Isoniazid

HIV-1 serodiscordant couples were randomly assigned in a 1:1 ratio to either an early or delayed strategy for receipt of antiretroviral therapy. Permuted-block randomization was used with stratification according to site. In the early-therapy group, antiretroviral therapy was initiated in the partner with HIV-1 infection at enrollment. In the delayed-therapy group, therapy was initiated after two consecutive measurements in which the CD4 count was 250 cells per cubic millimeter or less or after the development of an illness related to the acquired immunodeficiency syndrome (AIDS). HIV-1–infected participants who had active tuberculosis were excluded, and isoniazid prophylaxis was available, according to local guidelines and practice standards.
After enrollment, study participants were asked to attend three monthly visits, which were followed by quarterly visits unless they became ill or needed additional antiretroviral medications. HIV-1–infected participants who were receiving antiretroviral therapy had one additional visit 2 weeks after starting therapy. HIV-1–uninfected partners were encouraged to return for all visits together for counseling on risk reduction and the use of condoms, for treatment of sexually transmitted infections, and for management of other medical conditions. Some HIV-1–infected participants received trimethoprim–sulfamethoxazole prophylaxis, according to local guidelines.
HIV-1–uninfected partners were tested for HIV-1 seroconversion on a quarterly basis. Samples from all seroconversion events were evaluated at a central laboratory, and results were reviewed by an independent HIV end-point committee. Partners with seroconversion were released from the study and referred to a prearranged local clinic for care.
After the initiation of antiretroviral therapy, virologic failure for HIV-1–infected participants was defined as two consecutive plasma HIV-1 RNA measurements of more than 1000 copies per milliliter at 16 weeks or later. Assessment for clinical signs and symptoms, laboratory measurements, interviews about sexual behavior, review of adherence to the antiretroviral regimen (including a self-reported questionnaire and pill counts), and adherence counseling were conducted at each visit. (Details regarding study procedures, including guidelines for adherence counseling, are provided in the Supplementary Appendix.)
Any woman who was pregnant at enrollment or became pregnant was provided antiretroviral therapy appropriate for use during pregnancy at the start of the second trimester. On the basis of the judgment of the site investigator, women in the delayed-therapy group discontinued antiretroviral therapy after delivery or when breast-feeding ended. A new partner could be enrolled with an HIV-1–infected participant if the original partner was released from the study and the new HIV-1– uninfected partner met all inclusion or exclusion criteria.