The results of the review were as follows:
The annual rate of active tuberculosis in patients who did not receive prophylaxis was 6.7% (range: 3.4 - 10).
The all-cause annual mortality rate in HIV-infected patients without history of tuberculosis was 10% (range: 5 - 10).
All-cause annual mortality rate in HIV-infected patients after diagnosis of active tuberculosis increased to 35% (range: 20 - 35).
The effectiveness of prophylaxis in reducing the incidence of active tuberculosis:
12-month course of isoniazid prophylaxis, 83% (range: 50 - 90);
(a) isoniazid daily for 6 months, 67% (range: 23 - 86);
(b) isoniazid twice weekly for 6 months, 75% (range: 30 - 85);
(c) isoniazid and rifampin daily for 3 months, 60% (range: 14 - 82);
(d) isoniazid, rifampin and pyrazinamide daily for 3 months, 49% (range: 0 - 76);
(e) rifampin and pyrazinamide twice weekly for 2 months, 73% (range: 30 - 85);
and (f) rifampin and pyrazinamide daily for 2 months, 82% (range: 30 - 90).
It was assumed that effectiveness lasted for 3 years, and diminished to zero after 5 years.
The rates of mild adverse events for the 7 treatment regimes were assumed to be:
12-month course of isoniazid prophylaxis, 10.1% (range: 1 - 30);
(a) isoniazid daily for 6 months, 5.4% (range: 1 - 30);
(b) isoniazid twice weekly for 6 months, 3% (range: 1 - 30);
(c) isoniazid and rifampin daily for 3 months, 3.6% (range: 1 - 30);
(d) isoniazid, rifampin and pyrazinamide daily for 3 months, 16.9% (range: 1 - 30);
(e) rifampin and pyrazinamide twice weekly for 2 months, 3% (range: 1 - 30);
and (f) rifampin and pyrazinamide daily for 2 months, 12.5% (range: 1 - 30).
Moderate to severe adverse event rates were:
12-month course of isoniazid prophylaxis, 0.6% (range: 0 - 10);
(a) isoniazid daily for 6 months, 0.7% (range: 0 - 10);
(b) isoniazid twice weekly for 6 months, 0% (range: 0 - 10);
(c) isoniazid and rifampin daily for 3 months, 1.1% (range: 0 - 10);
(d) isoniazid, rifampin and pyrazinamide daily for 3 months, 2.8% (range: 0 - 10);
(e) rifampin and pyrazinamide twice weekly for 2 months, 0% (range: 0 - 10);
and (f) rifampin and pyrazinamide daily for 2 months, 0% (range: 0 - 10).
Finally, fatal adverse event rates were:
12-month course of isoniazid prophylaxis, 0.002% (range: 0 - 0.02);
(a) isoniazid daily for 6 months, 0.001% (range: 0 - 0.02);
(b) isoniazid twice weekly for 6 months, 0.001% (range: 0 - 0.02);
(c) isoniazid and rifampin daily for 3 months, 0.001% (range: 0 - 0.02);
(d) isoniazid, rifampin and pyrazinamide daily for 3 months, 0.001% (range: 0 - 0.02);
(e) rifampin and pyrazinamide twice weekly for 2 months, 0% (range: 0 - 0.02);
and (f) rifampin and pyrazinamide daily for 2 months, 0% (range: 0 - 0.02).
The 1-year survival rate for HIV infected patients with an episode of active tuberculosis was 65% and without an episode of active tuberculosis was 90%.
Quality-of-life value for HIV-infected patients with a CD4 count < 200 cells/mm3 and no opportunistic infections, and for patients with tuberculosis was 0.65 (range: 0.4 - 1.0) and 0.62 (range: 0.4 - 1.0).
The utility values for patients who had mild and moderate to severe adverse events were 0.5 for 1 week (0.25 for 1 month - 0.99 for 1 week) and 0.25 for 3 weeks (0.25 for 1 month - 0.99 for 1 week). These data were used as inputs to the model.