Mosimaneotsile, Barudi RN, MPH*; Mathoma, Anikie RN*; Chengeta, Bafanana RN, MSN*; Nyirenda, Samba MD*; Agizew, Tefera B MD, MPH*; Tedla, Zegabriel MD*; Motsamai, Oaitse I RN, MPH†; Kilmarx, Peter H MD‡; Wells, Charles D MD§; Samandari, Taraz MD, PhD*§
Tuberculosis (TB) remains a principal source of morbidity and mortality in sub-Saharan Africa in HIV-infected adults.1 In Botswana, it has been demonstrated that 40% of AIDS deaths were due to TB2 and that over 80% of TB patients are coinfected with HIV.3,4 Isoniazid preventive therapy (IPT) has been shown to be efficacious in reducing morbidity from TB in several clinical trials in African settings.5,6
The World Health Organization (WHO) recommended a 6-month course of IPT for persons living with HIV (PLWH) in TB-endemic countries in 1993.7 Although policies for IPT currently exist in 84 countries, 15 years after making this recommendation, the WHO reported that, “Provision of IPT remains at very low levels, with reported numbers treated with IPT reaching only 27,056 in 2006-equivalent to less than 0.1% of the estimated 33 million people estimated to be infected with HIV globally.”8
Two commonly cited reasons why national program managers are reluctant to provide IPT are the concerns that isoniazid may inadvertently be provided to recipients with active TB and thereby increase rates of isoniazid-resistant TB, and that the rate of adverse events particularly among those concomitantly taking antiretroviral therapies (ARTs) may be excessive. Despite such concerns and given the rapidly expanding TB epidemic it faced, Botswana's Ministry of Health embarked on a National IPT Program in 2001 which does not require a positive tuberculin skin test (TST) for enrolment. It accounted for 70% of the total number of PLWH reported to be on IPT worldwide in 2006.8
Botswana is also unique as it established a free national ART program in 2002 that by 2008 reached 90%-95% of the estimated PLWH needing ART.9 By itself ART is known to reduce the incidence of TB by 80%10,11 and it may be that the addition of IPT to ART may reduce TB by a further 58%-76%.12,13
Six months of open-label isoniazid was provided prospectively to a cohort of HIV-infected adults as per the National IPT Program guidelines. It was after the first 6 months of IPT that participants received randomly assigned, coded medication (isoniazid or placebo) for an additional 30 months as part of a clinical trial. The objective of the clinical trial was to determine whether continuous isoniazid prophylaxis is superior to a 6-month course because, in countries such as Botswana where TB is endemic, reinfection with TB is common.14-16
In this report we document our experience with the provision of 6 months of open-label isoniazid to a large cohort of HIV-infected adults in a country where ART is widely available. We present information about enrolment, adverse events, TB incidence, and most importantly adherence. As the IPT regimen used in Botswana is presently recommended by WHO for PLWH in TB-endemic countries, our observations are valuable to other countries similarly affected by these 2 diseases.
Potential candidates for the clinical trial were approached at 8 public health clinics in the 2 largest cities in Botswana between November 24, 2004 and June 30, 2006. These clinics provide services to PLWH from the National IPT Program and the National ART Program. Typically, after being diagnosed with HIV infection, candidates were referred for IPT as their first point of HIV care. Once they presented for IPT from the National Program they were, without screening, referred to a clinical trial nurse based at a different office within the same clinic. The nurse then began the screening process for interested candidates. After enrolment, if a participant's CD4 lymphocyte count was found to be <200 cells/mm3 (CD4 < 200), ART was offered through the National ART Program. The first-line ART provided was zidovudine, lamivudine, and either nevirapine or efavirenz.
