients treated with DOT were significantly more likely than patients treated with SAT to have positive sputum smears (57% of patients treated with DOT), be homeless, have either injection or noninjection drug use, be infected with HIV, have alcohol abuse, be less than 21 years old, be in a correctional facility, or be too infirm to self-manage therapy (Table 1). These differences are a consequence of the criteria used to assign patients to DOT. The majority (71%) of patients who were started on DOT were treated by DOT throughout. Among those who were changed from DOT to SAT, the median time for the change was 85 days. Eighty-nine percent of those who were started on SAT received SAT throughout. Among those who were changed from SAT to DOT, the median time for the change was 61 days.
Outcomes of Treatment between Those Treated by DOT compared with Those Treated by SAT
Treatment Outcomes among HIV-infected Patients Among the 50 patients with HIV infection and culture-positive pulmonary tuberculosis during the study period, the rate of treatment cure was not significantly higher among those treated by DOT at the start of therapy compared with those treated by SAT at the start of therapy (93.8% vs. 77.8%, p = 0.20). Treatment failure or relapse occurred in none of the 37 HIV-infected patients treated by DOT at start but in 1 of 13 treated by SAT (p = 0.26). Three HIV-infected patients (all treated by SAT at start of therapy) died during treatment, but 2 died within 3 days of the start of therapy and were excluded (p = 0.23).
Misassignment of Patients to SAT Because we are testing the efficacy of DOT when used in patients at risk for nonadherence, we divided the patients who received SAT at the start of therapy into two groups: those with risk factors for nonadherence (n = 97) and those without established risk factors (n = 126). When we compared patients who received DOT to patients with risk factors for nonadherence who received SAT (i.e., inappropriate SAT), the difference in outcomes was even stronger: treatment cure occurred in 97.8% of patients treated by DOT compared with only 80.3% of patients treated by SAT (p < 0.0001, relative risk [RR] = 1.22, 95% confidence interval [CI] = 1.09–1.37); death due to tuberculosis occurred in 0 patients treated by DOT compared with 10.6% of patients treated by SAT (p < 0.0001, RR = 0.03, 95% CI = 0.002–0.52) (Table 4). In contrast, when we compared the patients treated by DOT with patients treated by SAT not at risk for nonadherence (i.e., appropriate SAT) there were no significant differences in the major outcomes (Table 5).
This study shows that patients with tuberculosis treated by DOT at the start of therapy had significantly higher cure rates compared with those treated by SAT at the start. Patients treated by DOT at start also had lower rates of tuberculosis-related mortality. However, because the majority of deaths occurred during the first 2 weeks of treatment, DOT may not have been the major factor that contributed to this decreased mortality. No significant difference was found in the rate of acquired drug resistance between the two groups of patients, likely because only 2 of 372 patients (both treated by SAT) had acquired drug resistance. The study population included all culture-positive patients in San Francisco treated for initially drug-susceptible pulmonary tuberculosis in 1998 through 2000. During those years, San Francisco County had the highest tuberculosis incidence rate of all counties in the state of California (16). San Francisco has a long history of using DOT to improve adherence to tuberculosis treatment, beginning in the early 1980s when case rates abruptly increased. Patients assigned to DOT are treated by so-called enhanced DOT in which adherence-promoting measures are given, including, for example, social service support, treatment incentives and enablers, housing assistance, and referral for treatment of substance abuse. Patients treated by DOT have been those thought to be at highest risk for treatment nonadherence and failure or at greatest risk for transmission of infection. In our study population, the patients treated by DOT had higher rates of HIV infection (p < 0.001) and drug-resistant tuberculosis (we excluded these patients), factors known to be associated with poor response to treatment (17–20). In addition, our patients treated by DOT had high rates of homelessness, substance abuse with alcohol or other drugs, and history of being in a correctional facility, all factors thought to compromise treatment adherence. Despite these differences, we observed improvement in cure rates and in the rate of treatment-related mortality among those patients treated by DOT at the start of treatment compared with SAT. Our findings strongly suggest that DOT is the preferred management strategy to achieve the highest possible cure rate and lowest rate of death due to tuberculosis. A remarkable finding from our study was that patients treated by DOT had significantly fewer tuberculosis-related deaths compared with those treated by SAT. This is especially notable given the higher rates of sputum smear–positive disease, alcohol and drug abuse, and HIV infection in the DOT group, all factors thought to increase tuberculosis mortality (5). In addition, just under one-half of patients who received