The Role of Social Biases, Race, and Condom Use in Willingness to Prescribe HIV Pre-exposure Prophylaxis to MSM: An Experimental, Vignette-Based Study : JAIDS Journal of Acquired Immune Deficiency Syndromes

Secondary Logo

Journal Logo

Prevention Research

The Role of Social Biases, Race, and Condom Use in Willingness to Prescribe HIV Pre-exposure Prophylaxis to MSM: An Experimental, Vignette-Based Study

Bunting, Samuel R. MDa; Feinstein, Brian A. PhDb; Calabrese, Sarah K. PhDc; Hazra, Aniruddha MDd; Sheth, Neeral K. DOe; Wang, Gary BSf; Garber, Sarah S. PhDg

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes 91(4):p 353-363, December 1, 2022. | DOI: 10.1097/QAI.0000000000003072



Men who have sex with men (MSM) experience disproportionately high incidence of HIV.1 Centers for Disease Control and Prevention (CDC) data showed that Black MSM accounted for approximately 25% of new HIV infections in 2019.1 Daily pre-exposure prophylaxis (PrEP) using emtricitabine tenofovir disoproxil fumarate/alafenamide (FTC-TDF/TAF) is effective at preventing HIV transmission among at-risk MSM.2–4 Multiple studies have confirmed the efficacy of PrEP, reaching 99% with daily dosing.3,4

Although the effectiveness of PrEP is well-established, only approximately 35% of the estimated 1.4 million American MSM with PrEP indications were prescribed PrEP in 2019.5–7 Uptake of PrEP remains especially low among Black MSM, with 42% of White MSM using PrEP compared with 26% of Black MSM.5,8 Considering these gaps, understanding barriers to wider PrEP prescription is imperative.

Since approval in 2012, there have been concerns that PrEP use would contribute to increased frequency of condomless sex, less disclosure of HIV status, or increased numbers of sexual partners (so-called “risk compensation”).9 Literature regarding risk compensation has been mixed, with some studies reporting increased numbers of sexual partners and condomless sex, but others finding no meaningful differences among patients after beginning PrEP.10,11 A related concern has been suboptimal adherence to PrEP leading to HIV resistance to FTC-TDF/TAF, an exceedingly small risk outweighed by the benefits of PrEP.12 Previous studies of physicians have identified assumed risk compensation and concerns about nonadherence as reasons for not prescribing PrEP.13

In addition, growing evidence suggests that interpersonal stigma and provider biases, like heterosexism (bias toward nonheterosexual people), can influence willingness to prescribe PrEP and also bias judgments about risk compensation.6,13,14 A study of clinical documentation revealed that multiple partners, inconsistent condom use, and “admissions of homosexual behavior” were cited by clinicians as reasons for not prescribing PrEP to requesting patients.15 For decades, people at risk for HIV—particularly MSM—who did not use condoms or had multiple sexual partners were been stereotyped as sexually irresponsible and “promiscuous.”16–19 The stigma surrounding condomless sex has continued but shifted and is now applied to those who are taking the responsible step of using the most effective available HIV prevention method (PrEP), such that people taking PrEP are viewed as irresponsible and promiscuous, with PrEP precipitating sexual indiscretion because of the reduction in HIV risk.14,16,18,19

Furthermore, a long-standing theme of US public health messaging and interventions has been the notion of “personal responsibility” because it relates to sexual health and condom use.20 Exclusive reliance on PrEP for HIV prevention may be viewed as a violation of an implicit, personal responsibility to uphold public health.18 These views are likely connected to concerns about risk compensation after PrEP initiation. Although extremely effective at preventing HIV, PrEP offers no protection from other STIs. Reliance on PrEP for protection during sex may be superior for HIV; however, there is the theoretical risk of increasing transmission of STIs if condoms are discontinued, thus theoretically presenting a public health controversy.

Related to the stigmatizing forces surrounding condom use are those related to relationship dynamics and the moral superiority often ascribed to monogamous relationships.21 In studies assessing attitudes toward nonmonogamous relationships, individuals in nonmonogamous relationships have been viewed as less trustworthy and less likely to attend to basic personal responsibilities like flossing teeth and caring for pet animals, in addition to being viewed as less trustworthy.22 People in a nonmonogamous relationship have also been judged as being more likely to transmit STIs, including HIV, compared with those in a monogamous relationship.22 However, evidence suggests the contrary; people in nonmonogamous relationships are more likely to be tested for STIs and more likely to use condoms.23 Internalization of these beliefs could lead healthcare providers to stigmatize patients seeking PrEP who opt to forgo condoms or have multiple partners, thereby impeding PrEP prescription.14

It is also pertinent to consider how biases interact with each other for patients with multiple minoritized identities, who experience unique forms of stigma related to intersecting systems of oppression.24 Based on intersectionality theory, the stigma experienced by a Black gay man is unique from the stigma experienced by Black men and gay men as individual groups.24 In the context of healthcare, this means that individuals with intersecting minoritized identities may experience unique discrimination that presents a unique barrier to access to care.25 Biased judgments based on race may be present alongside judgments related to same-sex sexual activity and judgments related to personal decisions about condom use, creating unique barriers to PrEP prescription for Black MSM who choose to rely exclusively on PrEP for HIV prevention.