For the purposes of the clinical trial, a sample size of 2000 PLWH was derived from estimates of TB infection in the general population. A candidate had to be 18 years of age or older. Screening occurred in 2 stages. The first screening (screen 1) used the exclusion criteria of the National IPT Program: lack of documentation of HIV infection, illness (eg, fever, cough, AIDS-defining illnesses,17 or on physical exam lymphadenopathy, jaundice, or pulmonary disease), history of active TB in the last 3 years, current receipt of anti-TB therapy (ATT), previous IPT, current pregnancy, a history of hepatitis, nonadherence to chronic therapies, or adverse reactions to isoniazid. A positive TST was not a criterion for receipt of IPT through the National Program or the clinical trial. An additional study criterion at screen 1 required that participants have a Karnofsky score above 60%, that is, ability to perform activities of daily living with no more than occasional assistance, and requiring medical care less than every 3 months.
Upon passing screen 1, candidates underwent the second screening (screen 2) which had exclusion criteria specific to the clinical trial: an abnormal chest radiograph with no history of TB or pneumonia, neutrophil count <1000 cells/mm3, hemoglobin <6.5 g/dL, platelets <75,000/mm3, serum aspartate aminotransferase >85 U/L (2.5 times above the upper limit of normal, ULN), serum alanine aminotransferase >103 U/L (2.5× >ULN), total bilirubin >39 μmol/L (1.5× >ULN), serum creatinine >39 μmol/L, or a positive urine pregnancy test. Candidates indefinitely taking phenytoin, warfarin, or carbamazepine were also excluded as isoniazid may interfere with their metabolism. Participants had 2 opportunities to pass a 20-question comprehension test with a minimum passing score of 80%.
HIV status was determined by government nonstudy personnel using 2 ELISAs run in parallel (Ortho Kit HIV-1 and HIV-2 or Abbot Murex) or dual rapid ELISA (Determine Assay, Unigold, Capillus, or Orasure). Two positive ELISA tests were considered diagnostic for HIV infection. TSTs were administered by placing 5 tuberculin units (0.1 mL) of purified protein derivative (RT/23 Statens Seruminstitut, Copenhagen, Denmark) subcutaneously. The tests were read by study nurses within 48-72 hours by the ruler method and after September 2005 using digital calipers. A TST ≥5 mm was considered positive. Although CD4 count was measured, candidates with any CD4 count, whether or not receiving ART, could be enrolled.
At enrolment, as per Botswana national guidelines, we provided 100-mg isoniazid tablets to be taken daily at a dose of 300 mg for weight 30-49 kg and 400 mg for weight ≥50 kg. All participants received 25 mg of pyridoxine daily. In late 2005, the national guidelines changed and beginning January 1, 2006, all study participants were provided 300 mg of isoniazid daily. Each month, all participants returned for an evaluation and to receive the next month's supply of isoniazid and pyridoxine. An acceptability questionnaire was administered at the 6-month visit.
All participants identified with TB (incident cases) during the 6-month open-label period had clinical evidence consistent with TB and a clinical response to ATT. TB was categorized as definite, probable, or possible by the following criteria: definite TB ≥1 sputum culture positive for TB, probable TB if ≥1 smear positive for AFB, possible TB if both smear and cultures were negative. A possible TB death was a death in a participant with clinical or verbal autopsy evidence consistent with TB as the proximate cause of death.
Adverse events were graded from 1 to 5 as per National Cancer Institute Common Terminology Criteria for Adverse Events v3.018 with a severe adverse event defined as ≥grade 3. Grades 3, 4, and 5 correspond to severe, life-threatening or disabling, and death, respectively.
Nonadherence was defined as missing ≥1 visit during the 6-month course of IPT or, in the case of those who suffered death, adverse events or TB, missing >17% (one-sixth) of visits up until the time of the event. Pills in returned bottles were counted at months 1, 3, and 6 and nurses used a chart to assess adherence.
Data were collected using standardized case report forms and double entered into a Clindex database (Fortress Medical Systems, Minneapolis, MN). The impact of the stages of screening on the population referred was described, demographics of the cohort enrolled were assessed, and adherence to treatment by the cohort was determined. Proportions were compared with the uncorrected Chi square test using Epi Info (CDC, Atlanta, GA). Using SAS (SAS Institute, version 9.1, Cary, NC), relative risks (RRs) and odds ratios were determined to more closely examine factors associated with adherence, a logistic regression model (adjusted odds ratios, aORs) was developed and interactions between significant variables were assessed. P values of less than 0.05 were considered statistically significant.
