PrEP Uptake and Discontinuation Among a U.S. National Sample of Transgender Men and Women

Abstract

Pre-exposure prophylaxis (PrEP) has revolutionized HIV prevention efforts by effectively preventing the sexual transmission of HIV infection. Few studies have examined PrEP uptake and discontinuation among transgender individuals who are underserved in HIV prevention efforts. An online U.S. nationwide survey screened 294 HIV-negative transgender people for participation in a larger cohort study. We analyzed associations between demographic characteristics and PrEP use and discontinuation. We conducted thematic analyses of open-ended responses to identify reasons for PrEP discontinuation. Fifty-one participants (17.4%) reported ever receiving a PrEP prescription. Transgender men reported higher PrEP use than transgender women (χ² = 18.06; p < 0.0001). PrEP discontinuation was reported by 49.0% of individuals who reported PrEP use. Reasons for PrEP discontinuation were related to side effects, insurance coverage, relationship status, and access to providers. HIV prevention efforts to increase PrEP should include mechanisms for PrEP uptake and re-initiation among transgender individuals.

INTRODUCTION

Transgender (trans) women are disproportionately affected by HIV and trans men have been historically overlooked in HIV prevention (1). The estimated prevalence of HIV among trans women is 34 times that of cisgender adults (2,3), and incidence among trans women is considerably high (4). While the estimated HIV prevalence for adults in the U.S. is less than 0.5%, a recent metanalysis found that an estimated 9.2% of transgender people are living with HIV: 14.1% are transgender women and 3.2% are transgender men (5). More than half of trans people diagnosed with HIV are Black (44%) or Hispanic/Latinx (26%) (3). Many approaches to end the HIV epidemic use a status-neutral approach and begin with increased HIV testing to identify appropriate prevention interventions or link HIV-positive people with appropriate care (6). Since 2012, these prevention interventions for HIV-negative persons have included pre-exposure prophylaxis (PrEP).

PrEP is highly effective and prioritized for preventing HIV infection among populations with the highest HIV risk. However, studies have shown that trans people are not successfully reached by current HIV testing strategies (7), a precursor to PrEP initiation. Prior research indicates that trans women report generally positive attitudes toward PrEP but low rates of awareness and uptake (8-10), particularly among trans women of color (11). Identified barriers to PrEP among trans women include HIV-related stigma, stigma surrounding sexual behavior, concerns about drug interactions with hormone replacement therapy, and lack of representation in advertising (12,13). Far less is known about HIV testing and the PrEP cascade among trans men (14-17). In addition, few studies have explored both PrEP initiation and discontinuation among transgender people (18). As such, we sought to determine barriers to HIV testing and engagement in the PrEP cascade among a national sample of transgender individuals residing in the U.S.

METHODS

We analyzed screening data collected during recruitment for a large national cohort. As described previously (19,20), the team used internet-based strategies to recruit a geographically and racially diverse cohort of HIV-negative, sexual and gender expansive individuals not currently taking PrEP. For the present analyses, we included screened individuals who were transgender, self-reported HIV negative, and 16-49 years of age. Individuals were recruited via geosocial networking apps from October 2017 through May 2018. No incentive was provided for completion of this screening survey. All study procedures were reviewed and approved by the Institutional Review Board of the City University of New York (CUNY) Graduate School of Public Health and Health Policy.

Measures

The online screening questionnaire included questions about demographic characteristics, HIV testing and STI history, PrEP use, and other health behaviors. One open-ended question assessed reasons for PrEP discontinuation among individuals who were ever prescribed PrEP.

Respondent Characteristics

We operationalized gender through the two-step process of asking sex assigned at birth and gender identity (21). Response options included: Male; Female; Transgender Female; Transgender Male; and Something else. Participants who selected “Something Else” and then indicated a non-binary or gender fluid identity were excluded from these analyses. Thus, this analysis is restricted to participants who selected “Transgender Female” or “Transgender Male” and participants who indicated a sex assigned at birth that did not align with their “Female” or “Male” gender identity. Thus, participants assigned male sex at birth who identify as trans or female were categorized as “trans women” and participants assigned female who identified as trans or male were categorized as “trans men.”

Race/ethnicity categories were coded as Black, white, Latinx, and multi/other race. We dichotomized age to ages 16-26 and age 27+ to reflect the last age that young adults are eligible to be listed on their parent’s health insurance plans. Sexual orientation responses included gay, straight, bisexual, and other. Income was organized into three categories: Less than $20,000, $20,000-$49,000; and $50,000+. Geographic regions included South, Northeast, West, and Midwest. HIV testing frequency was operationalized in three categories: never, more than one year ago, or in the past 12 months. Lifetime sexually transmitted infection (STI) status, a binary variable, indicates participants who reported ever having an STI as “yes” or “no.”

