Secondary Logo

Journal Logo

Prevention Research

Patterns of PrEP Retention Among HIV Pre-exposure Prophylaxis Users in Baltimore City, Maryland

Wu, Linxuan MHSa; Schumacher, Christina PhDa; Chandran, Aruna MD, MPHb; Fields, Errol MD, PhD, MPHa; Price, Ashley MSc; Greenbaum, Adena MDc; Jennings, Jacky M. PhD, MPHa,b;  the IMPACT Partner Collaborative

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: December 15, 2020 - Volume 85 - Issue 5 - p 593-600
doi: 10.1097/QAI.0000000000002506
  • Free

Abstract

BACKGROUND

In the United States, approximately 40,000 individuals are newly HIV diagnosed annually,1 despite the availability of highly effective HIV prevention interventions. Pre-exposure prophylaxis (PrEP), the daily use of HIV antiretroviral therapy among HIV uninfected individuals, may reduce individual-level HIV acquisition risk by up to 99%.2–5 Several studies suggest that widespread PrEP implementation may reduce population-level HIV incidence throughout the United States by up to 31%.6–8 PrEP implementation to decrease population-level HIV incidence will depend on individuals initiating and continuing PrEP during at-risk periods for HIV acquisition (ie, PrEP retention). Maximizing PrEP implementation will require real-world studies on PrEP retention and factors associated with PrEP retention. There have been, however, a number of limitations in the current PrEP implementation studies to-date.

Current estimates of PrEP retention rates range from 15% to 92%.9–18 One reason for the observed variability in estimated PrEP retention rates is differences in study designs. The highest PrEP retention rates (69%–92%) were observed in open-label research studies and demonstration projects. These studies incentivized follow-up visit attendance,9 provided PrEP free of charge,9,15,16 and conducted intensive follow-up with patients to ensure retention.9,15,16 In contrast, substantially lower PrEP retention rates (15%–62%) were observed in 6 real-world clinical studies, which did not provide financial incentives or intensive follow-up. All but one17 of these studies primarily focuses on evaluating PrEP retention among gay, bisexual and other men who have sex with men (MSM) and transgender women, and did not include9–11,13–16 or only included a small proportion (5%–12%)12,18 of other priority populations for HIV transmission [ie, persons who inject drugs (PWID) or those at risk for HIV through heterosexual sex]. PrEP retention among non-MSM key populations remains understudied.

Another reason for observed variability in retention rates may be because of differences in PrEP retention measures. These measures include: number of PrEP discontinuations over a fixed time period,10 patterns of the timing of PrEP discontinuation (eg, early discontinuation, late discontinuation, and persistent PrEP use),12 visit constancy (proportion of time intervals with at least one completed clinical visit during a specified observation time),13,14 PrEP retention rates at a set time point after PrEP initiation (ie, cross-sectional),17 and time to PrEP discontinuation.11 However, some of these measurements, specifically, those which are cross-sectional and time-to-event, may not capture the natural variability in PrEP use over time.

The overall goal of this work was to inform real-world efforts to decrease HIV acquisition by supporting the implementation and delivery of PrEP in real-world clinical settings. We sought to surmount the limitations of previous work by evaluating PrEP retention patterns in a more heterogeneous population (ie, including heterosexuals) and by measuring PrEP retention at the visit- and individual-level using 3-month visit constancy. Visit constancy, a widely used retention measurement in HIV care research, may be a better approach for accounting for spacing between visits, is more suitable for long-term observation periods, and can provide the aggregated individual-level patterns to reflect the overall retention structure.19,20 The objectives were to: (1) describe visit-level patterns of PrEP retention, (2) describe individual-level patterns of PrEP retention, and (3) identify individual-level characteristics associated with low-level PrEP retention after PrEP initiation among HIV negative individuals receiving PrEP care from one of 7 clinical sites in Baltimore City, Maryland.

