First-line therapy for HIV is typically initiated when a patient's CD4⁺ T-cell count drops below 200 per cubic millimeter. Severe and colleagues sought to determine whether it might be beneficial to begin treatment earlier. They recruited 816 HIV patients in Haiti who hadn't previously received antiretroviral therapy, hadn't experienced an AIDS illness, and had a CD4⁺ T-cell count between 200 and 350 per cubic millimeter. All patients in this open-label trial received prophylactic trimethoprim-sulfamethoxazole and nutritional support. Those with a positive skin test for tuberculosis were given isoniazid.

Patients were randomized to the early or standard treatment group (n = 408, each) over a three-year period. The early treatment group began receiving lamivudine, zidovudine, and efavirenz within two weeks of enrollment. Drug substitutions were allowed if patients experienced toxic effects, and patients were switched to second-line treatment if first-line therapy failed. Patients in the standard care group received the same treatment when their CD4⁺ T-cell count fell to 200 per cubic millimeter or lower.

Patients were followed for a median of 21 months. A total of 160 patients in the standard group received antiretroviral drug treatment during the trial; their median CD4⁺ T-cell count at the initiation of treatment was 166 per cubic millimeter. Sixteen patients in the standard group died before they were able to start treatment. During the trial, only 1% of patients in the early group died compared with 6% in the standard group. There were 18 incident tuberculosis cases in the early group compared with 36 in the standard group.

The authors concluded that early treatment (started at a CD4⁺ T-cell count of less than 350 per cubic millimeter) decreased death by 75% and tuberculosis incidence by 50%. They recommend initiating treatment when CD4⁺ T-cell counts reach 350 per cubic millimeter instead of waiting until they drop to 200.