All participants signed informed consents and received a transportation reimbursement. The protocol was approved by the Botswana and Centers for Disease Control and Prevention (CDC) ethics committees. The clinical trial registration number is NCT00164281 and may be viewed at clinicaltrials.gov.
Figure 1 shows the stages at which participants were screened, ineligibility by IPT Program criteria and then by clinical trial criteria and the number enrolled. Seventy-three percent of candidates who agreed to be screened were eligible for IPT using Program criteria. Clinical trial criteria excluded 26% (776/2934) from enrolment and 1995 participants started 6 months of open-label IPT. Candidates excluded from the trial but eligible for the Program were referred to the National IPT Program. The proportion of men (35%, 1504/4331) excluded at screen 1 was significantly higher (P < 0.001) than the proportion of men enrolled (28%, 559/1995). Those enrolled represented 46% (1995/4331) of PLWH initially screened.
The principal reason for exclusion at screen 1 (Table 1) was illness (66%, 727/1096). The main reasons for exclusion at screen 2 were having a chest radiograph potentially compatible with TB (36%, 296/826), followed by a low neutrophil count (17%, 141/826) and living too far from the study site (15%, 120/826). We have described elsewhere the outcomes of those screened with abnormal chest radiographs.19
In the cohort enrolled, 72% (1436/1995) were women, the median age was 32 years, 65% (1302/1995) had at least a secondary education, and 32% (635/1994) were unemployed (Table 2). Body mass indices showed that 18% (329/1855) were underweight, 17% (321/1855) were overweight, and 9% (169/1855) were obese. Although all participants were recruited at the 8 clinics, 34% (682/1994) were referred there from voluntary counseling and testing centers. Twenty-four percent (468/1917) had a positive TST, 31% (607/1990) had a CD4 < 200, and 1.4% (28/1995) were on ART at enrolment. By the end of 6 months, 2 were lost to follow-up and 22% (444/1993) of all participants initiated ART at a median of 3 months postenrolment.
Outcomes During the 6 Months of Observation
Figure 2 shows key outcomes among the 1995 enrolled PLWH during 6 months of IPT. Eight participants (0.40%, 8/1993) were diagnosed with TB: 3 definite and 2 possible incident cases and 3 possible TB deaths. In the first month after IPT initiation, one definite TB case was diagnosed. This participant had a retrocardiac mass on the screening chest radiograph that was not noted initially; the other 7 cases had normal chest radiographs at baseline. All but 1 case had pulmonary TB: 1 possible TB death may have had extrapulmonary TB. One incident case of TB was diagnosed at month 1, one at month 2, 2 at month 3 (one was a possible TB death), one at month 5 and 3 at month 6 (2 of the latter 3 were possible TB deaths). Four of the TB cases had a CD4 < 200. In 2 cases, TB was diagnosed after 2 months of ART; although immune reconstitution disease was theoretically possible, in these cases it was considered unlikely.
Isoniazid resistance was not detected among the Mycobacterium tuberculosis isolates cultured from the 3 definite cases: 2 were sensitive to isoniazid, rifampin, ethambutol, and streptomycin and 1 was resistant to both rifampin and streptomycin. The latter case remained on second line ATT for polyresistant TB 26 months after ATT initiation. With the exception of the latter case and 3 possible TB deaths, the others received standard 6-month courses of ATT with resolution of symptoms and chest radiographic abnormalities.