METHODS

Study Population

This analysis included information on individuals receiving PrEP care at one of 7 clinical sites collaborating with the Baltimore City Health Department on a program designed to increase PrEP provision. The 7 clinical sites located in Baltimore City, Maryland, included 2 city health department sexual health clinics, a federally qualified health center, 2 academic adolescent clinics, and 2 academic PrEP specialty clinics. Eligibility for this, PrEP program was based on the Centers for Disease Control and Prevention's PrEP guidelines19 and local HIV epidemiology. Eligible individuals were those who were: cisgender males who self-identified as MSM or who reported male sex partners in the past 12 months, transgender individuals, or any individual who reported any of the following: injection drug use (past 12 months); buying or selling sex (past 12 months); an HIV positive sex partner (past 12 months); >2 uses of nonoccupational postexposure prophylaxis (past 12 months) or diagnosis of syphilis, gonorrhea or rectal chlamydia infection (past 3 months). HIV negative individuals who initiated PrEP (received their first PrEP prescription between September 30, 2015 and February 28, 2018) were included and followed for 12 months after initiating PrEP. Censoring occurred 13 months after PrEP initiation or on the date of a positive HIV laboratory test, whichever occurred first. All information from PrEP initiation and clinical visits during the observation period were obtained through medical record abstraction using standardized forms. Information abstracted included demographics, sexual behaviors, sexually transmitted infection (STI) diagnoses, clinical visits, PrEP prescriptions, and PrEP use status. Individuals included in the analyses were those who had complete demographic information and had nonmissing information on sexual behaviors or STI diagnoses at PrEP initiation.

Measures

The primary outcomes of visit- and individual-level PrEP retention were based on clinical visit attendance at a collaborating clinical site during each quarterly interval (0–3, 4–6, 7–9 and 10–12 months; hereafter referred to as 3-, 6-, 9- and 12-month follow-up visits) after PrEP initiation. For each 3-month interval, a 1-month buffer was allowed to account for the fact that PrEP follow-up visits may not have occurred exactly 3 months after the prior visit.

Visit-level PrEP retention (yes/no) was assessed at each quarterly follow-up visit, and was defined as a clinical visit with documentation of PrEP use, including a PrEP prescription or a notation indicating individuals were currently on PrEP. Individuals without documentation of PrEP use at any visit or who did not present for a clinical visit were considered to have discontinued PrEP. Individual-level retention over the 12 months after PrEP initiation was dichotomized as high- and low-level retention. Similar to definitions of high-level retention in studies of HIV care retention,21,22 high-level PrEP retention was defined as having documented PrEP use at ≥3 quarterly follow-up visits. Low-level PrEP retention was defined as having documented PrEP use at <3 quarterly follow-up visits.

Demographics (age, gender, race/ethnicity), year of PrEP initiation, HIV transmission risk category, sexual behaviors, and STI diagnoses were assessed at PrEP initiation. Age was categorized as ≤24 years, 25–34 years, and ≥35 years. Gender was defined as cisgender male, cisgender female, and transgender. Race/ethnicity was defined as non-Hispanic/Latinx White, non-Hispanic/Latinx African American, Hispanic/Latinx, non-Hispanic/Latinx Other. Transmission risk was based on standard Centers for Disease Control and Prevention surveillance definitions and defined as MSM, transgender, heterosexual, and PWID.1 We examined whether PrEP retention was associated with sexual behaviors in the past 12 months (except where otherwise indicated) that have previously associated with HIV acquisition.23 Information on sexual behaviors was self-reported and included: sex partner living with HIV, anonymous sex partners (a sex partner for whom the individual has limited identifying information such as only knowing a geosocial app username or alias/nickname), and number of sex partners (past 3 months). Reported number of sex partners was dichotomized as above/below the median. STI diagnoses at PrEP initiation was defined as a laboratory confirmed diagnosis of syphilis (any stage), gonorrhea, or chlamydia (any anatomic site) ± 30 days of the PrEP initiation date. Because there was substantial missing information on sexual behaviors and STI status, we included a composite binary sexual behaviors/STI diagnoses variable to examine whether any factor previously associated with increased risk of HIV acquisition (reporting anonymous sex partners, sex partners living with HIV, ≥3 sex partners in the past 3 months or had a laboratory confirmed STI diagnosis ± 30 days of PrEP initiation) was associated with retention to include in multivariable models.

Statistical Analysis

Summary statistics were generated to describe characteristics at PrEP initiation and by individual-level PrEP retention. Regression models were used to assess characteristics associated with low-level PrEP retention. We accounted for the nested nature of the data (ie, individuals within clinics) using Poisson regression with cluster robust standard errors. Multivariable regression models included any variable significant (P < 0.1) in bivariate models. Because of the collinearity of gender and HIV transmission risk, 2 multivariable regression models were generated. Model 1 included age, year of PrEP initiation, and gender. Model 2 included age, year of PrEP initiation, and HIV transmission risk category. All analyses were conducted in Stata version 15.0 (Stata Corp., College Station, TX).

