• Cumulative treatment failure rate, determined by comparing a primary health care level first line antiretroviral therapy regimen to a doctor monitored treatment regimen
  
• 48-week cumulative virology failure rate, determined by comparing community-based directly observed therapy (DOT) by identified and trained community members to continued clinic-based treatment support for subjects who have failed first line therapy
  

• Overall clinical safety of antiretroviral therapy, as measured by the occurrence of clinical and laboratory Grade 3 and 4 adverse events, between primary health care monitoring arms in adults and children
  
• CD4+ count, viral load, drug toxicity, adherence, participation in MTCT programs, intercurrent medical conditions, age as factors potentially contributing to cumulative treatment failure in participants age 16 and older
  
• cumulative treatment failure rate between groups, defined by clinical staging (CDC and WHO classification) prior to initiation of antiretroviral therapy and monitoring arms in adults and children
  
• cumulative treatment failure rate of second line therapy from time of Phase 1 randomization in adults and children, between the doctor and the primary health care sister-based monitoring arm
  
• comparison of immune reconstitution in adults and children, using change in CD4+ percent from baseline between the two treatment monitoring models over the duration of the study
  
• comparison of HIV resistance mutations in adults and children who have demonstrated virologic failure between the doctor and the primary health care sister-based monitoring arms
  
• comparison of adherence to antiretroviral treatment in Phase 1 in both adults and children as measured by pill count (or weight or volume of drug solution) between the two primary health care monitoring models
  
• comparison of adherence to antiretroviral treatment in adults as measured by pill count between those randomized to community-based DOT and those randomized to continued clinic-based adherence support
  
• comparison of HIV disease progression/death in adults and children between the two primary health care monitoring arms
  
• total and incremental costs of adding the 4 approaches for the provision of antiretrovirals to primary health care services in each study site
  
• potential costs to the government of provincial and national based scale-up for the provision of antiretroviral therapy in the short, long and medium term and resulting net costs/savings of each approach as a national strategy
  
• relative cost effectiveness of the introduction of the four alternative models of HAART provision versus no provision, using life-years gained and quality adjusted life-years gained
  
• household preferences for various aspects of antiretroviral therapy provision program design, including assessing their relationship with adherence
  
• personnel required for a nurse-based monitoring program, including the training process and implementation
  

The benefit of antiretroviral therapy is well established but limited to wealthy nations. A predefined, simple sequence of treatment regimens focused on extending the durability of limited treatment options has the best potential to be implemented in resource poor countries. South Africa has 15% of the world's HIV/AIDS patients and a limited number of physicians to treat them (l per 1,600 and less than 5 infectious diseases specialists). HIV patient care in the primary care setting must therefore be delivered by personnel other than doctors. Further, treatment strategies should include entire households to ensure maximum adherence and minimize sharing of drugs.

This study will have two parts. The first part will compare a first-line antiretroviral therapy regimen administered and monitored by primary health care sisters (nurses) with the same regimen administered by doctors. The second part of the study will determine if community-based directly observed therapy (DOT) is significantly superior to continued clinic-based treatment support for patients who have failed first-line therapy, as measured by cumulative virology failure rate. The project will also evaluate the cost and economic impact of a predetermined schedule of antiretroviral therapy; treatment outcomes in terms of morbidity, opportunistic and endemic infections, and mortality; and factors contributing to treatment failure, including toxicity, resistance, compliance, and treatment interruption.

In Part 1, households will be randomly assigned to receive first-line antiretroviral therapy under the monitoring and care of either an HIV-trained medical doctor supported by adherence counselors or an HIV-trained primary health care sister (nurse with training in diagnosis and treatment prescription). Members of the household who are HIV infected will receive stavudine, lamivudine, and efavirenz (nevirapine or nelfinavir may be used for special populations).

Participants who fail first-line antiretroviral therapy in Part 1 of the study will be entered into Part 2 of the study. Participants in Part 2 will receive zidovudine, didanosine, and lopinavir/ritonavir. Participants will be randomly assigned to have their treatment monitored through either a clinic-based treatment support group or through community-based directly observed treatment (DOT). For the DOT arm, a community member will observe therapy for at least one dose a day, five days a week, at the home or work of the participant.

HIV infected children age 3 months to 16 years who live in a participating household will also be included in the study. These children will receive first-line treatment with clinic visits monitored by either the assigned sister (nurse) or doctor along with their households. In Part 2, children will be provided with a second-line treatment regimen with continued daily monitoring of doses in the household.

The study will last 5 years.