Severe adverse events were observed in 1.4% (28/1993) of participants: 2 definitely, 12 probably, and 14 possibly associated with isoniazid. Twenty-three of these were grade 3, 4 were grade 4, and there was 1 death (grade 5) from hepatic encephalopathy probably associated with isoniazid. Among the 28 severe events, 19/1993 (0.95%) were hepatitis (1 died from hepatic encephalopathy, 2 had jaundice, 3 had other symptoms, 13 had no symptoms, 7 were concomitantly receiving ART), 5/1993 (0.25%) had rash, 2 had seizures (1 of these was a suicide attempt by overdosing on isoniazid), 1 had vertigo, and 1 had a headache. The single death from isoniazid-associated hepatitis represents a rate of 50/100,000 persons initiating IPT. As 15/28 (54%) of participants with severe adverse events were symptomatic, 0.75% (15/1993) of all participants had symptomatic severe adverse events. The cases of hepatitis are analyzed in more detail elsewhere (Tedla et al, submitted).
Excluding the case of hepatic encephalopathy, there were 20 deaths (1.0% of 1993 participants) in the following disease categories: central nervous system disease (5), gastroenteritis (6), respiratory illness (2), phenytoin-associated toxic epidermal necrosis (1), automobile accident (1), Kaposi sarcoma (1), cardiac (1), and unknown (3). The median CD4 count of those who died was 98 cells/mm3 (range 4-625). Nine of 16 (56%) deaths with sufficient information were due to AIDS-defining illnesses.
Adherence to IPT
Eighty-six percent (1710/1995) were adherent to the entire 6-month course of IPT (Fig. 2). The 213 nonadherent participants who did not have valid reasons for nonadherence consisted of 8 who voluntarily withdrew, 2 who were lost to follow-up, and 203 who had fewer than 6 visits. Using pill counts for those with pill counts available for months 1, 3, and 6 (n = 1511), 91% (1375/1511) took ≥80% of their pills.
Women were significantly more adherent to IPT than men (Table 3). Those receiving ART during the first 6 months were 1.41-fold more adherent [95% confidence interval (CI) 1.04 to 1.91], and participants with CD4 < 200 at enrolment were more likely to be nonadherent. Lack of education or being underweight was not associated with nonadherence. In a multivariate logistic regression model that included significant variables the association between adherence to IPT and being on ART increased (aOR 4.38, 95% CI 2.86 to 6.71). There was no significant interaction in the model between CD4 < 200 and being on ART.
IPT was provided to 1995 PLWH in a clinical trial managed within public health clinics in 2 Botswana cities. The National IPT Program provides its services through these same clinics among hundreds of other clinics. As CD4 count is not an eligibility criterion for IPT, IPT is typically the first service provided to PLWH after their diagnosis of HIV infection. We present data relevant to the National Program in relation to screening and observation during the course of 6 months of IPT in a large cohort that had an exceptionally low loss to follow-up.
Excluding TB disease at screening is a key step before IPT initiation. At screen 1, the reason for exclusion of two-thirds was due to illness-similar to the proportion reported from the pilot of the IPT Program in Botswana.20 Among asymptomatic screened PLWH from our cohort-most of whom were excluded-we have shown a 1.6% prevalence of TB disease,19 which is higher than the rate reported in the pilot. Among those enrolled and presented in this report, it is likely that a single participant who had asymptomatic TB disease at screening was diagnosed in the first month of IPT. The TB rate during IPT in our study (0.40%) was similar to a 0.28% rate reported in a clinical trial from Thailand21 but lower than a 2% rate reported in a clinical trial from Uganda22 and lower than a 3% rate in a programmatic evaluation in Uganda.23 We speculate that the lower rate of TB observed by the Thai trial and ours was due to participants' concomitant receipt of ART, 10.6% and 22%, respectively.
