Sociodemographic Characteristics of HIV Pre-Exposure Prophylaxis Use and Reasons for Nonuse Among Gay,Bisexual,and Other Men Who Have Sex with Men from Three US Cities

Measures

Demographic characteristics were categorized by age (18–29, 30–39, ≥40 years), race/ethnicity (White, Black/African American, Hispanic/Latino, Other/mixed), education (<4-year college degree, 4-year college degree, post-Bachelor's education), health insurance (private, public only, uninsured), sexual orientation (gay, bisexual/other), and city/MSA of residence (Atlanta, Detroit, New York City).

For CAS, participants were asked “In the past 3 months, have you had any anal sex with casual male partner in which a condom was not used from start to finish?” (Yes/No). Risky drinking was assessed by using the Alcohol Use Disorder Identification Test-Concise (AUDIT-C), ³⁸ a three-item abbreviated form of the original AUDIT questionnaire. ³⁹ As recommended in literature, scores ≥5 out of 12 were coded “Yes” for risky drinking. ^39,40 Problematic drug use was assessed by using Drug Use Dependency Identification Test (DUDIT), an 11-item self-reported questionnaire. ⁴¹ Scores ≥6 out of 44 were categorized as “Yes” for problematic drug use as recommended in literature. ^{41 –43}

Participants were asked several questions about their PrEP use and discontinuation. MSM who responded “Yes” to “Are you currently taking PrEP to prevent HIV?” were categorized as currently using PrEP. Men not currently using PrEP were asked; “In the past 3 months, were you taking PrEP but stopped?” and “Have you ever used PrEP to prevent HIV?”; those who responded “Yes” to either question were categorized as previously used PrEP. Participants who responded “No” to “Have you ever used PrEP to prevent HIV?” were considered as never users of PrEP.

Reasons for never using PrEP and PrEP discontinuation were assessed by asking “What are some of the reasons you've never used (stopped using) PrEP?” (Select all that apply). Examples of response options (see full list in Table 3) include “I didn't know where to get it,” “I couldn't afford it,” and “I was worried about the safety of PrEP.” Additional reasons were presented to participants but were not reported in Table 3, due to low response frequency.

Statistical analysis

The full RCT study included HIV-positive and HIV-negative MSM (N = 1226); however, current analyses were limited to HIV-negative MSM (N = 778). Key characteristics were assessed for the sample and by PrEP use/nonuse. Bivariate analyses were conducted by using chi-square tests to examine differences among the three PrEP use/nonuse categories by age group, race/ethnicity, education level, insurance type, sexual orientation identification, city/MSA, CAS, risky drinking, and problematic drug use, selected based on prior literature. ^8,19 We conducted multinomial logistic regression models for PrEP use [current vs. never (Model 1), previous vs. never (Model 2), and current vs. previous (Model 3)], with all variables in each model to determine their independent associations with PrEP use. Participants who never heard of PrEP (n = 26) were excluded from the reasons for PrEP nonuse analysis, as these items were not asked. We calculated frequencies of reasons for not using and for discontinuing PrEP and conducted Chi-square/Fisher exact tests to examine differences based on CAS. Analyses were performed by using SAS^® 9.4.

Reasons for PrEP nonuse

Reasons for PrEP nonuse (never and discontinued use) were categorized as: cost and availability; health and safety; lower perceived need and committed relationship; and other. Among the MSM who never used PrEP (n = 450), commonly reported reasons were related to cost and availability (Table 3): Not knowing where to get PrEP (27%); Not able to afford PrEP (25%); and Having insurance that wouldn't cover PrEP (15%). Lower perceived need and committed relationship reasons were Decided didn't need PrEP (34%) and Starting a committed relationship (19%). Health and safety reasons were Worrying about side effects of PrEP (38%) and Doctor did not recommend PrEP (20%). In bivariate analysis (p < 0.05), more men who reported CAS (n = 151) [vs. did not (n = 299)] reported that Insurance wouldn't cover PrEP (20% vs. 12%, respectively), and Didn't know where to get it (33% vs. 24%). Fewer men who reported CAS (vs. did not) said they Didn't need PrEP (23% vs. 39%) or Started a committed relationship (7% vs. 25%) as reasons for never use.

Reasons for PrEP discontinuation

MSM who previously used PrEP (n = 59), Starting a committed relationship (32%) and Deciding they didn't need PrEP anymore (29%) were most reported reasons for discontinuation (Table 3). Cost and availability reasons were: Not being able to afford PrEP (19%) and Lack of insurance covering PrEP (12%). Health and safety related reasons were: Didn't like side effects (17%) and Worry about safety of PrEP (10%). More men who reported CAS (n = 26) [vs. did not (n = 33)] said Worry about safety of PrEP (19% vs. 3%, respectively) was a reason for discontinuing PrEP.