RESULTS

Between September 30, 2015 and February 28, 2018, 529 HIV negative individuals enrolled in the program, among whom 85.1% (n = 450) initiated PrEP. Among the 450 individuals initiating PrEP, 412 (91.6%) had complete demographic information and had nonmissing information on sexual behaviors or STI diagnoses and were included in the analytic sample. Compared with those who were included in the analytic sample, individuals who were excluded did not differ significantly by PrEP retention levels, age, race/ethnicity, year of PrEP initiation, and transmission risk (data not shown). However, compared with cisgender males, cisgender females were more likely to be excluded. Only one individual was censured before 12 months, after 171 days because of documentation of a positive HIV test.

Overall, the mean and median age of PrEP users at PrEP initiation was 31.8 (SD: 10.6) and 28.7 years (range: 16.1–69.2), respectively; over one-quarter (26.9%, n = 111) were adolescents/young adults (aged ≤ 24), and 1.2% (n = 5) were aged under 18 (Table 1). Most were cisgender male (83.7%, n = 345), non-Hispanic/Latinx Black (56.3%; n = 232), and MSM (74.0%; n = 305). About one-fifth (20.4%, n = 84) had heterosexual HIV transmission risk.

TABLE 1. - Characteristics of Individuals Prescribed HIV PrEP at PrEP Initiation, Baltimore City, Maryland, September 30, 2015 to February 28, 2018
N = 412
n %
Age
 ≤24 yrs 111 26.9
 25–34 yrs 184 44.7
 ≥35 yrs 117 28.4
Year of PrEP initiation
 2015 15 3.6
 2016 149 36.2
 2017 201 48.8
 2018 47 11.4
Gender identity
 Cisgender male 345 83.7
 Cisgender female 47 11.4
 Transgender 20 4.9
Race/ethnicity
 Non-Hispanic/Latinx White 135 32.8
 Non-Hispanic/Latinx Black 232 56.3
 Hispanic/Latinx 28 6.8
 Non-Hispanic/Latinx Other 17 4.1
Transmission risk
 Gay, bisexual and other MSM 305 74.0
 Transgender 20 4.9
 Heterosexual 84 20.4
 PWID 3 0.7
Clinic
 Publicly funded sexual health clinics 190 46.1
 Federally qualified health center 147 35.7
 Adolescent clinic 1 25 6.1
 Adolescent clinic 2 5 1.2
 Academic PrEP clinic 1 29 7.0
 Academic PrEP clinic 2 16 3.9
Sex partner living with HIV, past 12 mo (n = 387) 105 25.5
Anonymous sex partner, past 12 mo (n = 387) 183 44.4
Number of sex partners, past 3 mo (n = 296)
 0–2 partner(s) 135 32.8
 ≥3 partners 161 39.1
Buying or selling sex, past 12 mo (n = 387) 12 2.9
STI diagnoses* ± 30 d of PrEP initiation (n = 359) 85 20.6
Composite sexual behaviors/STI diagnosis variable 286 69.4
*Includes diagnoses of syphilis, gonorrhea or chlamydia at urogenital, rectal or oropharyngeal sites.
Individuals who reported sex partner living with HIV (past 12 months), anonymous sex partner (past 12 months), ≥3 sex partners (past 3 months) at PrEP initiation, or had laboratory confirmed syphilis, gonorrhea, or chlamydia diagnosis ± 30 days of PrEP initiation.

In the 12 months before PrEP initiation, 25.5% (n = 105) reported a sex partner living with HIV, and 44.4% (n = 183) reported anonymous sex partners. In the 3 months before PrEP initiation, the mean and median number of sex partners was 4.4 (SD: 5.3) and 3 (range: 0–50), respectively; 39.1% (n = 161) reported ≥3 sex partners. Very few individuals reported buying or selling sex in the 12 months before PrEP initiation (2.9%, n = 12), and this variable was not included in regression models. One quarter (20.6%, n = 85) were diagnosed with an STI ± 30 days of PrEP initiation. More than half (69.4%, n = 286) reported any of the following factors: sex partners living with HIV, anonymous sex partners, ≥3 sex partners or had an STI diagnosis ± 30 days of PrEP initiation.

Visit-Level Patterns of PrEP Retention

Figure 1 shows visit-level PrEP retention at each quarterly PrEP care follow-up visit 12 months after PrEP initiation. Three months after PrEP initiation, 69.4% (n = 286) had a clinical visit with documentation of PrEP use. At the 6-month visit, just over half (51.9%, n = 214) presented for a clinical visit and had documentation of PrEP use. At both 9- and 12-month quarterly visits, less than half (9-months: 44.5%, n = 183; 12-months: 41.1%, n = 169) presented for a visit and had documentation of PrEP use. At each quarterly visit, few participants (5.1%–7.0%) presented for a clinical visit and did not have documentation of PrEP use; the remaining individuals did not present for a clinical visit that quarter (3-months: 23.5%, n = 97; 6-months: 42.5%, n = 175; 9-months: 50.4%, n = 207; 12-months: 53.0%, n = 218).