The observed rates of hepatitis, hepatic encephalopathy, and rash associated with isoniazid were within the expected ranges: 0.5%-5.3%,6,24,25 0-56/100,000,26-28 and 0.3%-0.9%5,29,30 respectively. The National IPT Program does not screen for biochemical hepatitis for enrolment and only stops IPT if a recipient develops symptoms; we found that 0.75% had symptomatic severe adverse events including hepatitis, which is lower than the report from a program setting in Uganda, where 1.7% of IPT recipients stopped therapy due to side effects.23 It is quite possible that among those who had only chemical hepatitis (ie, asymptomatic elevations in transaminases) in our study, some may have progressed to symptomatic disease but most would have resolved without any symptoms while continuing therapy.31
The death rate in this study was likely mitigated by the provision of ART to study participants: two-thirds of PLWH with a screening CD4 < 200 were on ART by the end of the 6-month period. A 2.0% 1-year mortality of participants in this study (data not shown) was comparable to the 1-year mortality observed in other observational cohorts in resource-constrained countries where ART was available.32-35
The 86% adherence reported in this study as measured by attendance at all 6 scheduled monthly visits is somewhat higher than reports from most clinical trials. This measure of adherence is more stringent than the National IPT Program's requirement as the Program permits completion of the equivalent of 6 months IPT within 9 months. TB prevention clinical trials in several countries reported 69%-86% adherence rates using various definitions.21,22,25,36
We observed that adherence to IPT seems to have benefited 4.38-fold through participation in the ART Program (95% CI 2.86 to 6.71). This effect was independent of having a baseline CD4 < 200 or female sex. Moreover, it is unlikely that this effect is merely due to survivor bias among those who were adherent as the effect remained significant in the multivariate model even if those who became nonadherent within the first 3 months were excluded. Provision of IPT while PLWH also receive ART must be tempered by the fact that use of ART concomitantly with IPT may result in a higher rate of isoniazid-associated hepatitis possibly requiring monthly monitoring of hepatic enzymes in the first 3 months. Integration of ART and IPT services to PLWH would certainly improve service delivery to PLWH but may also be particularly effective as preliminary evidence suggests that the concomitant or sequential provision of IPT and ART synergistically reduces TB incidence in PLWH.12,13 Finally, as there are 635 clinics providing IPT and only 100 ART clinics nationwide, IPT should continue to be provided through current venues.
Although PLWH were referred for IPT from local clinics and voluntary counseling and testing center, few were referred from either prevention of mother-to-child transmission (PMTCT, 0.6%) or ART (1.0%) clinics. Because PLWH are eligible for IPT at any CD4 count and are thus first referred to IPT, most will have received IPT before enrolment in the ART Program. The IPT-followed-by-ART sequence may reduce the incidence of the unmasking form of the immune reconstitution inflammatory syndrome as observed in our cohort. PMTCT programs could improve upon the number of HIV-infected women referred for IPT postpartum.
Findings of this report are useful to the National IPT Program in assessing its performance by showing the proportion that it may expect to exclude, the proportion of recipients that may develop active TB after screening and the rates of various adverse events. Several findings may not be as relevant to the National IPT Program and are limitations to generalization: (1) the clinic sites in the 2 major cities may not be representative of all clinics providing IPT nationwide, and, as rural people tend to move to cities seeking care, our clinics may have had a higher proportion of PLWH with more advanced disease (personal communication, O.M.); (2) adherence in this clinical trial was likely much higher than what would be expected within a program given that the cohort was motivated to participate in a 36-month clinical trial, was required to live near the study sites, and was provided incentives and intensive adherence counseling; (3) it is possible that the clinical trial identified more deaths, adverse events, and TB cases than would be expected in the programmatic context although deaths may also have been lower as the enrolled participants were likely healthier given the minimum hemoglobin and neutrophil count inclusion criteria. Another limitation to our conclusions is that as a measure of adherence, clinic attendance is imperfect. Routine monitoring for urine isoniazid was considered but proved impractical for this large cohort. Although pill counts or electronic monitoring systems are useful, they too have inherent limitations such as participant obfuscation as observed in some participants in this trial. Similar to the preference of an expert panel in a Ugandan TB prevention study,36 we chose attendance as our measure of adherence as it is a good indication of a participant's intent to comply.