PrEP Indicators

Ever prescribed PrEP, a dichotomous variable, was assessed by the following question: “Have you ever been prescribed HIV medications (e.g., Truvada) for use as PrEP (Pre-Exposure Prophylaxis)?” The original four response options included: I don’t know what PrEP is; No, never taken PrEP; Yes, I am currently on PrEP; and Yes, but I am not currently taking PrEP. Participants who reported ever receiving a prescription for PrEP were given a “1” whereas participants who were unaware of PrEP or reported never taking PrEP were re-coded “0”. Participants who reported previous PrEP prescriptions who were not currently taking PrEP were subsequently asked an open-ended write-in question asking why they stopped taking PrEP.

We assessed PrEP eligibility using existing indications for PrEP based on the 2014 CDC guidelines available during the study period (22). Participants were deemed eligible for PrEP if they reported at least one of the following: one or more receptive condomless anal sex acts with a male partner in the past 3 months; two or more insertive condomless anal sex acts with a male partner in the past 3 months; methamphetamine use in the past 3 months; rectal gonorrhea/chlamydia in the past 12 months; syphilis in the past 12 months; post-exposure prophylaxis use in the past 12 months; or shared injection drug needles in the past 12 months. An additional measure assessed whether participants reported two or more male sex partners in the past 90 days.

Analyses

Data were analyzed in SAS 9.4. We used descriptive statistics to describe the sample of screened participants. We used chi-square tests to investigate associations between demographic characteristics, gender, and PrEP indicators. Free-response data on reasons for PrEP discontinuation were categorized by the first author using thematic analysis (23).

RESULTS

Approximately 43,000 people began the five-minute eligibility screener and 25,000 individuals completed the screener. Of those, 304 identified as transgender. The final analytic sample size includes 294 transgender individuals whose self-reported HIV status was either negative or unknown. Excluded participants included nine trans women and one trans man living with HIV who were not asked the PrEP series due to skip patterns in the screener questionnaire.

Respondent Characteristics

Descriptive and bivariate analyses of the sample are shown in Table I. Over half of the sample (53.4%) identified as trans men. About half of the participants identified as white (45.6%), followed by Hispanic/Latinx (21.8%), Black (11.6%) and more than one race or another identity (21.1%). More than half of the participants (53.7%) were aged 27 or older, 82.0% identified as gay or bisexual, and 43.2% reported some college education. The majority of participants screened (60.5%) reported incomes of less than $20,000 annually. Participants primarily resided in the South (48.6%), followed by the West (23.6%), Midwest (15.6%), and Northeast (12.2%). More than half the sample met indications for PrEP use (59.9%) and 52.0% of participants reported at least two male sex partners in the past 90 days.

Table 1:

Demographic Characteristics of the Sample and History of PrEP Prescription (N = 294)

