Holstad 2011.

Methods	Randomized controlled trial
Participants	The study location was 5 HIV clinical sites in a large south eastern metropolitan city, USA 104 participants were randomized to the intervention group and 103 participants were randomized to the control group The inclusion criteria were (1) HIV infected; (2) female by birth; (3) prescribed antiretroviral therapy; (4) 18 years of age or older; (5) English speaking; (6) mentally stable as determined by a screening assessment; (7) willing to participate by completing computerized assessments, use electronic drug monitoring (EDM) caps, be randomly assigned and participate in group sessions
Interventions	Intervention: KHARMA INTERVENTION The KHARMA intervention consisted of 8 group sessions using motivational interviewing delivered in a group format. It was designed to empower women to make decisions and develop strategies about taking ART as prescribed and consistently using risk reduction behaviors, such as condom use and to overcome resistance/ambivalence to both. The sessions lasted about 1.5 to 2 hours, and were led by trained MI nurses. The first and last session focused on both adherence and risk behavior, 3 sessions were devoted to adherence only and 3 to risk reduction behaviors only, including a session on disclosure which is important to risk reduction as well as adherence. Each session included a discussion of goals and goal setting related to the topic. MI techniques were incorporated into every session. In keeping with the autonomy support spirit of MI, participants as a group chose which topic they wanted to address first. The majority of groups (n = 14) chose medication adherence as the first topic Control: HEALTH PROMOTION PROGRAM The control group sessions were equivalent in length and time to the MI group and were led by trained nurses and a health educator. This group used health education techniques of lecture/discussion/educational games and focused on nutrition, exercise, stress recognition, and women's health issues tailored to the HIV positive woman. Participants received a manual containing content and supplementary materials for each session. Adherence and risk reduction behaviours (RRB) were not addressed in the health promotion program (HPP) and facilitators were instructed to redirect the group if these issues came up
Outcomes	The measures of adherence were Electronic Drug Monitoring (EDM). One medication from the regimen was electronically monitored using the MEMS® cap and an algorithm was used to determine which medication would receive the cap. In general, that medication was the primary protease inhibitor (PI), or the non‐nucleoside reverse transcriptase inhibitor (NNRTI) if a PI was not prescribed. A sticker was placed on the monitored medication bottle and the participant was instructed to write dates and times on the sticker when the medication was pocketed, or if the cap was opened by mistake. At each download, information on this sticker was reviewed and an EDM questionnaire was completed. The information collected on the questionnaire included changes in medication regimen, problems with the cap, if someone else had administered the medication, and if the medication had been stopped. If the medication had been stopped, the participant was asked who stopped it and the reason, and if it had been restarted. If the participant was taken off the medication and brought the cap to the study office, it was stored until the participant re‐started medication and those days were marked as 'non‐monitored.' EDM data were adjusted by setting certain days as 'non‐monitored' on which the following events occurred: cap malfunction, lost cap, medication stopped by healthcare provider, someone else administered the medication (e.g. hospital, group home), incarceration, pocketing pills, reported exclusive use of pill box, and excessive openings (a form of cap malfunction defined as more than twice the dose plus one). The EDM data were captured over time in phases consistent with the 5 assessment periods. Data from the MEMS report on the Percentage (%) of Prescribed Doses Taken and Percentage (%) of Prescribed Doses Taken on Schedule were used for this analysis. The MEMS Track® caps were used by participants during the entire study period (about 13 months) including 2 weeks prior to the baseline to obtain an estimate of baseline adherence The patient outcomes were CD4 counts and viral loads. Data were extracted from participants' medical records during the time they were enrolled in the study. Dates that the laboratory tests were drawn were grouped according to proximity with the study assessment periods. For example, laboratory tests drawn as close as possible and prior to or on the day of the baseline assessment were classified as the baseline CD4 and viral load results. The median was about 27 days prior to the baseline. Participants did not always have both tests performed on the same date and therefore may have only had one test result available during an assessment period. The standard of care is to order these tests every 3 to 4 months once a patient has stabilized. Consequently, by the end of the follow‐up period, participants had fewer laboratory tests available that corresponded to study time points
Notes	Low attendance rates to the intervention. Also, not all the participants' viral loads were calculated by the same method
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	(pg 3) The study was a randomized controlled behavioral clinical trial in which participants were randomized within sites to the intervention or control condition using a computer‐generated randomization scheme
Allocation concealment (selection bias)	Unclear risk	No information given on allocation concealment
Selective reporting (reporting bias)	Low risk	Author's note: "We report on page 7: CESD scores were significantly but mildly correlated (r < 0.2) with adherence measures, but not significantly different between treatment groups at baseline, thus we chose not to control for depressive symptoms"
Other bias	High risk	From discussion (pg 10): "In 2006, our main recruitment site implemented an ongoing "Prevention for Positives" initiative where all providers were required to elicit and document discussion about risky behaviors with each patient every 12 months. When indicated, patients were referred for risk reduction education and counseling. Our results may be impacted by the addition of this program. Given this occurred for women in both groups, it might have diluted between group differences. In addition, risk behaviors in this study were measured by self‐report, and could be subject to bias from poor recall or social desirability. Due to cost constraints, laboratory results were extracted from medical records and dates of lab tests did not always correspond in time to follow‐up assessments. As the study progressed, there were fewer lab data for the 9 month assessment period. In the future blood draws consistent with the follow‐up periods are recommended."