In summary, we observed that a 6-month course of IPT for PLWH in Botswana was relatively safe and well tolerated even with concomitant receipt of ART by over a quarter of the participants. Few cases of TB occurred during the observational period. Adherence to IPT was excellent, particularly among those receiving ART. As ART clinics presently do not provide IPT, we recommend that in addition to public health clinics perhaps initially as a pilot program, ART clinics also begin provision of IPT because patient inconvenience will be reduced and adherence will improve. As Botswana and other sub-Saharan African countries expand access to both ART and IPT, it will be important to assess their combined impact on the reduction of TB where TB and HIV remain a major coepidemic.
The authors thank the study participants, the data entry team, and the hard work of the research nurses who carried out the day-to-day work of the IPT Trial. The study would not have been possible without the generous cooperation of Ministry of Health and Ministry of Local Government health workers. We acknowledge the guidance and support of the members of the Data and Safety Monitoring Board: Drs. Jonathan Levin, Ndwapi Ndwapi, Andrew Nunn (Chair), and Karin Weyer, and greatly value the key role of the Trial's Endpoints Committee: Drs. John L. Johnson, Marape Marape, Abraham Miranda (Chair), and Helmuth Reuter. The coauthors appreciate the contributions of Drs. Elizabeth Talbot, Themba Moeti and Howard Moffat to the original conception and design of the protocol.
1. World Health Organization. Global Tuberculosis Control: Surveillance, Planning, Financing
. Geneva, Switzerland: World Health Organization; 2007;376.
2. Ansari NA, Kombe AH, Kenyon TA, et al. Pathology and causes of death in a group of 128 predominantly HIV-positive patients in Botswana, 1997-1998. Int J Tuberc Lung Dis
3. Gammino VM, Mboya JJ, Samandari T, et al. Baseline evaluation of routine HIV testing among tuberculosis patients in Botswana, 2005. Int J Tuberc Lung Dis
. 2008;12(Suppl 1):92-94.
4. Talbot EA, Hay B, Hone NM, et al. Tuberculosis serodiagnosis in a predominantly HIV-infected population of hospitalized patients with cough, Botswana, 2002. Clin Infect Dis
5. Mwinga A, Hosp M, Godfrey-Faussett P, et al. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS
6. Whalen CC, Johnson JL, Okwera A, et al. A trial of three regimens to prevent tuberculosis in Ugandan adults infected with the human immunodeficiency virus. Uganda-Case Western Reserve University Research Collaboration. N Engl J Med
7. Anonymous. Tuberculosis preventive therapy in HIV-infected individuals. A joint statement of the WHO Tuberculosis Programme and the Global Programme on AIDS, and the International Union Against Tuberculosis and Lung Disease. Weekly Epidemiol Rec
8. World Health Organization. Global Tuberculosis Control: Surveillance, Planning, Financing
. Geneva, Switzerland: World Health Organization; 2008;393.
9. Ojikutu B, Makadzange AT, Gaolathe T. Scaling up ART treatment capacity: lessons learned from South Africa, Zimbabwe, and Botswana. Curr Infect Dis Rep
10. Badri M, Wilson D, Wood R. Effect of highly active antiretroviral therapy on incidence of tuberculosis in South Africa: a cohort study. Lancet
11. Santoro-Lopes G, de Pinho AM, Harrison LH, et al. Reduced risk of tuberculosis among Brazilian patients with advanced human immunodeficiency virus infection treated with highly active antiretroviral therapy. Clin Infect Dis
12. Golub JE, Saraceni V, Cavalcante SC, et al. The impact of antiretroviral therapy and isoniazid preventive therapy on tuberculosis incidence in HIV-infected patients in Rio de Janeiro, Brazil. AIDS
13. Golub JE, Pronyk P, Mohapi L, et al. Isoniazid preventive therapy, HAART and tuberculosis risk in HIV-infected adults in South Africa: a prospective cohort. AIDS
14. Easterbrook PJ, Gibson A, Murad S, et al. High rates of clustering of strains causing tuberculosis in Harare, Zimbabwe: a molecular epidemiological study. J Clin Microbiol
15. Gilks CF, Godfrey-Faussett P, Batchelor BI, et al. Recent transmission of tuberculosis in a cohort of HIV-1-infected female sex workers in Nairobi, Kenya. AIDS
16. Glynn JR, Crampin AC, Yates MD, et al. The importance of recent infection with Mycobacterium tuberculosis in an area with high HIV prevalence: a long-term molecular epidemiological study in Northern Malawi. J Infect Dis
17. Interim WHO Clinical Staging of HIV/AIDS and HIV/AIDS Case Definitions for Surveillance, African Region. WHO/HIV/2005
19. Agizew T, Arwady MA, Yoon JC, et al.Tuberculosis among asymptomatic HIV-infected adults with abnormal chest radiographs screened for TB preventive therapy. Int J Tuberc Lung Dis
. In press.