Several previous studies have begun to investigate the role of provider biases related to patients' sexual behavior and race in clinical encounters with patients seeking PrEP prescriptions. One study found medical students believed that a Black MSM would be more likely to engage in condomless sex if prescribed PrEP as compared with a White MSM,26 although this finding has not been consistently upheld.27 Additional research has focused specifically on the role of patients' condom use on medical students' willingness to prescribe PrEP.28 This work identified the “prevention paradox” because medical students were less willing to prescribe PrEP to a patient who was not using condoms or who planned to discontinue condom use compared with a patient at lower risk—one who used condoms and planned to continue using condoms if prescribed PrEP.28 Medical students were also less willing to prescribe PrEP to hypothetical patients with multiple partners compared with a single partner.28 Studies of physicians also showed that respondents were more willing to prescribe PrEP to MSM in HIV-serodifferent relationships rather than single MSM.29,30

The current study builds on this previous work in several important ways.28 First, we systematically varied patient race and examined the interaction between patient race and condom behavior in relation to willingness to prescribe PrEP. Second, we examined assumptions about the patient as potential mediating mechanisms, and we tested social biases as potential moderating factors. Finally, previous work presented limited clinically relevant information, presenting patients as “An HIV- man…. He wants to take PrEP and continue using condoms.”28 Our study presented patients in the format of a clinical vignette with supporting information to mirror actual practice more closely.

Specifically, the present study sought to determine: (1) Do assumptions about adherence to PrEP and HIV risk differ based on MSMs' condom use and/or race? (2) Does willingness to prescribe PrEP differ based on patients' condom use and/or race? (3) Do assumptions about adherence and HIV risk mediate the association between patients' condom use and willingness to prescribe PrEP? (Fig. 1A) (4) Do social biases related to sexual orientation and nonmonogamy moderate indirect associations between patients' condom use and willingness to prescribe PrEP? (Fig. 1B) and (5) Do patient race and implicit racism interactively moderate the indirect association between condom use and willingness to prescribe PrEP? (Fig. 1C).

Conceptual diagrams of the parallel mediation, moderated mediation, and moderated moderation models. All analyses controlled for participants' year of training, sexual orientation, gender identity, and race. For all analyses, the planned continuation condom use behavior was taken as the reference group. Model A: Model estimated without any moderating variables. Model B: Moderated mediation model estimated twice, once considering explicit heterosexism (MHS) as the moderator and then again using attitudes toward nonmonogamy (ATCNM) as the moderating variable. Model C: Moderated moderation model estimated using implicit racism as a moderating variable of the effect of Black patient race relative to White.

Regarding differences by condom use, we hypothesized: The patient who is using condoms and intends to continue using them if prescribed PrEP would be viewed as being (1) more likely to adhere to PrEP (given that participants may view condom nonuse to be suggestive of nonadherence) and (2) at lower HIV risk than condom users who intended to discontinue and nonusers. Based on previous research, we further hypothesized that participants would be more willing to prescribe PrEP to patients who were using condoms and intended to continue using them if prescribed PrEP compared with the other 2 groups.28 Regarding differences by patient race, we hypothesized that the Black patient would be viewed as being (1) less likely to adhere to PrEP and (2) at higher HIV risk compared to the White patient. We did not hypothesize a racial difference in willingness to prescribe PrEP, consistent with previous studies.26,27,31 Finally, we hypothesized that negative attitudes toward nonmonogamous relationships would be associated with decreased willingness to prescribe PrEP based on a theoretical fear of condoning a less responsible relationship pattern compared with monogamy.


Study Population and Recruitment

The present study was conducted as a component of a larger research initiative focused on medical students' knowledge, decision-making, and biases related to PrEP.32,33 Data were collected between March and May 2021. Participants were allopathic and osteopathic medical students from 16 US medical schools (10 allopathic and 6 osteopathic) with a combined enrollment of 12,660 students.

