Continued attendance in a PrEP program despite low adherence and non-protective drug levels among adolescent girls and young women in Kenya: Results from a prospective cohort study

Background In sub-Saharan Africa (SSA), adolescent girls and young women (AGYW) ages 15 to 24 years represent <10% of the population yet account for 1 in 5 new HIV infections. Although oral pre-exposure prophylaxis (PrEP) with tenofovir disoproxil fumarate/emtricitabine (TDF/FTC) can be highly effective, low persistence in PrEP programs and poor adherence have limited its ability to reduce HIV incidence among women. Methods and findings A total of 336 AGYW participating in the PEPFAR-funded DREAMS PrEP program in western Kenya were enrolled into a study of PrEP use conducted between 6/2019 to 1/2020. AGYW, who used daily oral TDF/FTC, completed interviews and provided dried blood spots (DBS) for measurement of tenofovir-diphosphate (TFV-DP) concentrations at enrollment and 3 months later, and 176/302 (58.3%, 95% confidence interval [95% CI 52.3 to 63.8]) met our definition of PrEP persistence: having expressed intention to use PrEP and attended both the second interview and an interim refill visit. Among AGYW with DBS taken at the second interview, only 9/197 (4.6%, [95% CI 1.6 to 7.5]) had protective TFV-DP levels (≥700 fmol/punch) and 163/197 (82.7%, [95% CI 77.5 to 88]) had levels consistent with no recent PrEP use (<10 fmol/punch). Perception of being at moderate-to-high risk for HIV if not taking PrEP was associated with persistence (adjusted odds ratio, 10.17 [95% CI 5.14 to 20.13], p < 0.001) in a model accounting for county of residence and variables that had p-value <0.1 in unadjusted analysis (age, being in school, initiated PrEP 2 to 3 months before the first interview, still active in DREAMS, having children, having multiple sex partners, partner aware of PrEP use, partner very supportive of PrEP use, partner has other partners, AGYW believes that a partner puts her at risk, male condom use, injectable contraceptive use, and implant contraceptive use). Among AGYW who reported continuing PrEP, >90% indicated they were using PrEP to prevent HIV, although almost all had non-protective TFV-DP levels. Limitations included short study duration and inclusion of only DREAMS participants. Conclusions Many AGYW persisted in the PrEP program without taking PrEP frequently enough to receive benefit. Notably, AGYW who persisted had a higher self-perceived risk of HIV infection. These AGYW may be optimal candidates for long-acting PrEP.


Introduction
Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 6 Introduction Paragraph 3 Importantly, many women in clinical trials of daily oral PrEP in SSA continued to attend study visits but did not adhere to daily dosing.
Objectives 3 State specific objectives, including any prespecified hypotheses 7, 8 Introduction Paragraph 6 Methods Paragraph 3 The main questions examined were the extent of program persistence (a measure focusing on visit attendance), and the level of protection from HIV infection among program attendees as assessed by measuring PrEP metabolites in blood.
identification of factors associated with PrEP persistence and adherence.

Study design 4
Present key elements of study design early in the paper 7 Methods paragraph 1 A prospective study was conducted from 6/2019 -1/2020 among AGYW participants in the DREAMS PrEP program. This study included face-to-face interviews at enrolment and 3 months later, and DBS sample collection for assays of intracellular drug metabolites (tenofovir diphosphate, TVF-DP, as an objective measure of PrEP adherence) from AGYW who reported PrEP use at the time of interview.
Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data collection 7, methods paragraph 1 conducted from 6/2019 -1/2020 residing in Kisumu and Homa Bay counties Participants 6 (a) Cohort study-Give the eligibility criteria, and the sources and methods of selection of participants. Describe methods of follow-up Case-control study-Give the eligibility criteria, and the sources and methods of case ascertainment and control selection. Give the rationale for the choice of cases and controls Cross-sectional study-Give the eligibility criteria, and the sources and methods of selection of participants 7-8 Methods paragraph 2 From the DREAMS database, we identified 613 AGYW who met our eligibility criteria of 18-to-24 years old, residing in Kisumu and Homa Bay counties, enrolled in the PrEP program for 2-9 months with a visit for PrEP initiation or refill between October 2018 and April 2019, and had returned for a refill in the two months prior to study start in June 2019 (to exclude AGYW who might have recently stopped using PrEP). From the 613 eligible AGYW, we randomly sampled (using the 'sample' command in STATA) to select and enroll 359 AGYW who met the above eligibility criteria, were able to speak English, Dholuo, or Kiswahili and had indicated on the DREAMS enrollment consent form willingness to be contacted for future studies.
(b) Cohort study-For matched studies, give matching criteria and number of exposed and unexposed . We included cut-offs at 350 fmol/punch, (~one to two days per week) as well as the lower limit of quantification (200 fmol/punch) and the lower limit of detection (10 fmol/punch); levels below 10 were taken as consistent with no PrEP use in the recent past.
We conducted one-on-one in-person structured interviews with all participants in their preferred language, entering data electronically via Open Data Kit (ODK). The instrument captured information on: 1) socio-demographic characteristics (including age, marital status, current school attendance, living condition), 2) participation in and perceptions of the DREAMS program (including being in a PrEP support group, being active in the DREAMS program), 3) experience with PrEP and support for PrEP use among partners, family members and the community, 4) self-reported continued use of and adherence to PrEP, 5) HIV risk perception, 6) contraceptive use, 7) alcohol use (AUDIT17), 8) intimate partner violence (HITS, 4 items19), 9) social support (MOS Social Support Scale, 19 items) 20, 10) depression (PHQ-9, 10 items 21) and used scales adapted for previously used in Kenya or other African countries (ref 17, 19, 20, 21).
In addition to the study data collected at the two interviews, pharmacy PrEP refill information was obtained from the DREAMS program data. Data sources/ measurement 8* For each variable of interest, give sources of data and details of methods of assessment (measurement). Describe comparability of assessment methods if there is more than one group