FIGURE 1.
FIGURE 1.:
Visit-level PrEP retention 12 months after initiating HIV PrEP, Baltimore city, Maryland, September 30, 2015 to February 28, 2018.

Individual-Level Patterns of PrEP Retention

Figure 2 shows individual-level PrEP retention at each quarterly visit for 12 months after PrEP initiation. Only 19.7% (n = 81) of individuals had complete PrEP retention for a full year after initiating PrEP (defined as presenting for a clinical follow-up visit with documentation of PrEP use during each quarter). An additional 21.8% (n = 90) presented for a clinical follow-up visit with documentation of PrEP use at 3 of the 4 quarterly visits. Together, these 41.5% (n =171) of individuals were considered to have high-level PrEP retention. Fifty-nine percent (58.5%, n = 241) had low-level PrEP retention (defined as documentation of a PrEP prescription for <3 quarterly follow-up visits), among whom 33.6% (n = 81) had evidence of episodic use. Of those 81 individuals, 41 percent (n = 33) missed their 3-month visit and had evidence of PrEP use at only one other quarterly visit. Fifty-nine percent (n = 48) had evidence of PrEP use at 2 nonsequential quarterly visits throughout the year after PrEP initiation. Fifteen percent (15.0%, n = 62) did not have documentation of PrEP at any quarterly follow-up visit. An additional 16.3% (n = 67) only presented for a clinical visit with documentation of PrEP use at 3-month visit. Eight percent (7.5%, n = 31) had presented for clinical visits with documentation of PrEP use at the 3- and 6-months, but did not have documentation of PrEP use at 9- and 12-month visits.

FIGURE 2.
FIGURE 2.:
Individual-level PrEP retention 12 months after initiating HIV PrEP, Baltimore city, Maryland, September 30, 2015 to February 28, 2018. Each horizontal line represents an individual. aHigh-level PrEP retention was defined as ≥3 quarters with a PrEP prescription during the12 months after PrEP initiation. bLow-level PrEP retention was defined as <3 quarter(s) with a PrEP prescription during the12 months after PrEP initiation.

Characteristics Associated With Low-Level PrEP Retention

Table 2 shows individual characteristics associated with low-level PrEP retention. Race/ethnicity, sex partner living with HIV, anonymous sex partners, STI diagnoses, nor the composite sexual behaviors/STI diagnoses variable were not associated with low-level PrEP retention. In bivariate models, individuals aged 25–34 were less likely to have low-level PrEP retention; heterosexual (vs. MSM) and those who initiated PrEP in 2018 (vs. 2015) were slightly more likely to have low-level retention. However, these associations were not statistically significant in multivariable models. In bivariate models, reporting ≥3 sex partners (vs. 0–2 sex partners) [relative risk (RR): 0.85; 95% confidence interval: (0.78 to 0.93)] were associated with decreased risk of low-level PrEP retention.