	Total		Gender identity						Ever Prescri bed PrEP
			Trans woman		Trans man
	(N = 294)	%	(N = 137)	%	(N = 157)	%	χ2	p- value	(N = 51)	%	χ2	p- value
							-	-			18.06	< 0.0001
Trans man	157	53.4	-		-				41	26.1
Trans woman	137	46.6	-		-				10	7.3
							44.07	< 0.0001			1.05	0.7895
Black	34	11.6	23	16.8	11	7.0			5	14.7
White	134	45.6	37	27.0	97	61.8			26	19.4
Latinx	64	21.8	47	34.3	17	10.8			9	14.1
Multi/Other	62	21.1	30	21.9	32	20.4			11	17.7
							5.09	0.0240			7.04	0.0080
16-26	136	46.3	73	53.3	63	40.1			15	11.0
27+	158	53.7	64	46.7	94	59.9			36	22.8
							50.58	< 0.0001			14.89	0.0019
Gay/Queer	136	46.3	46	33.6	90	57.3			35	25.7
Bisexual	105	35.7	44	32.1	61	38.9			14	13.3
Straight	41	14.0	39	28.5	2	1.3			1	2.4
Other	12	4.1	8	5.8	4	2.6			1	8.3
							29.06	< 0.0001			10.51	0.0052
HS / GED	85	28.9	56	40.9	29	18.5			8	9.4
Some college	127	43.2	61	44.5	66	42.0			20	15.8
College grad	82	27.9	20	14.6	62	39.5			23	28.1
							10.02	0.0067			3.24	0.1981
Less than $20,000	178	60.5	91	66.4	87	55.4			31	17.4
$20,000 to $49,000	83	28.2	39	28.5	44	28.0			11	13.3
$50,000+	33	11.2	7	5.1	26	16.6			9	27.3
							22.22	< 0.0001			6.33	0.0967
South	140	48.6	84	63.2	56	36.1			19	13.6
Northeast	35	12.2	10	7.5	25	16.1			11	31.4
West	68	23.6	21	15.8	47	30.3			12	17.7
Midwest	45	15.6	18	13.5	27	17.4			9	20.0
							6.82	0.0090			3.77	0.0521
Yes	109	37.1	40	29.2	69	44.0			25	22.9
No	185	62.9	97	70.8	88	56.1			26	14.1
							8.21	0.0165			50.97	< 0.0001
Never	64	21.8	37	27.0	27	17.2			0	0.0
More than 1 year	69	23.5	37	27.0	32	20.4			0	0.0
12 months or less	161	54.8	63	46.0	98	62.4			51	31.7
							8.31	0.0039			28.79	< 0.0001
Negative	172	58.5	68	49.6	104	66.2			47	27.3
Don't know	122	41.5	69	50.4	53	33.8			4	3.3
							11.13	0.0009			0.21	0.6443
Yes	176	59.9	96	70.1	80	51.0			32	18.2
No	118	40.1	41	29.9	77	49.0			19	16.1
							10.29	0.0013			0.57	0.4483
Yes	153	52.0	85	62.0	68	43.3			29	19.0
No	141	48.0	52	38.0	89	56.7			22	15.6

The demographic composition of the sample varied by gender identity. The majority of trans men were white (61.8%), aged 27 or older (59.9%), gay/queer (57.3%), and reported at least some college education (42.0%) or a college degree (39.5%). In comparison, only 27.0% of trans women identified as white (χ² = 44.07; p < 0.0001). The majority of trans women were aged 16-26 (53.3%) (χ²= 5.09; p = 0.0240), reported a high school GED (40.9%) or some college (44.5%) (χ²=29.06); p < 0.0001), and lived in the South (63.2%) (χ²= 22.22; p < 0.0001).

HIV testing, STI history, and PrEP Eligibility

Overall 37.1% of the sample reported ever having an STI, a smaller percentage of trans women reported ever having an STI (29.2%) compared to trans men (44.0%) (χ² = 6.82; p = 0.0090). Knowledge of HIV status was low in general among participants: 41.5% of the sample were unaware of their HIV status, and half of trans women were unaware of their status (50.4%), compared to only 33.8% of trans men (χ² =8.31; p = 0.0039). Moreover, 21.8% of participants reported never having been tested for HIV and 23.5% reported not having been tested in the past 12 months. Trans men reported more recent HIV testing compared to trans women: 62.4% of trans men reported receiving an HIV test in the past 12 months compared to only 46.0% of trans women (χ² = 8.21; p = 0.0165).

PrEP eligibility also varied by gender: 70.1% of trans women were eligible for PrEP based on CDC guidelines compared to 51.0% of trans men (χ² = 11.13; p = 0.0009). The percentage of trans women reporting at least two male sex partners in the past 90 days (62.0%) was also higher than trans men (43.3%) (χ² = 10.29; p = 0.0013).

PrEP Prescriptions

PrEP use was relatively low across the sample. Although 91.5% of participants reported having heard of PrEP, only 51 participants (17.4%) reported ever receiving a PrEP prescription, comprising a greater percentage of transgender men (26.1%) compared to transgender women (7.3%) (χ² = 18.06; p < 0.0001). Moreover, 49.0% of participants who were ever prescribed PrEP were not currently taking PrEP (data not shown). We found no differences in PrEP prescriptions by race or ethnicity, however PrEP prescriptions were higher among participants aged 27 or older (22.8%) compared to individuals age 26 and younger (11.0%) (χ² = 7.04; p = 0.0080). In addition, a greater percentage of participants who identified as gay/queer reported PrEP prescriptions (25.7%) compared to those who identified bisexual (13.3%), straight (2.4%) or reported another sexual orientation (8.3%) (χ² = 14.89; p = 0.0019). A higher percentage of participants with a college degree (28.1%) reported receiving a PrEP prescription than those with some college education (15.8%) or high school/GED equivalent (9.4%) (χ² = 10.51; p = 0.0052).

