Sample sizes for each objective of the Adherence Guideline impact evaluation study in South Africa
| Objective | Sample size | Rationale |
| Objective 1—fast track ART initiation counselling | 720 patients | The RapIT study of rapid ART initiation,16 conducted at a well-managed PHC in Gauteng Province, found that about 60% of ART-eligible patients initiated under standard care within 30 days. Conservatively assuming 60% initiation without the intervention and 75% with the intervention, 30 subjects in each of the 24 clusters for 720 total subjects will be required to detect a difference of 15%. We have increased this by 20% to account for ineligible patients. |
| Objective 2—adherence clubs | 576 patients | Data from Themba Lethu Clinic17–19 show that about 80% of patients made all of their medication pickups over a 3-month period. It is anticipated that 24 subjects per clinic for a total of 576 patients will be needed to detect a difference of 15%. We have increased this by 20% to account for ineligible patients. |
| Objective 3—decentralised medication delivery | 576 patients | Data from Themba Lethu Clinic17–19 show about 80% of patients made all of their medication pickups over a 3-month period. It is anticipated that 24 subjects per clinic for a total of 576 patients will be needed to detect a difference of 15%. We have increased this by 20% to account for ineligible patients. |
| Objective 4—enhanced adherence counselling | 1008 patients | Data from KwaZulu-Natal Province indicate that 52% of patients with a detectable viral load resuppress after one session. It is anticipated that 42 subjects per clinic for a total of 1008 patients will be needed to detect a difference of 15%. We have increased this by 20% to account for ineligible patients. |
| Objective 5—early tracing of patients lost to follow-up | 576 patients | Data from various Right to Care clinics19 suggest that the proportion of patients who are lost from care who return with no or little intervention is low, between 20% and 35%. It is anticipated that 24 subjects per clinic for a total of 576 patients will be needed to detect a difference of 15% assuming a baseline of 30% loss to follow-up without intervention. We have increased this by 20% to account for ineligible patients. |
ART, antiretroviral therapy; PHC, primary healthcare clinic; RapIT, Rapid Initiation of Treatment Trial.