TABLE 2. - Low-Level PrEP Retention* 12 Months After Initiating HIV PrEP, Baltimore City, Maryland, September 30, 2015 to February 28, 2018
Total, N = 412 Low-Level PrEP Retention Unadjusted Relative Risk Adjusted Relative Risk, Model 1 Adjusted Relative Risk, Model 2
n N Row% RR 95% CI aRR 95% CI aRR 95% CI
Overall 412 241 58.5
Age
 ≤24 yrs 111 68 61.3 Ref Ref Ref
 25–34 yrs 184 101 54.9 0.90 0.83 to 0.97 0.97 0.89 to 1.04 0.95 0.86 to 1.05
 ≥35 yrs 117 72 61.5 1.00 0.86 to 1.18 0.98 0.84 to 1.14 0.95 0.77 to 1.17
Year of PrEP initiation
 2015 15 8 53.3 Ref Ref Ref
 2016 149 70 47.0 0.88 0.59 to 1.30 0.85 0.59 to 1.22 0.84 0.57 to 1.22
 2017 201 129 64.2 1.20 0.85 to 1.70 1.16 0.82 to 1.64 1.14 0.77 to 1.68
 2018 47 34 72.3 1.36 1.02 to 1.80§ 1.30 0.97 to 1.75 1.28 0.92 to 1.76
Gender identity
 Cisgender male 345 190 55.1 Ref Ref
 Cisgender female 47 36 76.6 1.39 1.08 to 1.79 1.36 1.04 to 1.77§
 Transgender 20 15 75.0 1.36 0.89 to 2.08 1.45 0.95 to 2.20
Race/ethnicity
 Non-Hispanic/Latinx White 135 76 56.3 Ref
 Non-Hispanic/Latinx Black 232 138 59.5 1.06 0.83 to 1.35
 Hispanic/Latinx 28 15 53.6 0.95 0.89 to 1.02
 Non-Hispanic/Latinx other 17 12 70.6 1.25 0.81 to 1.94
Transmission risk
 Gay, bisexual and other MSM 305 165 54.1 Ref Ref
 Transgender 20 15 75.0 1.39 0.91 to 2.12 1.46 0.96 to 2.24
 Heterosexual 84 58 69.0 1.28 1.03 to 1.58§ 1.25 0.95 to 1.65
 PWID 3 3 100.0 1.85 1.46 to 2.34 1.64 1.07 to 2.53§
Sex partner living with HIV, past 12 mo (n = 387) 105 54 51.4 0.87 0.69 to 1.10
Anonymous sex partner, past 12 mo (n = 387) 183 100 54.6 0.92 0.76 to 1.12
Number of sex partners, past 3 mo (n = 296)
 0–2 partner(s) 135 80 59.3 Ref
 ≥3 partners 161 81 50.3 0.85 0.78 to 0.93
STI diagnoses ± 30 d of PrEP initiation (n = 359) 85 44 51.8 0.90 0.78 to 1.04
Composite sexual behaviors/STI diagnosis variable# 286 157 54.9 0.82 0.61 to 1.12
*Low-level PrEP retention was defined as <3 quarter(s) with a PrEP use during the 12 months after PrEP initiation.
Poisson regression with cluster robust standard errors to account for nesting of PrEP users within clinics.
P < 0.01.
§P < 0.05.
P < 0.001.
Includes diagnoses of syphilis, gonorrhea, or chlamydia at urogenital, rectal, or oropharyngeal sites.
#Individuals who reported sex partner living with HIV (past 12 months), anonymous sex partner (past 12 months), ≥3 sex partners (past 3 months) at PrEP initiation, or had laboratory confirmed syphilis, gonorrhea or chlamydia diagnosis ± 30 days of PrEP initiation.
CI, confidence interval; RR, relative risk.

Compared with cisgender males, cisgender females were 39% more likely [adjusted relative risk (aRR): 1.39; (1.04 to 1.77)] to have low-level PrEP retention, when adjusting for age and year of PrEP initiation (Table 2, model 1). When adjusting for age, year of PrEP initiation, heterosexuals and PWID (vs. MSM) were 25% [aRR: 1.25; (0.95 to 1.65)] and 64% [aRR: 1.64; (1.07 to 2.53)] more likely to have low retention, respectively, though only the association between PWID and PrEP retention was statistically significant (Table 2, model 2).

DISCUSSION

Retaining individuals at elevated risk for HIV acquisition on PrEP is critical to reducing HIV incidence.2–4,19 Among individuals who initiated PrEP, we found that in the year after PrEP initiation, the proportion of individuals retained in PrEP care decreased over time, with the largest attrition (23.5%) occurring by the first 3-month PrEP follow-up visit. We also found that nearly three-fifth (58.5%) of individuals had low-level retention, among whom 25.7% had no evidence of any PrEP prescription (ie, lost-to-follow by the 3-month visit) and 27.8% had documented PrEP use at the 3-month visit, but no evidence of PrEP use at the 6-, 9- or 12-month visits. Low-level retention was significantly associated with cisgender females (vs. cisgender males), and was not associated with race/ethnicity, transmission risk, and composite sexual behaviors/STI diagnoses variable.

