Adherence Measurement using Unannounced Pill Counts

We used unannounced Pill Counts to measure adherence to ART treatment. This measure was applied because unannounced Pill Counts conducted in patients’ homes have demonstrated validity for monitoring medication adherence [27]. Due to funding limitations, we were not able to conduct follow-up visits for pill counts for 247 participants. However, all these participants completed other study-related follow-up assessments and thus were not completely lost to follow-up. The participants without pill count data were distributed across 17 clinics, with the clinics contributing the fewest number of participants having 7 ALHIV, while the clinic with the highest number of participants had 34 ALHIV.

For the Suubi + Adherence study, we followed the same approach. Specifically, in the Suubi + Adherence study, the research team designed a protocol for unannounced pill counts which specified the following steps for the Research Assistant/Interviewer to follow: (1) ascertain each participant’s regimen before actual counting; (2) use a pharmacy pill tray, spatula, cup for excessive pills, and alcohol prep wipes: (3) on visiting the home, ask the participant bring out all medications in the home, including closed bottles, pillboxes, and doses kept in pockets, purses, backpacks, etc.; (4) on a flat surface, and in the presence of the participant, sort their medications into clusters; (5) report prescription numbers on each medication, refill date, and dispensed quantity; (6) ask the participant if they had lost or gained any pills since their previous count and whether they had taken any of the drugs that day; (7) in the presence of participant, count pills using the study-provided pharmacist tray and cup; (8) repeat the procedure to double count. Moreover, since the procedure was based on prescription data, it took medication changes into account. To ensure issues of confidentiality are maintained, the team visits participants at an undisclosed time and day to count their pills once every 1–3 months throughout the study. Adherence was calculated as the difference between the current and previous pill count divided by the prescribed number of doses for the same period.

While calculating the number of pills taken, we adjusted for pharmacy refills during the review period. Although most participants took more than one type of medicine, for analysis purposes, only one drug was used to determine adherence. We used a non-nucleoside reverse transcriptase inhibitor (NNRTI), including Nevirapine or Efavirenz for participants of first-line treatment, or a protease inhibitor such as Ritonavir/Lopinavir for patients on second-line therapy. Previous studies have used an NNRTI drug selection to calculate pill count adherence [28, 29]. As applied in similar studies, we adjusted for adherence greater than 100–100% [27]. After these adjustments, we calculated ART adherence with a possible range of 0 to 100%. Participants who took 85% or more of their medication were considered to have good adherence based on the Consolidated guidelines for the prevention and treatment of HIV in Uganda [26].

In addition, using a paper-based questionnaire administered during an interview that lasted about 60 min, we collected sociodemographic information such as age, sex, orphanhood status, and household size. Using a 20-item asset index, we assessed asset ownership, including houses, land, and other properties reported by participants’ families. Each owned item is coded as 1; otherwise, it’s 0. The scale’s range is 0 to 20, where higher values indicate greater asset ownership.