PrEP Discontinuation

PrEP discontinuation was reported by 49.0% (n = 25) of individuals who reported ever having received a PrEP prescription. Of those, seven were trans women (28%) and 18 were trans men (72%). Thus, among participants who ever received a PrEP prescription, 70% of trans women reportedly discontinued PrEP compared to 43.9% of trans men. More than half of the individuals who reportedly discontinued PrEP were people of color (n = 13; 52%). In addition, 40.0% of individuals who discontinued PrEP were between the ages of 16 and 26. Text responses to reasons for PrEP discontinuation provided by participants were organized around four themes: side effects, insurance, changes in relationship status, and access to providers (see Table II).

Table 2.

Reasons for PrEP discontinuation among transgender participants (N = 25)

Cost or insurance coverage
Ran out of medication, too expensive now	31 year-old, Trans man, white
Never started, too expensive	23 year-old, Trans man, Latinx
Lost insurance	34 year-old, Trans woman, white
Expensive	24 year-old, Trans woman, Multi/Other
Cost is too high, I can't afford	28 year-old, Trans man, Multi/Other
Side effects
Had issues taking it	29 year-old, Trans man, Multi/Other
High STD rate	36 year-old, Trans man, Latinx
I felt like it did more harm to my body than needed.	30 year-old, Trans man, Latinx
It upset my stomach	19 year-old, Trans man, white
I was having complications with my GI tract (believed to be unrelated) and needed surgery	33 year-old, Trans man, white
Adverse side effects	20-year-old, Trans man, white
Side effects	27 year-old, Trans man, white
Side effects and not sexually active	33 year-old, Trans man, white
Stomach and digestive issues	29 year-old, Trans man, Latinx
Access to provider
New doc- have to see his referral	39 year-old, Trans man, white
I moved	21 year-old, Trans woman, Latinx
My doctor said he was no longer "comfortable" prescribing it after my last potential exposure	31 year-old, Trans man, white
Relationship status
I'm not sexually active lately	31 year-old, Trans woman, Multi/Other
Headaches and relationship with positive partner ended	22 year-old, Trans man, white
Don't know or never started taking it
I don't know	20-year-old, Trans woman, Black
I don't take it	21 year-old, Trans man, white
Never started	21 year old, Trans woman, Multi/Other
Never started	40 year old, Trans man, white
Never took it	26 year-old, Trans man, Latinx
Never took it	29 year-old, Trans woman, Black

Participants indicated cost and insurance coverage as prohibitive to PrEP persistence and initiation. For example, participants indicated they never started taking PrEP because it was too expensive, they lost insurance coverage, or ran out of medication. Side effects mentioned by participants who discontinued PrEP included upset stomach and other digestive issues among other unspecified side effects. Three participants noted that access to providers was a barrier to PrEP continuation, indicating inaccessibility due to a new doctor making a referral, moving, and one provider no longer willing to prescribe PrEP. Three participants reported no longer being sexually active or changes in partner status as reasons for discontinuation. Notably, six participants reported that although they received PrEP prescriptions, they never began to take the medication.

DISCUSSION

Optimal implementation of PrEP to End the HIV Epidemic (24) will necessitate identifying strategies to scale up PrEP among trans populations, including understanding discontinuation following initiation in the PrEP cascade. The results of this study support prior research demonstrating that trans men and women report mismatched HIV testing and awareness levels despite elevated risk for HIV infection (7), and are not being adequately reached by current HIV prevention efforts, including PrEP services. Moreover, we found relatively high PrEP discontinuation among trans individuals who reportedly ever were prescribed PrEP: 70% of trans women reportedly discontinued PrEP followed by 43.9% of trans men. In comparison, a recent prospective cohort study among young Black men who have sex with men (MSM) who initiated PrEP found 69% discontinued after a median of 165 days and 40% discontinued during the 24-month study period (25).

We found that trans men were more likely to report PrEP prescriptions and prior STIs compared to trans women. Although more trans women than trans men met at least one PrEP eligibility criterion, 51.0% of trans men in the sample met at least one risk factor cited in the PrEP guidelines, and 43.3% of trans men reported at least two male sex partners in the past 90 days. These findings highlight the need for increased efforts to increase PrEP uptake for trans men.

