‚TB control policy in sub-Saharan Africa should include preventive therapy‚ conclude Jensa Bell and her colleagues on page 1549 . They note that their study is ‚strongly grounded in the assumption that tuberculosis preventive therapy will prolong life‚. The Markov model that they use estimates that life expectancy of a hypothetical cohort of tuberculin skin test (TST)-positive people living with HIV (PLWH) in Uganda could be increased by about 6 months if they were given preventive therapy. The first randomized trial of preventive therapy in PLWH included 63 people with positive TST and did find fewer deaths in the isoniazid group . However, the most recent meta-analysis included five studies and around 1500 TST-positive PLWH from Uganda, Zambia, Kenya, Haiti and Mexico and did not find a significant effect of isoniazid preventive therapy (relative risk 0.79; 95% confidence interval 0.37-1.70) on mortality . Nonetheless, in only one trial  was death more common in the control arm; consequently, the model may reflect reality.
If the effect of preventive therapy on mortality is small and hard to demonstrate, varying the estimate of it as the denominator in a cost-effectiveness analysis will have large effects on costs per year of life saved or other similar measures.
While preventive therapy may have little effect on mortality, it clearly does reduce the incidence of tuberculosis. Furthermore, in many of the analyses presented, preventive therapy actually saves money. Tuberculosis is an infectious disease and, therefore, it is reasonable to take into consideration some of the additional cases that would be prevented by preventing a case that becomes a potential source of infection. Bell and colleagues use a conservative estimate of each patient infecting 4.7 contacts. They predict that the 38126 patients in their cohort will infect about 180000 people of whom 19667 will develop tuberculosis. Such a low number of secondary cases from each primary case suggests that the tuberculosis epidemic should have been self-limiting while, in fact, numbers have risen inexorably. However, even when this conservative estimate is used, their model demonstrates that every TST-positive PLWH who is started on preventive therapy will save medical care money.
Similar arguments, albeit with less-elegant models, have led WHO and UNAIDS to recommend that ‚Preventive therapy should be part of a package of care for people living with HIV/AIDS‚ .
However, to jump from that conclusion to the statement that ‚TB control policy in sub-Saharan Africa should include preventive therapy‚ is a wild leap. Most people in sub-Saharan Africa have no idea whether they are living with HIV or not. Most clinics do not have reagents to perform TST nor are many staff trained in how to interpret it. In cities with a high prevalence of HIV, as many as 25% of adults may be infected with the virus. The sensitivity of TST to detect those infected with Mycobacterium tuberculosis falls markedly with HIV infection . The randomized trials of preventive therapy in Kenya  and Zambia , for instance, that recruited PLWH found that less than 25% were TST positive. Preventive therapy is inappropriate for PLWH who already have signs or symptoms of active tuberculosis. Others cannot yet accept their HIV positive status and so refuse a treatment that may remind them of it daily. Feasibility studies of implementing preventive therapy suggest that less than 50% of PLWH in high-prevalence countries who might be eligible for treatment actually start it . Bell and colleagues show that for every 100000 TST-positive PLWH who take preventive therapy, about 8000 cases will be prevented over the following 8 years or so. To achieve this impact would need about 400000 PLWH to undergo TST, which implies about 800000 PLWH to enter the programme. To find these PLWH will need about 2400 000 adults to have undergone testing. None of the costs of establishing and running such services are included in their analysis.
At the same time, many countries in sub-Saharan Africa are having great difficulties in establishing the most basic requirements for tuberculosis control: quality-assured microscopy services to make an accurate diagnosis of the infectious cases, secure supplies of sufficient antituberculosis chemotherapy to treat the cases diagnosed; adequate supportive systems to enable patients to complete their treatment, efficient supervision and reporting to ensure that they know how many patients are being cured, and the resources needed to improve the situation when patients are not being cured. These must remain the first priorities for tuberculosis control policy. Where these requirements are in place, control programmes may look to expand their policies and preventive therapy is among the options.
Nonetheless, tuberculosis preventive therapy should now be an important facet of HIV care programmes. It is particularly important for those brave enough to choose to find out their HIV status before they become unwell with HIV-related diseases, since this is the population in whom it is easiest to exclude active tuberculosis and who are least likely to be anergic on TST. The pressure to increase access to and availability of voluntary counselling and testing services is rising. Medical options for PLWH are increasing, for example antiviral therapy to prevent vertical transmission from pregnant women  and cotrimoxazole to reduce morbidity  and mortality  in some groups. There is also increasing awareness of the direct benefits of counselling  and testing on coping mechanisms and, through behavioural change, even on HIV incidence .
There are further benefits to establishing tuberculosis preventive therapy services within HIV care programmes. The need to ensure adequate exclusion of active tuberculosis among their clients and to care for those found to have it should enhance collaboration between tuberculosis and HIV programmes. Preventive therapy may become one of the reasons that people are more ready to choose to be tested for HIV. If this leads to even a small reduction in the stigma that still surrounds HIV in much of sub-Saharan Africa it will be a significant step. Ironically, if the result of 2400 000 people undergoing HIV counselling and testing is that HIV incidence falls, either through greater openness concerning HIV in the community or through appropriate behavioural changes in those tested, it is quite possible that more than 8000 cases of tuberculosis will be prevented before a single table of isoniazid is dispensed.
1. Bell JC, Rose DN, Sacks HS. Tuberculosis preventive therapy for HIV-infected people in sub-Saharan Africa is cost-effective. AIDS
2. Pape JW, Jean SS, Ho JL, Hafner A. Johnson WD Jr.
Effect of isoniazid prophylaxis on incidence of active tuberculosis and progress of HIV infection. Lancet
3. Bucher H, Griffith L, Guyatt G et al.
Isoniazid prophylaxis for tuberculosis in HIV infection: a meta-analysis of randomized controlled trials. AIDS
4. Mwinga A, Hosp M, Godfrey-Faussett P, et al.
Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS
5. WHO and UNAIDS. Policy statement on preventive therapy against tuberculosis in people living with HIV.
6. Markowitz N, Hansen N, Wilcosky T et al.
Tuberculin and anergy testing in HIV-seropositive and HIV-seronegative persons. Ann Intern Med
7. 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
8. Aisu T, Raviglione MC, van Praag E, et al.
Preventive chemotherapy for HIV-associated tuberculosis in Uganda: an operational assessment at a voluntary counselling and testing centre. AIDS
9. Shaffer N, Chuachoowong R, Mock P et al.
Short course zidovudine for potential HIV-1 transmission in Bangkok, Thailand: a randomised controlled trial. Lancet
10. Anglaret X, Chene G, Attia A, et al.
Early chemoprophylaxis with trimthoprim-sulphamethoxazole for HIV-1 infected adults in Abidjan, Cote d‚Ivoire. Lancet
11. Wiktor S, Sassan-Morokro M, Grant A, et al.
Efficacy of trimethoprim-sulphamethoxazole prophylaxis to decrease morbidity and mortality in HIV-1-infected patients with tuberculosis in Abidjan, Cote d‚Ivoire: a randomised controlled trial. Lancet
12. Kayewe I, Weinreich S, Chana S, et al.
Clients‚ views on HIV counselling and testing - Is it helpful? XII International Conference on AIDS.
Geneva, June 1998 [abstract 134/33265].
13. Coates T, Sangiwa G, Balmer D, Gregorich S, Kamenga C. Voluntary HIV counselling and testing (VCT) reduces risk behaviour in developing countries: results from the voluntary counselling and testing study. XII International Conference on AIDS.
Geneva, June 1998 [abstract 133/33269].