Use of PrEP currently, previously, and never

This study examined correlates of current, previous, and never use of PrEP in a sample of HIV-negative MSM from three cities. We assessed reasons for PrEP nonuse by CAS, ²⁸ in a sample with more current PrEP users (31%) compared with prior research; ^8,9,19,44 another 8% were previous PrEP users, similar to past studies. ^19,21,44 Having representation of current, previous, and never users in our sample allowed us to examine sociodemographic and behavioral associations with PrEP use status, and reasons for nonuse, in an era of increasingly common PrEP use where gaps persist in access, uptake, and persistence of use for MSM. ^11,12,16,20

Compared with never users, current and previous PrEP users were similar in reporting CAS, younger age, and higher education. The association between CAS and PrEP use is not surprising, because CAS is an indication for PrEP use. ^19,28,44 In addition, participants reported high adherence to PrEP (84% reported 100% adherence for past 7-day), which could offer protection against HIV and may explain the association of current PrEP use and CAS. However, among MSM previously used PrEP, higher CAS indicates heightened risk for HIV infection for those who discontinued PrEP use. These findings are telling in the context of expanding community PrEP use, where measurement and understanding of unprotected sex (condomless and PrEP less sex as “unprotective” and condom or PrEP use as “protective”) are changing. ⁴⁵ Thus, our finding of previous PrEP users reporting CAS underscores the need for new interventions in this era to re-initiate condom use among MSM who discontinue PrEP. ^20,35,44 Messaging is also needed to enhance PrEP use among less educated MSM, if they are PrEP eligible, given their associations with lower likelihood of PrEP use found here; for example, messaging about PrEP through community outreach ads and other social marketing efforts. ^46,47

Contrary to earlier studies on substance use and current PrEP use and adherence, ^19,48 we found that risky drinking and problematic drug use were not associated with current and previous PrEP use, as have others recently. ^26,49 This may suggest that risky drinking and problematic drug use may not necessarily interfere with PrEP use for some men. ^49,50 However, considering the evidence on the association of alcohol use and substance use with sub-optimal ART adherence, ^{51 –53} future studies might examine PrEP use among MSM who report drug and alcohol use. We found no differences in PrEP use for Black and Hispanic/Latino MSM compared with White MSM, similar to other recent studies ^8,44,54,55 but in contrast with earlier research. ^{10,16 –18} This could be due to biases in our sample, which was recruited from urban areas and highly educated, similar to studies that found no significant racial/ethnic differences in PrEP use when controlled for urbanicity and education levels. ^8,44 However, given that MSM of color have higher HIV incidence than White MSM, equivalent levels of PrEP use might still not translate to equitable prevention and widen disparities in HIV incidence. ⁵⁶ It is important to continue addressing PrEP access and uptake issues for MSM of color.

Not having health insurance was associated with lower likelihood of current PrEP use. Although improving in some parts of the United States, ^{57 –59} lack of insurance is still a barrier to PrEP access, ^8,10,19 despite current implementation of financial assistance programs for covering PrEP costs. ^{10,33,44,60 –62} Further efforts are needed in this area, including implementation of PrEP-Drug Assistance Programs and consideration for expansion of Medicaid, ¹³ and other mechanisms to reduce in PrEP costs for the uninsured. ⁶³

Reasons for PrEP nonuse

Primary reasons for not using PrEP (both never and discontinuation) were health and safety concerns, cost and availability, and lower perceived need. These findings highlight the need for messaging about: relatively low side effects to ease concerns about PrEP, personal risk assessment and indications for PrEP, and PrEP clinics and assistance programs to help cover PrEP costs. ^{12,16,20,21,64 –68} Evidence of effective messaging through advertisement campaigns and social networks to enhance PrEP uptake has been encouraging and might be expanded to reach more MSM who need PrEP. ^46,47

We found that insurance coverage and access were prominent reasons for never using PrEP among MSM who reported CAS. More MSM who did not report CAS indicated lower perceived need and starting a committed relationship as appropriate reasons for never use. For discontinuation, more MSM who reported CAS noted concerns about safety of PrEP as a reason. Although marginally significant, more MSM who did not report CAS said starting a committed relationship was an appropriate reason for PrEP discontinuation. A considerable proportion (23% of never and 31% of previous users) of MSM who reported CAS indicated lower perceived need (Decided I didn't need it/anymore) as a reason. This finding underscores the discrepancy between perceived need and PrEP eligibility. ^{21,32,68 –70} Efforts are needed to educate MSM about personal risk assessment and further discussion between MSM and their provider on PrEP eligibility.

Many reasons noted by MSM who reported CAS may be considered barriers (e.g., lack of access, insurance coverage, affordability, concern for side effects), although some reasons for not using PrEP are appropriately determined by an individual and their provider (e.g., provider not recommending, actual side effects). Reasons for PrEP nonuse noted by MSM who did not report CAS could be appropriate (e.g., starting committed relationship), until their risk status and PrEP eligibility changes. This distinction may help in designing tailored prevention messaging to increase PrEP uptake. For example, messaging about safety and side effects could help alleviate some concerns of MSM who reported CAS. Providers should discuss with MSM about perceived vs experienced side effects of PrEP. Messaging all men on availability of PrEP could improve PrEP use if and when it is needed. ²¹

Our study has limitations, including behavioral assessment recall and reporting and a nonexhaustive list of reasons provided for PrEP nonuse, although they are consistent with previous research. ^19,20,26,44 The study sample had high education levels and was recruited only from three cities (i.e., nonrural), ^8,71 with accordingly limited generalizability. However, the sample was relatively large and diverse with sizeable representation of current, previous, and never PrEP users, which provides timely results in a transitional era of increasing PrEP use. Future research might continue to focus on current and previous PrEP users to better understand PrEP use over time in the context of long-acting injectable PrEP medications and other future products. ^{8,19,31,45,72}

Differing reasons for PrEP nonuse by CAS among MSM has important implications for public health and clinical decision making. Among MSM who reported CAS, access, and lack of insurance coverage, and worry about safety of PrEP were major barriers for never use and discontinuation, respectively. Lower perceived need was understandably a major reason for nonuse among MSM who did not report CAS. Future research might focus on better understanding barriers specific to MSM based on individual HIV risk level. The association of CAS with previous use could indicate increased HIV risk, providing messaging about re-initiating condom use for men who discontinue PrEP.