Participants in this study also reported a significant history of prior STIs (37.1%) and sub-optimal HIV testing: 21.8% of our sample reported never having tested for HIV in their lifetimes. Unsurprisingly, all participants who reported any PrEP use also reported having been tested for HIV in the past year, given that testing is a precursor to accessing PrEP. However, that none of the individuals who reported testing more than a year ago reported taking PrEP suggests that all of participants may have initiated PrEP relatively recently. This finding supports other published evidence that increased access to testing may be an important facilitator to PrEP awareness and engagement among transgender individuals (15,26). Moreover, because half of the participants who reported PrEP initiation subsequently discontinued taking PrEP, the open-ended responses to reasons for discontinuation are critical for efforts to understand fluctuations in PrEP uptake and opportunities to intervene to improve the PrEP cascade.

Reasons for PrEP discontinuation reflect themes from the extant literature focused on other populations including changing sexual practices or relationship status (27), cost, insurance, and provider-related barriers (20), and side effects (28). Of note, none of the participants attributed PrEP discontinuation to negative perceptions of PrEP among sexual partners or other influential peers, a reason for PrEP discontinuation recently identified in a qualitative study among MSM (29). Although we did not have the statistical power to identify factors associated with PrEP persistence in the sample, these preliminary findings support the need for more research to address barriers and facilitators to PrEP use and persistence among trans men and women.

We found that participants continue to report cost barriers, despite the existence of medication assistance programs. Beyond providing individual assistance with navigating insurance and co-pay relief programs, our findings indicate that lowering the cost of PrEP and expanding health care coverage would make PrEP more accessible for trans individuals. We also anticipate that providers may need further support in advocating for co-pay relief from non-drug related costs of PrEP monitoring, including HIV/STI testing. In addition, healthcare stigma and access to gender-affirming and competent care remain a barrier for transgender populations (30). Interventions that improve both provider knowledge and competency in the provision of trans health care, particularly comfort in testing trans people for HIV and other STIs and prescribing and monitoring PrEP in primary settings, may improve PrEP persistence over time. Moreover, although this study specifically mentioned Truvada, perceptions of side effects for Descovy should also be monitored in addition to variations in experience with long-acting injectables as they become available. Of note, this study was conducted before Descovy was approved for use and Descovy has yet to be studied in individuals assigned female sex at birth.

Finally, qualitative responses to reasons for PrEP discontinuation revealed that six out of the 51 participants who ever received a PrEP prescription never actually started taking the medication. Although the screener question asked participants whether they were ever prescribed PrEP, the response options included whether they ever and currently took PrEP. Thus, through the open-ended responses in the survey, we were able to ascertain that six individuals never began taking PrEP after receiving their prescription. To accurately monitor all the PrEP cascade, future studies should additionally explore barriers to PrEP initiation among trans individuals who received a PrEP prescription but never begin taking the medication. We recommend using mixed methods approaches to full capture all facets of the PrEP cascade.

LIMITATIONS

Our cross-sectional survey presents screening data from a larger U.S. national study, with a small sub-sample of transgender participants who discontinued or never started PrEP after receiving a prescription. The self-report data we report are subject to recall and social desirability bias, although social desirability bias is reduced when surveys are self-administered. Further, because the screener was made available online through geosocial networking apps, the results presented here may only be generalizable to trans individuals who utilize those networking apps.

We also note that we did not include vaginal sex in our indicator for PrEP treatment. The parent study from which these data were derived was primarily focused on enrolling cisgender men who have sex with men; however, we permitted gender expansive individuals to complete our screening survey as we saw this as an important opportunity to gather data on trans health. The PrEP eligibility metric was designed primarily for cisgender men who have sex with men in accordance with the 2014 CDC guidelines during the study period in 2017. It may be that were we to have included more questions on vaginal sex, a greater number of individuals would have met clinical indicators for PrEP treatment.

Finally, the study design may have resulted in potential bias due to advertisements being appealing to people who are interested in free HIV tests or were motivated to receive the incentive associated with the secondary, baseline survey of the parent study. Although the number of trans individuals who discontinued PrEP was small in this sample, ours is one of the first studies to use a national sample to explore reasons for PrEP discontinuation among transgender individuals.

CONCLUSIONS

Despite the heightened risk for HIV among transgender individuals, few interventions have been specifically developed that address the PrEP cascade for this diverse and underserved population (31). We found relatively low PrEP uptake among a sample of transgender individuals with relatively high PrEP discontinuation (49%). Nearly half of respondents who reported receiving a PrEP prescription discontinued the medication, and while our sample is small it is notable that half of the participants who discontinued PrEP represent priority populations of younger Black and Latinx transgender individuals. Efforts to support PrEP use and persistence among transgender individuals should consider reasons for PrEP discontinuation and identify mechanisms for potential re-initiation.