A total of 1,592 students indicated interest in participating in the current study and were sent an informational message about the study, purposely excluding the focus of the study as being about PrEP and HIV to limit selection bias. Interested participants who met the inclusion criteria (currently studying in a US allopathic or osteopathic medical education program, aged 18 years or older) were sent a follow-up email containing a link to access the study. The study was completed via Qualtrics (Provo, UT). After completing the study, participants were given a $10.00 credit to an online retailer and sent a debrief message.

Study Procedures and Instrument

A focus group of 5 allopathic medical students reviewed and provided feedback on the vignettes and study measures before broad distribution. We incorporated minor edits to vignette and item wording and instrument mechanics based on focus group feedback.

In the online study instrument, after completion of the informed consent, participants were randomized to 1 of 6 patient vignettes (described below). Randomization was completed using the Qualtrics randomization program, programmed to ensure a balanced number of participants in each of the 6 experimental conditions. The full study instrument is attached as Supplements, Supplemental Digital Content,

Patient Vignette

We systematically varied patient condom use and race in a 3 × 2 factorial design (see Figure S1, Supplemental Digital Content, Patient race was conveyed through an image of the fictional patient. Images were taken from the Chicago Faces Database, a public resource containing images of people across multiple races, ages, and gender identities.34 Images matched the race and age of the vignette patient. The Database also contains facial measurements and we matched images between the Black and White conditions to avoid potential confounding effects. The vignette described a MSM patient presenting to a primary care physician seeking PrEP (Fig. 2). All vignettes were presented in the format of a simulated electronic medical record. Laboratory results, including a negative HIV antigen/antibody test, were included.

Example vignette presentations are shown. The bolded text in the figure is for emphasis only and was not included in the experiment. All patients and their partners were presented as cisgender, as indicated by matching sex at birth and gender identity. All patients were described as in overall good health without significant medical history. Patient images are shown in the right pane of the figure, which were used to convey patient race. All remaining details were identical across vignette presentations.

Vignette Follow-up Items

After the vignette, participants responded to a series of follow-up items regarding the patient. Participants rated patients' assumed adherence to PrEP if prescribed (1 = least likely, 7 = most likely), and their assumption about the overall HIV-risk of the patient without PrEP (1 = extremely low, 7 = extremely high). Participants also indicated their willingness to prescribe PrEP to the patient (1 = least likely, 7 = most likely).

Social Bias Measures

Participants completed a series of social bias measures including the Attitudes Towards Consensual NonMonogamy Scale [⍺ = 0.82, 95% confidence interval (CI): 0.80 to 0.84],35 the Modern Homonegativity Scale (MHS; ⍺ = 0.92, 95% CI: 0.91 to 0.93),36 and an Implicit Association Test (IAT) to detect implicit racial biases toward Black people, adapted from the publicly available IATs from Harvard University's Project Implicit.37,38 The interpretable measurement of the IAT is the d-score, which ranges from −2 to +2, with a higher score suggesting greater implicit racism.

Manipulation Checks

We included 2 manipulation checks purposely separated from the vignette by several blocks of unrelated items. First, we asked participants: “In the clinical vignette, what was the patient's race/ethnicity?” Participants selected from a multiple-choice list. The second item asked: “Which of the following describes the patients' condom use?” Participants selected from the 3 condom use groups included in the study. Participants “passed” each manipulation check if they correctly identified race or condom use for the patient in the experimental condition to which they were randomized.

Statistical Analyses

Descriptive statistics and Pearson correlation coefficients were calculated for assumptions, bias measures, and willingness to prescribe PrEP. Analysis of covariance with Bonferroni pairwise post hoc tests was used to examine the effects of patient race, condom use, and their interaction on participants' assumptions about adherence, HIV risk, and willingness to prescribe PrEP.

To explore potential mechanisms by which patients' condom use practices affected willingness to prescribe PrEP, we conducted multiple mediation and moderated mediation analyses (Figs. 1A–C) using Hayes's PROCESS macro for SPSS (v3.5).39,40 First, we tested for indirect effects of condom use on willingness to prescribe PrEP via parallel mediators of assumed adherence to PrEP and HIV risk (Fig. 1A). We also evaluated whether social biases [heterosexism (MHS), attitudes toward nonmonogamous relationships (ATCNM)] moderated indirect effects of condom use on willingness to prescribe PrEP (moderated multiple mediation; Fig. 1B). Finally, we evaluated whether patient race and implicit racism moderated the indirect association between condom use and willingness to prescribe via assumed adherence and HIV risk (Fig. 1C).