8,9
Methods paragraph 5 The instrument captured information on: 1) socio-demographic characteristics (including age, marital status, current school attendance, living condition), 2) participation in and perceptions of the DREAMS program (including being active in the DREAMS program), 3) experience with PrEP and support for PrEP use among partners, family members and the community (e.g. being in a PrEP support group, have told partner, family members and friends of PrEP use, partner support of PrEP use, have friends on PrEP) 4) self-reported continued use of and Methods paragraph 7 Methods paragraph 8 adherence to PrEP (e.g. number of PrEP pills taken in the past week or month), 5) HIV risk perception (e.g. feeling that partner behavior put AGYW at risk, perception of being at moderate to high HIV risk if not taking PrEP), 6) condoms use and contraceptive use, 7) alcohol use (AUDIT17), 8) intimate partner violence (HITS, 4 items19), 9) social support (MOS Social Support Scale, 19 items) 20, 10) depression (PHQ-9, 10 items 21) and used scales previously used in Kenya or other African countries (ref 17, 19, 20, 21).
TFV-DP was analyzed using a validated liquid chromatography mass spectrometry (LC-MS/MS) with calibration curve range of 200-10,000 fmol/3mm punch. Internal/external quality control samples were included in each run, with external samples cross-validated with intra-laboratory testing. In addition to the study data collected at the two interviews, pharmacy PrEP refill information was obtained from the DREAMS program data.
The main outcome of interest in this study is PrEP persistence. Persistent AGYW were defined as having attended both interviews as well as interim visits for PrEP refills and stating that they were taking PrEP at both interviews. A secondary outcome is PrEP adherence based on the TFV-DP levels from the testing of the DBS samples. Adherent AGYW were defined as having TFV-DP levels of 700+ fmol/punch, the equivalent of 4+ doses per week taken regularly22 Bias 9 Describe any efforts to address potential sources of bias 9 Methods Paragraph 9 We used univariable and multivariable (adjusted) generalized estimating equations modification of logistic regression models accounting for clustering of AGYW within wards to assess the associations between persistence and factors measured at Interview 2. Study size 10 Explain how the study size was arrived at 8 Methods Paragraph 3 This sample size assumed a discontinuation or loss to follow-up rate of 50%, and a two-level factor potentially associated with PrEP adherence. Our sample size was sufficient to detect a minimum absolute difference of 15% in the PrEP adherence rates at the first interview between the levels of the factor at 80% power given a type I error rate of 5% and a two-sided test.

Quantitative variables 11
Explain how quantitative variables were handled in the analyses. If applicable, describe which groupings were chosen and why 9 Methods Paragraph 9 We summarized categorical variables as frequency and percentages and continuous variables as mean and standard deviation (SD) or median and interquartile range (IQR). We used univariable and multivariable (adjusted) generalized estimating equations modification of logistic regression models accounting for clustering of AGYW within wards to assess the associations between persistence and characteristics factors measured at Interview 2.
Statistical methods

12
(a) Describe all statistical methods, including those used to control for confounding 9 Methods Paragraph 9 We used univariable and multivariable (adjusted) generalized estimating equations modification of logistic regression models accounting for clustering of AGYW within wards to assess the associations between persistence and factors measured at Interview 2. The multivariable models adjusted for county of residence and factors that had a p-value<0.1 in the univariable analysis.
(b) Describe any methods used to examine subgroups and interactions (c) Explain how missing data were addressed 9 Methods Paragraph 9 The proportion of AGYW who attended Interview 2 and had missing information on factors of interest was <10%. The analyses were based on complete case data assuming missing completely at random for the missing data mechanism.
(d) Cohort study-If applicable, explain how loss to follow-up was addressed

Case-control study-If applicable, explain how matching of cases and controls was addressed
Cross-sectional study-If applicable, describe analytical methods taking account of sampling strategy 9 Methods Paragraph 9 AGYW who were lost to follow up or censored due to study closure prior to Interview 2 were excluded from the analysis (e) Describe any sensitivity analyses

Results
Participants 13* (a) Report numbers of individuals at each stage of study-e.g., numbers potentially eligible, examined for eligibility, confirmed eligible, included in the study, completing follow-up, and analysed (10) Figure 1 (b) Give reasons for non-participation at each stage (c) Consider use of a flow diagram (10) Figure 1 Descriptive data 14* (a) Give characteristics of study participants (e.g., demographic, clinical, social) and information on exposures and potential confounders (10) In conclusion, our study provides insight into oral PrEP use in a real-world programmatic setting, as well as predictors of PrEP persistence in this context. The results reveal that most AGYW may be better protected by long-acting injectable PrEP than by oral daily PrEP. Future research is needed to clarify whether persistence without adequate adherence is as common among AGYW in other settings, and whether new long-acting PrEP formulations, adequately supported by facilitators identified during use of existing PrEP agents, can afford higher level protection to this very vulnerable population.
Generalisability 21 Discuss the generalisability (external validity) of the study results

Discussion
Paragraph 5 Limitations of this study include both size and generalizability, as we included a relatively small sample size from a large programmatic project, which may be most generalizable to AGYW in western Kenya.

Other information
Funding 22 Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based 15 The study is supported by NIH grants R01HD094682 and 3R01HD094682-02S1. Funders were not involved in study design, execution, or interpretation of results.
*Give information separately for cases and controls in case-control studies and, if applicable, for exposed and unexposed groups in cohort and cross-sectional studies.
Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is available at www.strobe-statement.org