Ial, they had the most participants enrolled in FEM-PrEP, and they were the sites with the highest number of participants who seroconverted. As described in more detail elsewhere [14], participants were selected from a list of former FEM-PrEP participants who had previously given permission to be contacted for future research or through community-wide promotion of the study; all had previously had their blood specimens analyzed for drug concentrations s11606-015-3271-0 as part of other FEM-PrEP analyses. Based on their adherence drug concentrations, we placed the participants into three interview groups: “high,” “moderate,” and “none/scarce.” We did not require a minimum number of specimens for purchase Pedalitin permethyl ether inclusion in a group. Rather, we chose to consider all available specimens (up to 13 per participant) when deciding in which adherence group to place participants, as we believed this would provide for a richer discussion of adherence. Because of the early closure of the trial, however, many participants were unable to complete 52 weeks on the study drug and therefore had fewer specimens available for testing of drug concentrations. Moreover, aTable 1. Qualitative adherence composite scores, corresponding TFV and TFV-DP concentrations, and estimated doses per interval [7]. Adherence composite score 0 1 2 3 TFV in plasma and TFV-DP in upper fpsyg.2014.00822 layer packed cells No detectable TFV and <10,000 fmol/mL TFV-DP Detectable TFV but <10,000 fmol/mL TFV-DP 10,000?00,000 fmol/mL TFV-DP, regardless of TFV <10 ng/ml TFV and >100,000 fmol/mL TFV-DP >10 ng/ml TFV and 100,000?1,000,000 fmol/mLTFV-DP >10 ng/ml TFV and >1,000,000 fmol/mL TFV-DP Estimated doses per interval A low number of doses or no doses at all in the interval A few doses in the entire interval 1? doses per week Several doses early in the interval, followed by a stop in the week or two leading up to the sampling visit 4? doses per week Approximately daily dosing4doi:10.1371/journal.pone.0125458.tPLOS ONE | DOI:10.1371/journal.pone.0125458 April 13,3 /Facilitators of Study Pill Adherence in FEM-PrEPcomposite adherence score was not calculated for a study visit interval when a participant missed a study visit, when the participant did not have study pills available to use (e.g., because of a protocol-defined product withdrawal or a missed pill supply visit), or when there was no or insufficient specimen for analysis [7]. Consequently, some participants had composite adherence scores for all 13 study visit intervals, but many had fewer scores available. We therefore categorized participants based on a combination of the composite adherence score for each study visit interval and the number of available specimens. Participants who had composite adherence scores of 4 or 5 for the majority, if not all, of their available samples were identified for recruitment into the high adherence group. For the moderate adherence group, we identified participants for recruitment who had composite adherence scores that fluctuated over time among the available samples, remained steady between scores 2 and 4, or were PP58 web higher at the beginning of the trial and lower as time passed. Participants were identified for the none/scarce group if the majority of their recorded measurements had scores of 0. We ranked the order in which to recruit women based on the number of specimens available (i.e., women who had a higher number of recorded specimen measurements were placed higher on the recruitment list for each group). We aimed to interview appr.Ial, they had the most participants enrolled in FEM-PrEP, and they were the sites with the highest number of participants who seroconverted. As described in more detail elsewhere [14], participants were selected from a list of former FEM-PrEP participants who had previously given permission to be contacted for future research or through community-wide promotion of the study; all had previously had their blood specimens analyzed for drug concentrations s11606-015-3271-0 as part of other FEM-PrEP analyses. Based on their adherence drug concentrations, we placed the participants into three interview groups: “high,” “moderate,” and “none/scarce.” We did not require a minimum number of specimens for inclusion in a group. Rather, we chose to consider all available specimens (up to 13 per participant) when deciding in which adherence group to place participants, as we believed this would provide for a richer discussion of adherence. Because of the early closure of the trial, however, many participants were unable to complete 52 weeks on the study drug and therefore had fewer specimens available for testing of drug concentrations. Moreover, aTable 1. Qualitative adherence composite scores, corresponding TFV and TFV-DP concentrations, and estimated doses per interval [7]. Adherence composite score 0 1 2 3 TFV in plasma and TFV-DP in upper fpsyg.2014.00822 layer packed cells No detectable TFV and <10,000 fmol/mL TFV-DP Detectable TFV but <10,000 fmol/mL TFV-DP 10,000?00,000 fmol/mL TFV-DP, regardless of TFV <10 ng/ml TFV and >100,000 fmol/mL TFV-DP >10 ng/ml TFV and 100,000?1,000,000 fmol/mLTFV-DP >10 ng/ml TFV and >1,000,000 fmol/mL TFV-DP Estimated doses per interval A low number of doses or no doses at all in the interval A few doses in the entire interval 1? doses per week Several doses early in the interval, followed by a stop in the week or two leading up to the sampling visit 4? doses per week Approximately daily dosing4doi:10.1371/journal.pone.0125458.tPLOS ONE | DOI:10.1371/journal.pone.0125458 April 13,3 /Facilitators of Study Pill Adherence in FEM-PrEPcomposite adherence score was not calculated for a study visit interval when a participant missed a study visit, when the participant did not have study pills available to use (e.g., because of a protocol-defined product withdrawal or a missed pill supply visit), or when there was no or insufficient specimen for analysis [7]. Consequently, some participants had composite adherence scores for all 13 study visit intervals, but many had fewer scores available. We therefore categorized participants based on a combination of the composite adherence score for each study visit interval and the number of available specimens. Participants who had composite adherence scores of 4 or 5 for the majority, if not all, of their available samples were identified for recruitment into the high adherence group. For the moderate adherence group, we identified participants for recruitment who had composite adherence scores that fluctuated over time among the available samples, remained steady between scores 2 and 4, or were higher at the beginning of the trial and lower as time passed. Participants were identified for the none/scarce group if the majority of their recorded measurements had scores of 0. We ranked the order in which to recruit women based on the number of specimens available (i.e., women who had a higher number of recorded specimen measurements were placed higher on the recruitment list for each group). We aimed to interview appr.