For all models, indirect effects were evaluated by generating 10,000 bootstrapped samples from which 95% CI were established. Indirect effects were considered statistically significant if the 95% CI did not include zero. Moderation analyses were interpreted only if the index of moderated mediation (IMM) was statistically significant based on a 95% CI that did not include zero.41 Analyses were completed using IBM SPSS v27 (Armonk, NY). This study was reviewed and approved by the institutional review board of Rosalind Franklin University.


Overall, 808 students initiated the study (response rate = 50.2%). We removed responses with missing data (n = 78) and with invalid IAT results, including those that were more than 3 SDs from mean completion time (n = 12). Analyses were restricted to only those participants who passed both manipulation check items (analytic sample = 600). The number of participants randomized to each condition ranged from 93 to 103.


Over half of participants were in allopathic medical education programs (n = 333, 55.5%), and the largest percentage were in their first year of training (n = 173, 28.8%). Most participants identified as heterosexual (n = 502, 83.7%), White (n = 367, 61.2%), as cisgender women (n = 377, 62.8%), and training in the Western United States (n = 310, 51.7%). Participants' mean age was 26.1 (SD = 2.82) years. Full demographic information is provided in Table 1.

TABLE 1. - Demographics (N = 600)
Type of Training n %
Medicine (allopathic-MD) 333 55.5
Medicine (osteopathic-DO) 267 44.5
Phase of training*
 1st yr 173 28.8
 2nd yr 152 25.3
 3rd yr 132 22.0
 4th yr 143 23.8
Sexual orientation
 Heterosexual (straight) 502 83.7
 Gay/lesbian 33 5.5
 Bisexual 47 7.8
 Other sexual orientation 18 3.0
 African American or Black 21 3.5
 Caucasian or White 367 61.2
 Hispanic or Latino/a/x 35 5.8
 American Indian or Alaska Native 4 0.7
 Asian 190 31.7
 Other race 20 3.3
Gender identity
 Man 218 36.3
 Woman 377 62.8
 Other gender identity§ 5 0.8
 South 57 9.5
 Northeast 109 18.2
 Midwest 124 20.7
 West 310 51.7
Demographics for participants in the analytic sample. We removed 28 participants who failed the first manipulation check, 64 who failed the second manipulation check, and 26 who failed both manipulation checks.
*The year in training numbers also include 14 students who were in combined MD/MBA, MD/MPH, or MD/PhD programs, who were categorized into the year of medical school they indicated currently being enrolled, or the last year of medical school completed before transitioning to graduate coursework.
Includes sexual orientations other than the listed options, including asexual, demisexual, and pansexual.
Includes participants who indicated their race was other than the listed options.
§Includes those gender identities other than cisgender identities, including transgender, gender fluid, agender, and gender nonbinary.
Northeast: CT, ME, MA, NH, NJ, NY, PA, RI, VT; Midwest: IL, IN, IA, KS, MI, MN, MO, NE, ND, OH, SD, WI; South: AL, AR, DE, DC, FL, GA, KY, LA, MD, MS, NC, OK, SC, TN, TX, VA, WV; West: AK, AZ, CA, CO, HI, ID, MT, NV, NM, OR, UT, WA, WY.

Demographics of the participants randomized to each condition were compared without identifying significant differences suggesting successful randomization (see Table S1, Supplemental Digital Content, Demographics of the sample were also compared with national populations of allopathic and osteopathic medical students nationally to ensure accurate representation (see Table S2, Supplemental Digital Content,

Adherence to PrEP

Assumed adherence to PrEP differed between the 3 condom use groups (F[2,585] = 42.2, P < 0.001) (Fig. 3). In post hoc analyses, the continued-use group was viewed as the most likely to adhere to PrEP (M = 6.27, 95% CI: 6.14 to 6.41). The continued-nonuse condition (M = 5.43, 95% CI: 5.29 to 5.56) was viewed as less likely to adhere compared with the planned-discontinuation condition (M = 6.14, 95% CI: 6.01 to 6.28, P < 0.001) and the continued-use condition (P < 0.001). Assumed adherence to PrEP also differed between patient race (F[1,585] = 5.38, P = 0.02). Black MSM were assumed more likely to adhere to PrEP (M = 6.04, 95% CI: 5.93 to 6.15) compared with White MSM (M = 5.85, 95% CI: 5.74 to 5.96, P = 0.02). The interaction between condom use and race was not significant (F[2,585] = 0.32, P = 0.73).

Comparisons of assumed adherence, overall HIV risk, and willingness to prescribe PrEP between the 3 condom use groups. All analyses of covariances controlled for participants' race, gender identity, sexual orientation, and year in training given their conceptual relevance. **P < 0.01, ***P < 0.001.

