O5.3 - Concordance between self-reported PrEP use using a diary app and intracellular tenofovir diphosphate in the Amsterdam PrEP demonstration project

Background:
HIV pre-exposure prophylaxis (PrEP), taken as a daily or event-driven regimen, is highly effective in preventing HIV acquisition when users adhere to the chosen regimen. Adherence can be measured through intracellular concentrations of tenofovir diphosphate (TFV-DP) from dried blood spots (DBS). Tracking adherence through a diary is less costly, but self-report may be subject to social desirability bias. We examine the concordance between self-reported pill intake in a diary app and intracellular TFV-DP concentrations.

Methods:
AMPrEP was a prospective demonstration study providing PrEP to men who have sex with men (MSM) and transgender women in Amsterdam, the Netherlands (2015-2020). Participants could opt to register PrEP intake in a mobile diary app. DBS were taken after 6,12,24, and 48 months. We included TFV-DP measurements in analysis if DBS sampling was preceded by regular diary completion (i.e., >90% of days in the prior six weeks). We examined the association between the natural log of self-reported pill intake during the six weeks before DBS and log-transformed TFV-DP using a linear mixed model. Furthermore, we examined the association between four-category pill intake (< 2, 2-3, 4-6 or 7 pills per week on average) and four-category TFV-DP concentration (< 350, 350-699, 700-1249 or ≥1250 fmol/punch), using linear regression.

Results:
In total, 1230 DBS samples were available. We excluded 333 samples without any diary data, and 138 samples preceded by < 90% diary completion. Finally, 759 DBS measurements (n=600 daily use periods; n=159 event-driven use periods) from 261 individuals (all MSM) were included. Median self-reported PrEP use during the six weeks before DBS was 41 pills (IQR 32-42), and median TFV-DP was 1158 fmol/punch (IQR 645-1496). Self-reported PrEP use was strongly and positively associated with TFV-DP concentration (p< 0.0001). Furthermore, categorized pill intake was associated with TFV-DP category (β=0.6;p< 0.0001) (Figure).

Conclusion:
Self-reported PrEP use in a diary app was strongly positively associated with intracellular tenofovir levels. Still, agreement between categories of self-reported pills taken and TFV-DP was not perfect. As such, pill diaries can be used to compare PrEP adherence between groups in research settings, but should be used with caution when assessing an individual‘s adherence within clinical settings.