If there were multiple valid results at a given visit, the first value at each visit window was used, unless a test was repeated and the result suggested that the initial value was an error, in which case the last observation within the visit window was used.
The expected visit window for the monthly pharmacy refill was 7 days early or 14 days late in the 30-day study month.
The visit window was included to allow some flexibility to the study participants and the time frame chosen was based upon practical considerations specific to the study area.
A control was defined as 1) a participant who continued to be on study medication and 2) last seen within the expected visit window.
Subjects were followed prospectively with study visits at weeks 4, 8, 12, 16, 24, 32, 40, and 48, and at time of suspected IRIS, if outside the scheduled visit window.
Data from a given participant were excluded if they met any of the following criteria: average response to PHA <400 spot-forming cells (SFC /200,000 PBMC; average of negative control wells >20 SFC/200,000 PBMC; day 2 viability less than 66%; or blood obtained outside the allowable visit window.
We begin by dividing the follow-up time into windows as follows: First visit window Second visit window ⋮ Time following last visit window [ t 11, t 12 ) [ t 21, t 22 ) [ t J + 1, 1, ∞ ).
If controls cannot be matched to the 6-month visit window period then they will be matched to within 12 months of the visit window period.
The VOICE protocol defined a missed clinic visit as non-attendance within the visit window.
If there was more than one value in a visit window, the results were averaged.
