Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Preventive therapy reduces the incidence of active tuberculosis in the short to medium term in patients with HIV


ACP J Club. 2001 May-June;134:102. doi:10.7326/ACPJC-2001-134-3-102

Source Citation

Wilkinson D. Drugs for preventing tuberculosis in HIV infected persons. Cochrane Database Syst Rev. 2000;(4):CD000171 (latest version 30 Aug 2000). [PubMed ID: 11034676] (All 2001 articles were reviewed for relevancy, and abstracts were last revised in 2007.)



In patients with HIV who are at risk for tuberculosis (TB), how effective is preventive therapy in reducing the development of active TB?

Data sources

Studies were identified by searching the Cochrane Controlled Trials Register, the trials register of the Cochrane Infectious Diseases Group, MEDLINE (1982 to 2000), and EMBASE/Excerpta Medica (1980 to 2000); by scanning the reference lists of known published works and reviews; and by contacting key agencies and current researchers in the field.

Study selection

Studies were selected if they were randomized controlled trials involving adults with HIV who were at high risk for TB infection and if TB-prevention drugs were compared with placebo or each other.

Data extraction

Data were extracted on study quality and staging of HIV infection. Outcomes extracted were diagnosis of active TB, mortality, adverse drug reaction, and progression of HIV disease.

Main results

7 studies were included; 5 were placebo controlled (4652 patients), and 3 compared isoniazid with rifampin plus pyrazinamide (1 placebo-controlled trial had an active therapy comparison) (2725 patients). When the placebo-controlled trials were pooled, fewer patients who received preventive therapy had active TB than did patients who received placebo (Table). The groups did not differ for mortality (Table). In patients with positive results on tuberculin skin testing (n = 2361), the incidence of TB and mortality were reduced (odds ratio [OR] 0.35, 95% CI 0.21 to 0.59 and OR 0.70, CI 0.50 to 0.98, respectively) among those who received preventive therapy. In patients with negative results (n = 2202), groups did not differ for TB or mortality (OR 0.82, CI 0.51 to 1.31 and OR 1.04, CI 0.86 to 1.28). Adverse drug reactions were seen in more patients who received preventive therapy than in those who received placebo (Table). 1 trial that assessed progression of HIV disease showed no difference between preventive therapy and placebo. 3 trials that compared isoniazid with rifampin plus pyrazinamide showed no difference for active TB or mortality.


In patients with HIV and a positive result on tuberculin skin testing, preventive therapy reduces the development of active tuberculosis in the short to medium term.

Sources of funding: Department for International Development UK and European Commission (Directorate General XII) Belgium.

For correspondence: Professor D. Wilkinson, South Australian Centre for Rural and Remote Health, University of South Australia-Whyalla Campus, Nicholson Avenue, Whyalla Norrie, South Australia 5608, Australia. FAX 61-8-8647-6164.

Table. Preventive therapy vs placebo in patients with HIV who are at risk for tuberculosis (TB)*

Outcomes at 15 to 33 mo Weighted event rates RRR (95% CI) NNT (CI)
Preventive therapy Placebo
Active TB 3.1% 5.3% 41% (21 to 56) 46 (30 to 112)
Mortality† 22% 21.6% 2% (-18 to 18) Not significant
Adverse drug reaction 4.3% 2.6% 79% (24 to 159) 59 (38 to 143)

*Abbreviations defined in Glossary; RRR, RRI, NNT, NNH, and CI calculated from data in article.
†Random effects used in analysis.

Revised Commentary(2007)

This important meta-analysis by Wilkinson of 7 published randomized controlled trials, updated in 2003 to include 11 trials (1), confirms that preventive TB therapy benefits adults with HIV. Only certain patients, however, showed benefit. 2 distinct HIV-patient groups were analyzed: those with a positive result on tuberculin skin testing and patients “at high risk” for TB with a negative skin test. The latter group is more heterogeneous than is the former group, and benefit could not be shown (given the period of prophylaxis and follow-up) for this diverse group.

Important ramifications exist for those with HIV and positive results on tuberculin skin testing. Clearly, isoniazid prophylaxis (6 to 12 mo) lowers the incidence of active TB and decreases mortality. It is therefore no longer debatable whether prophylaxis should be offered such patients. The only question is which therapy is most effective? These studies did not have the power to determine whether shorter combination therapy was superior to longer isoniazid prophylaxis. Furthermore, they were not designed to evaluate long-term (> than 3 y) benefits or to compare different durations of single or multiple-drug regimens.

Thus, we do not know whether lifelong prophylaxis is superior to limited-duration prophylaxis. Side effects requiring discontinuation were greater with short course multi-drug regimens compared with isoniazid (1). The use of 2-month combination regimens containing pyrazinamide is no longer recommended because of reports of severe liver injury (2).

This meta-analysis showed that providing drug prophylaxis for 46 tuberculin-positive patients with HIV prevents 1 additional active case of TB at 15 to 33 months. This information must be translated into clinical practice.

Aaron E. Glatt, MD
St. Vincent's Catholic Medical Centers, Brooklyn/Queens Region
Jamaica, New York, USA

References (2007)

1. Woldehanna S, Volmink J. Treatment of latent tuberculosis infection in HIV infected persons. Cochrane Database Syst Rev. 2004:CD000171. [PubMed ID: 14973947]

2. Centers for Disease Control and Prevention (CDC); American Thoracic Society. Update: adverse event data and revised American Thoracic Society/CDC recommendations against the use of rifampin and pyrazinamide for treatment of latent tuberculosis infection-United States, 2003. MMWR Morb Mortal Wkly Rep. 2003;52:735-9. [PubMed ID: 12904741]