Our observed retention rates are similar to those observed among PrEP users receiving care at an federally qualified health center in Chicago (complete retention: 15%)13 and an academic health system in New York (6-month: 42%),17 and lower than those observed in 2 primary care health networks in Boston (no PrEP interruptions greater than 7 days: 62%)10 and San Francisco (complete retention: 38%),12 and a sexual health clinic in San Francisco (complete retention: 62%).11 All but one17 of these studies, however, primarily focus on MSM and transgender women and may not be directly comparable to our population. It is important to note, however, that several mathematical modeling studies exploring the impact of PrEP use on HIV incidence assume that individuals who initiate PrEP stay on PrEP for at least 1 year.6,8,24 Our results, combined with previous work, suggest that PrEP retention in real-world settings is substantially lower than what these models assume. This also suggests, that without strategies to improve PrEP retention, the impact of PrEP on population level HIV incidence may be lower than what models have projected. We also found that episodic PrEP use (defined as no documentation of PrEP use at the 3-, 6- or 9-month follow-up visits with evidence presenting for a clinical visit with documented PrEP use in a subsequent quarter) was common. In addition, many individuals with low-level retention did not have documentation of PrEP use at 9- and 12- months after PrEP initiation (73.0%, n = 176/241) and were presumed to have discontinued PrEP for at least 6 months of our observation period. These patterns may reflect changes (i.e., decreases) in HIV acquisition risk over time. Although we were unable to ascertain reasons for PrEP discontinuation, PrEP cohorts from Los Angeles, United States25 and Montreal, Canada,26 reported that the HIV incidence among those who discontinued PrEP were much higher than that among individuals who stayed on PrEP (HIV incidence among those discontinued PrEP vs. those on PrEP: Los Angeles: 2.1/100 vs. 0.1/100 person-years, Montreal: 3.9/100 vs. 0/100 person-years). This suggests that some individuals who discontinue PrEP remain at elevated risk for HIV acquisition. Understanding more about reasons individuals discontinue PrEP, including individual- and provider-level perceived risk for HIV, may help us better understand changes in HIV acquisition risk and need for PrEP over time. In addition, although some barriers to PrEP retention have been identified (ie, transportation, scheduling problems, unstable housing, and low or no social support, substance use, mental health issues, and/or being un- or under-insured9,10,27), more research to improve understanding of barriers and facilitators to PrEP retention is warranted. Moreover, exploring strategies such as telemedicine,28–30 PrEP home delivery service,31,32 and pharmacist-driven PrEP prescribing33 may yield important new avenues to support PrEP retention. More research also is needed to inform alternative strategies like on-demand PrEP.

Identifying individual-level factors associated with low-level retention may inform interventions to improve PrEP retention. In our analysis, PrEP retention in cisgender females were more likely than cisgender male to have low-level retention. Similar to the study conducted by Lankowski et al,17 we found low-level PrEP retention were more likely among heterosexual individuals compared with MSM; however, this association was not statistically significant in the multivariable model. To better inform retention interventions, future research should explore differences in retention levels and barriers and facilitators to PrEP retention across key populations. In contrast to other studies,9,15,17 we found that low-level PrEP retention was similarly likely among racial/ethnic minorities compared to non-Hispanic/Latinx Whites.

There is concern that those at highest risk for HIV acquisition (ie, those reporting sexual behaviors previously associated with HIV acquisition and those with STI history) are less likely to be retained in PrEP care, although the evidence has been mixed.9,11,13,16 We found no association between PrEP retention and sexual behaviors/STI diagnoses. This adds to existing evidence that reported sexual behaviors and STI diagnoses at PrEP initiation do not differ among those who are retained on PrEP and those who discontinue PrEP.

Our study has several important limitations. First, PrEP's effectiveness depends on both adherence to daily PrEP use and retention in PrEP programs.34 We only were able to determine whether there is evidence of PrEP use in the medical record; we were unable to ascertain whether prescriptions were filled, or whether individuals took PrEP as prescribed. Our analysis assumed that individuals without documentation of PrEP use had discontinued PrEP and remained at elevated risk for HIV acquisition. However, we also were unable to determine reasons for PrEP discontinuation. It is possible that those lost follow-up may have discontinued PrEP because they are no longer at elevated HIV risk, or had a PrEP prescription refilled that was not documented in the medical record. Therefore, our results may underestimate true PrEP retention levels. There are some unmeasured individual- (ie, education, income, housing, and insurance status) and clinic-level factors (ie, cultural competency, navigation programs) that may affect PrEP retention and should be explored. In our analysis, cisgender females were more likely than cisgender males to be excluded from the analytic dataset; however, levels of PrEP retention among those excluded from the analysis were not significantly different from those included, suggesting selection bias is minimal. Our current definition of PrEP retention may not correctly measure the patterns of PrEP use among individuals who had irregular clinic visits (ie, those who had at least 2 documentations of PrEP use in one quarterly visit and did not have a documentation of PrEP use in a previous or latter quarterly visit), potentially misclassifying individuals with high-level retention as having low-level PrEP retention. However, we only found 4 (1.7%) individuals who may have been misclassified as low-level of PrEP retention, indicating the risk of misclassification bias is low. Finally, our results may not be generalizable to other settings. The providers participating in this program also provide routine HIV/STI prevention and clinical care services, and PrEP-related services offered by these clinics may not be representative of PrEP services available in other clinical settings.