Blinding of outcome assessment (detection bias) Adherence measure	Low risk	(PRIMARY) MEMS ‐ MEMS not likely to be affected by lack of blinding of outcome assessors
Blinding of outcome assessment (detection bias) Patient outcome	Unclear risk	(PRIMARY) CD4 COUNT AND VIRAL LOAD ‐ Unclear if extractor was blind. "Because of their high cost, we used data extracted from participants' medical records during the time they were enrolled in the study. Dates that the laboratory tests were drawn were grouped according to proximity with the study assessment periods. For example, labs drawn as close as possible and prior to or on the day of the baseline assessment were classified as the baseline CD4 and viral load results."
Blinding of participants (performance bias) Adherence measure	High risk	(PRIMARY) MEMS ‐ Author's note: patients were not blind. Also, the article suggests patients may have taken medication out of MEMS bottle to use other devices, and were then asked to remember to open the bottle when they took medication. (pg 10) "We monitored adherence primarily using the MEMS® Track Cap for electronic drug monitoring. Participants were asked to use this cap during the 13 months while enrolled in the study. The cap is cumbersome to use, and affects the portability of one's medications. Therefore, participants may have skipped using their caps for periods of time during the study. In addition, caps can malfunction, rendering data unavailable. Since many patients use pill boxes or pill trays to facilitate their adherence, it is difficult to remember to open the cap when removing pills from the pill box. For these reasons, future researchers who choose to use EDM might consider shortening the time for EDM, or using newer more user‐friendly technology available to conduct EDM."
Blinding of participants (performance bias) Patient outcome	Low risk	(PRIMARY) CD4 COUNT AND VIRAL LOAD ‐ Lack of blinding of the patients not likely to affect this outcome
Blinding of personnel (performance bias) Adherence measure	High risk	(PRIMARY) MEMS ‐ Author's note: key study personnel were not blinded to the study groups. Also, the article outlines complicated method for determining adherence based on MEMS results. "Electronic Drug Monitoring (EDM)—One medication from the regimen was electronically monitored using the MEMS® cap and an algorithm was used to determine which medication would receive the cap. In general, that medication was the primary protease inhibitor (PI), or the non‐nucleoside reverse transcriptase inhibitor (NNRTI) if a PI was not prescribed. A sticker was placed on the monitored medication bottle and the participant was instructed to write dates and times on the sticker when the medication was pocketed, or if the cap was opened by mistake. At each download, information on this sticker was reviewed and an EDM questionnaire was completed. The information collected on the questionnaire included changes in medication regimen, problems with the cap, if someone else had administered the medication, and if the medication had been stopped. If the medication had been stopped, the participant was asked who stopped it and the reason, and if it had been restarted. If the participant was taken off the medication and brought the cap to the study office, it was stored until the participant re‐started medication and those days were marked as 'non‐monitored.' EDM data were adjusted by setting certain days as 'non‐monitored' on which the following events occurred: cap malfunction, lost cap, medication stopped by healthcare provider, someone else administered the medication (e.g., hospital, group home), incarceration, pocketing pills, reported exclusive use of pill box, and excessive openings (a form of cap malfunction defined as more than twice the dose plus one). The EDM data were captured over time in phases consistent with the five assessment periods. Data from the MEMS report on the Percentage (%) of Prescribed Doses Taken and Percentage (%) of Prescribed Doses Taken on Schedule were used for this analysis. The MEMS Track® caps were used by participants during the entire study period (about 13 months) including 2 weeks prior to the baseline to obtain an estimate of baseline adherence."
Blinding of personnel (performance bias) Patient outcome	Low risk	(PRIMARY) CD4 COUNT AND VIRAL LOAD ‐ Unclear if key study staff were the extractors and if they were blind. "Because of their high cost, we used data extracted from participants' medical records during the time they were enrolled in the study. Dates that the laboratory tests were drawn were grouped according to proximity with the study assessment periods. For example, labs drawn as close as possible and prior to or on the day of the baseline assessment were classified as the baseline CD4 and viral load results."
Incomplete outcome data (attrition bias) Adherence measure	Unclear risk	(PRIMARY) MEMS ‐ Withdrawal fairly balanced, but wide variation in attendance. However, results also have been reported by taking only the participants with high attendance rates
Incomplete outcome data (attrition bias) Patient outcome	Unclear risk	(PRIMARY) CD4 COUNT AND VIRAL LOAD ‐ Withdrawal fairly balanced, but wide variation in attendance. However, results also have been reported by taking only the participants with high attendance rates