20. Mosimaneotsile B, Talbot EA, Moeti TL, et al. Value of chest radiography in a tuberculosis prevention programme for HIV-infected people, Botswana. Lancet
21. Hiransuthikul N, Nelson KE, Hiransuthikul P, et al. INH preventive therapy among adult HIV-infected patients in Thailand. Int J Tuberc Lung Dis
22. Lugada ES, Watera C, Nakiyingi J, et al. Operational assessment of isoniazid prophylaxis in a community AIDS service organisation in Uganda. Int J Tuberc Lung Dis
23. Mugisha B, Bock N, Mermin J, et al. Tuberculosis case finding and preventive therapy in an HIV voluntary counseling and testing center in Uganda. Int J Tuberc Lung Dis
24. Gordin FM, Matts JP, Miller C, et al. A controlled trial of isoniazid in persons with anergy and human immunodeficiency virus infection who are at high risk for tuberculosis. Terry Beirn Community Programs for Clinical Research on AIDS. N Engl J Med
25. Hawken MP, Meme HK, Elliott LC, et al. Isoniazid preventive therapy for tuberculosis in HIV-1-infected adults: results of a randomized controlled trial. AIDS
26. Kopanoff DE, Snider DE Jr, Caras GJ. Isoniazid-related hepatitis: a U.S. Public Health Service cooperative surveillance study. Am Rev Respir Dis
27. Nolan CM, Goldberg SV, Buskin SE. Hepatotoxicity associated with isoniazid preventive therapy: a 7-year survey from a public health tuberculosis clinic. JAMA
28. Snider DE Jr, Caras GJ. Isoniazid-associated hepatitis deaths: a review of available information. Am Rev Respir Dis
29. Byrd RB, Horn BR, Solomon DA, et al. Toxic effects of isoniazid in tuberculosis chemoprophylaxis. Role of biochemical monitoring in 1,000 patients. JAMA
30. Churchyard GJ, Fielding K, Charalambous S, et al. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS
31. Navarro VJ, Senior JR. Drug-related hepatotoxicity. N Engl J Med
32. Corey DM, Kim HW, Salazar R, et al. Brief report: effectiveness of combination antiretroviral therapy on survival and opportunistic infections in a developing world setting: an observational cohort study. J Acquir Immune Defic Syndr
33. Ferradini L, Jeannin A, Pinoges L, et al. Scaling up of highly active antiretroviral therapy in a rural district of Malawi: an effectiveness assessment. Lancet
34. Laurent C, Ngom Gueye NF, Ndour CT, et al. Long-term benefits of highly active antiretroviral therapy in Senegalese HIV-1-infected adults. J Acquir Immune Defic Syndr
35. Wester CW, Kim S, Bussmann H, et al. Initial response to highly active antiretroviral therapy in HIV-1C-infected adults in a public sector treatment program in Botswana. J Acquir Immune Defic Syndr
36. Pekovic V, Mayanja H, Vjecha M, et al. Comparison of three composite compliance indices in a trial of self-administered preventive therapy for tuberculosis in HIV-infected Ugandan adults. Uganda-Case Western Reserve University Research Collaboration. J Clin Epidemiol
© 2010 Lippincott Williams & Wilkins, Inc.