SAT as initial therapy should have been assigned to DOT, and controlling for these risk factors among patients who received SAT eliminated any differences in tuberculosis-related death between the two groups. This suggests that, among patients without risk factors for nonadherence (i.e., appropriate use of SAT), SAT results in outcomes similar to those who receive DOT. Misassignment of patients to SAT was a particular problem among private providers, who mistakenly assigned almost half of their patients to SAT. This suggests that educational programs should be undertaken to educate private providers concerning appropriate patients for SAT, and that all patients with newly diagnosed tuberculosis should be evaluated by the public health department for possible assignment to DOT. Specifically, patients who cannot bear any of the risks of nonadherence should be initiated on treatment by DOT to have the greatest impact on public health. Almost all prior studies of DOT have been observational and have lacked an appropriate comparison group that has therefore limited the ability to draw firm conclusions regarding the superiority of DOT compared with SAT. There have been four recent randomized controlled trials (10–13) that have assessed the direct observation component of DOT, known as DOT short course (DOTS), a strategy for tuberculosis care recommended by the World Health Organization and International Union Against Tuberculosis and Lung Disease (4–5). This strategy comprises five elements, of which direct observation of treatment by healthcare workers during at least the first 2 months is one component. The first such study, in South Africa, involved 216 patients and found that treatment of tuberculosis had significantly higher completion and cure rates among self-supervised patients than among those on direct observation of treatment (10). A study of 836 patients in Thailand showed a significantly higher cure rate among those treated by DOT although there was no difference in the proportion that had negative sputum cultures after 2 months of treatment (12). Another study of 497 patients in Pakistan showed similar completion and cure rates among patients randomized to direct observation of treatment by a healthcare worker, direct observation of treatment by a family member, or self-administered treatment (11). There are significant limitations to these studies, however, most notably that the rates of cure and completion reported with either direct observation or self-administration of treatment were in the 50–60% range, substantially (35–40%) less than those for patients treated by either method in our study. Thus, findings from those studies likely do not apply to areas with low incidence of tuberculosis such as San Francisco, in which we found that DOT significantly improves upon treatment outcomes in an already highly successful tuberculosis treatment program. There are several limitations to our study. First, we did not randomly assign patients to DOT or SAT; therefore, it is possible that patients who were selected for DOT were more likely to benefit from this intervention than those selected for SAT. Given the higher rates of homelessness, substance abuse, drug resistance, HIV infection, and sputum smear positivity among patients assigned to DOT, the patients treated by DOT represented an extremely difficult group of patients to treat, many of whom had risk factors for treatment nonadherence or failure. Thus, one would have expected the patients treated by SAT to have better outcomes than the patients treated by DOT, given the higher prevalence of these factors. However, we detected a significant improvement in treatment cure among the patients treated by DOT. Second, many of the patients who died were elderly and had significant comorbidities that may have contributed to their deaths. Although none of the patients who died were initially hospitalized, hospitalization and its attendant closer monitoring may have averted death in some of these patients. Finally, not all patients who died underwent a postmortem examination. However, charts were reviewed independently by two of the study authors before concluding that a patient died principally from tuberculosis. In summary, this study shows that tuberculosis patients treated by DOT at the start of therapy had significantly higher cure rates compared with those treated by SAT at the start. Patients treated by DOT also had lower rates of tuberculosis-related mortality, but it is uncertain if DOT was the major factor that contributed to the reduced mortality. We conclude that treatment plans that emphasize DOT from the start of therapy for at-risk patients have the greatest success in improving tuberculosis treatment outcomes.
This study was supported by grants from the National Institutes of Health (AI 01549). Conflict of Interest Statement: R.M.J. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; C.B.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; L.C.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; L.M.K. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; D.H.O. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; C.L.D. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form January 21, 2004; accepted in final form June 3, 2004
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