Overall, the observed PrEP retention in our study suggests that PrEP's effectiveness in reducing HIV transmission may be limited in this and other similar settings. Early considerations of interventions targeting individuals who are no longer engaged in PrEP care may improve PrEP retention. Future work should focus on identifying clinic-level and other structural factors to retention to inform interventions to improve PrEP retention. Our results also highlight the need to improve the understanding of reasons for PrEP discontinuation for different risk groups and reasons for different patterns of retention.

ACKNOWLEDGMENTS

The Initiative to Maximize Prevention, Access, Care, and Treatment: Kathleen Page (Baltimore City Health Department, Baltimore, MD), Maisha Davis (Chase Brexton Health Services, Baltimore, MD), Patrick Ryscavage (Department of Medicine, University of Maryland School of Medicine, Baltimore, MD), Joyce Leitch Jones (Department of Medicine, Johns Hopkins School of Medicine, Baltimore, MD), Jason Farley (Department of Community Public Health, Johns Hopkins School of Nursing, Baltimore, MD), Renata Arrington-Sanders (Department of Pediatrics, Johns Hopkins School of Pediatrics, Baltimore, MD), Vicki Tepper (Department of Pediatrics, University of Maryland School of Medicine, Baltimore, MD), and Mimi Demissew (The Pride Center of Maryland, Baltimore, MD).

REFERENCES

1. Centers for Disease Control and Prevention. HIV surveillance report, 2018 (preliminary); vol. 30. 2018. Available at: http://www.cdc.gov/hiv/library/reports/hiv-surveillance.html. Published November 2019. Accessed February 18, 2020.
2. Grant RM, Lama JR, Anderson PL, et al. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med. 2010;363:2587–2599.
3. Thigpen MC, Kebaabetswe PM, Paxton LA, et al. Antiretroviral preexposure prophylaxis for heterosexual HIV transmission in Botswana. N Engl J Med. 2012;367:423–434.
4. Baeten JM, Donnell D, Ndase P, et al. Antiretroviral prophylaxis for HIV-1 prevention among heterosexual men and women. N Engl J Med. 2012;367:399–410.
5. Anderson PL, Glidden DV, Liu A, et al. Emtricitabine-tenofovir concentrations and pre-exposure prophylaxis efficacy in men who have sex with men. Sci Transl Med. 2012;4:151ra125.
6. Jenness SM, Goodreau SM, Rosenberg E, et al. Impact of the Centers for Disease Control's HIV preexposure prophylaxis guidelines for men who have sex with men in the United States. J Infect Dis. 2016;214:1800–1807.
7. Desai K, Sansom SL, Ackers ML, et al. Modeling the impact of HIV chemoprophylaxis strategies among men who have sex with men in the United States: HIV infections prevented and cost-effectiveness. AIDS. 2008;22:1829–1839.
8. Khurana N, Yaylali E, Farnham PG, et al. Impact of improved HIV care and treatment on PrEP effectiveness in the United States, 2016–2020. AIDS Behav. 2017;21:2895–2903.
9. Liu AY, Cohen SE, Vittinghoff E, et al. Preexposure prophylaxis for HIV infection integrated with municipal-and community-based sexual health services. JAMA Intern Med. 2016;176:75–84.
10. Krakower D, Maloney KM, Powell VE, et al. Patterns and clinical consequences of discontinuing HIV preexposure prophylaxis during primary care. J Int AIDS Soc. 2019;22:e25250.
11. Hojilla JC, Vlahov D, Crouch PC, et al. HIV pre-exposure prophylaxis (PrEP) uptake and retention among men who have sex with men in a community-based sexual health clinic. AIDS Behav. 2018;22:1096–1099.
12. Spinelli MA, Scott HM, Vittinghoff E, et al. Missed visits associated with future preexposure prophylaxis (PrEP) discontinuation among PrEP users in a municipal primary care health network. Open Forum Infect Dis. 2019;6:ofz101.
13. Rusie LK, Orengo C, Burrell D, et al. Preexposure prophylaxis initiation and retention in care over 5 Years, 2012-2017: are quarterly visits too much? Clin Infect Dis. 2018;67:283–287.
14. Hevey MA, Walsh JL, Petroll AE. Prep continuation, HIV and STI testing rates, and delivery of preventive care in a clinic-based cohort. AIDS Educ Prev. 2018;30:393–405.
15. Hosek SG, Rudy B, Landovitz R, et al. An HIV preexposure prophylaxis demonstration project and safety study for young MSM. J Acquir Immune Defic Syndr. 2017;74:21–29.
16. Grant RM, Anderson PL, McMahan V, et al. Uptake of pre-exposure prophylaxis, sexual practices, and HIV incidence in men and transgender women who have sex with men: a cohort study. Lancet Infect Dis. 2014;14:820–829.
17. Lankowski AJ, Bien-Gund CH, Patel VV, et al. PrEP in the real world: predictors of 6-month retention in a diverse urban cohort. AIDS Behav. 2019;23:1797–1802.
18. Chan PA, Mena L, Patel R, et al. Retention in care outcomes for HIV pre-exposure prophylaxis implementation programmes among men who have sex with men in three US cities. J Int AIDS Soc. 2016;19:20903.
19. Centers for Disease Control and Prevention. Preexposure prophylaxis for the prevention of HIV infection in the United States—2017 update: a clinical practice guideline. Available at: https://www.cdc.gov/hiv/pdf/risk/prep/cdc-hiv-prep-guidelines-2017.pdf. Published March 2018. Accessed February 19, 2020.
20. Mugavero MJ, Davila JA, Nevin CR, et al. From access to engagement: measuring retention in outpatient HIV clinical care. AIDS Patient Care STDS. 2010;24:607–613.
21. Waldrop-Valverde D, Guo Y, Ownby RL, et al. Risk and protective factors for retention in HIV care. AIDS Behav. 2014;18:1483–1491.
22. Silva DS, De Boni RB, Lake JE, et al. Retention in early care at an HIV outpatient clinic in Rio de Janeiro, Brazil, 2000–2013. AIDS Behav. 2016;20:1039–1048.
23. Patel P, Borkowf CB, Brooks JT, et al. Estimating per-act HIV transmission risk: a systematic review. AIDS. 2014;28:1509–1519.
24. Kasaie P, Pennington J, Shah MS, et al. The impact of pre-exposure prophylaxis Among men who have sex with men: an individual-based model HHS public access. J Acquir Immune Defic Syndr. 2017;75:175–183.
25. Shover CL, Shoptaw S, Javanbakht M, et al. Mind the gaps: prescription coverage and HIV incidence among patients receiving pre-exposure prophylaxis from a large federally qualified health center in Los Angeles, California. AIDS Behav. 2019;23:2730–2740.
26. Greenwald ZR, Maheu-Giroux M, Szabo J, et al. Cohort profile: L'Actuel pre-exposure prophylaxis (PrEP) cohort study in Montreal, Canada. BMJ Open. 2019;9:e028768.
27. Arnold T, Brinkley-Rubinstein L, Chan PA, et al. Social, structural, behavioral and clinical factors influencing retention in pre-exposure prophylaxis (PrEP) care in Mississippi. PLoS One. 2017;12:e0172354.
28. Klausner J. Using innovations in telehealth to accelerate biomedical HIV prevention. Presented at: CFAR Seminar [Internet]; 2018; Seattle, WA.
29. Telehealth and PrEP plenary. Presented at: YTH Live Conference; 2018; Oakland, CA.
30. Stekler JD, McMahan V, Ballinger L, et al. HIV pre-exposure prophylaxis prescribing through telehealth. J Acquir Immune Defic Syndr. 2018;77:e40–e42.
31. John SA, Rendina HJ, Grov C, et al. Home-based pre-exposure prophylaxis (PrEP) services for gay and bisexual men: an opportunity to address barriers to PrEP uptake and persistence. PLoS One. 2017;12:e0189794.
32. Siegler AJ, Mayer KH, Liu AY, et al. Developing and assessing the feasibility of a home-based preexposure prophylaxis monitoring and support program. Clin Infect Dis. 2019;68:501–504.
33. Tung EL, Thomas A, Eichner A, et al. Implementation of a community pharmacy-based pre-exposure prophylaxis service: a novel model for pre-exposure prophylaxis care. Sex Health. 2018;15:556–561.
34. Fonner VA, Dalglish SL, Kennedy CE, et al. Effectiveness and safety of oral HIV preexposure prophylaxis for all populations. AIDS. 2016;30:1973–1983.
Keywords:

pre-exposure prophylaxis; HIV; retention in care

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.