HIV Risk

Assumed HIV risk differed between the condom use groups (F[2,585] = 75.5, P < 0.001) (Fig. 3). The continued-nonuse group was viewed as being at the highest HIV risk (M = 6.34, 95% CI: 6.18 to 6.51), and this group was assumed as being at higher HIV risk compared with the planned-discontinuation group (M = 5.52, 95% CI: 5.36 to 5.69, P < 0.001) and the continued-use group (M = 4.89, 95% CI: 4.73 to 5.05, P < 0.001). Additionally, the planned-discontinuation group was viewed as being at higher HIV risk compared with the continued-use group (P < 0.001). HIV risk did not differ (F[2,585] = 2.20, P = 0.14) between Black (M = 5.51, 95% CI: 5.38 to 5.65) and White (M = 5.66, 95% CI: 5.52 to 5.79) MSM. The interaction between condom use and race was not significant (F[2,585] = 0.08, P = 0.92).

Willingness to Prescribe PrEP

Willingness to prescribe PrEP differed between the condom use groups (F[2,585] = 7.13, P < 0.001) (Fig. 3). Willingness was highest for the continued-nonuse group (M = 6.35, 95% CI: 6.18 to 6.52) and was lower for the planned-discontinuation group (M = 5.91, 95% CI: 5.75 to 6.08, P = 0.001). Willingness to prescribe was lower for the planned-discontinuation group compared with the continued-use group (M = 6.27, 95% CI: 6.10 to 6.43, P = 0.01). Willingness to prescribe did not differ (F[1,585] = 1.71, P = 0.19) between Black (M = 6.24, 95% CI: 6.10 to 6.38]) and White (M = 6.11, 95% CI: 5.96 to 6.25) MSM. The interaction between condom use and race was not significant (F[2,585] = 2.41, P = 0.09).

Higher assumed HIV risk (r = 0.26, P < 0.001) and greater assumed adherence to PrEP (r = 0.22, P < 0.001) were correlated with willingness to prescribe PrEP. Explicit heterosexism (r = −0.09, P = 0.02) and attitudes regarding nonmonogamy (r = −0.11, P = 0.01) were correlated with lower willingness to prescribe PrEP (Table 2).

TABLE 2. - Correlations and Means of Key Study Variables
1 2 3 4 5 6 M (SD)
1 Overall HIV risk 5.58 (1.31)
2 Anticipated adherence to PrEP −0.18* 5.95 (1.06)
3 MHS 0.10 −0.03 2.26 (0.67)
4 ATCNMS 0.11 −0.10 0.29* 4.69 (1.12)
5 Implicit racism (IAT) −0.07 0.02 0.11 0.07 0.27 (0.41)
6 Willingness to prescribe PrEP 0.26* 0.22* −0.09 −0.11 0.02 6.18 (1.23)
*P < 0.001.
P < 0.05
P < 0.01.
ATCNMS, Attitudes Towards Consensual NonMonogamy Scale.

Indirect and Moderating Effects

In the first model (Fig. 1A; Table 3), indirect effects of patients' condom use on willingness to prescribe PrEP were evaluated via parallel mediators of assumed HIV risk and assumed adherence to PrEP. Indirect effects of both planned-discontinuation (effect = 0.19, 95% CI: 0.09 to 0.30) and continued-nonuse (effect = 0.43, 95% CI: 0.26 to 0.61) on willingness to prescribe PrEP were identified via higher assumed HIV risk relative to the continued-use group. An indirect effect of continued-nonuse on willingness to prescribe PrEP was also identified via assumed adherence (effect = −0.25, 95% CI: −0.35 to −0.17) because lower adherence was assumed for nonuse relative to continued use (Table 3). Analyses were repeated using the continued-nonuse group as the reference group (see Table S3, Supplemental Digital Content,

TABLE 3. - Coefficients for Multiple Mediation Analyses
Patient Condom Use Effects*
Planned Discontinuation (1) vs. Continued Use (0) Continued Nonuse (1) vs. Continued Use (0)
Effect SE 95% CI Effect SE 95% CI
Relative indirect effects
 HIV risk 0.19 0.10 0.09 to 0.30 0.43 0.09 0.26 to 0.61
 PrEP adherence −0.04 0.03 −0.09 to 0.01 −0.25 0.05 −0.35 to −0.17
Relative direct effect −0.50 0.12 −0.72 to −0.27 −0.09 0.13 −0.35 to 0.17
Relative total effect −0.35 0.12 −0.59 to −0.11 0.08 0.12 −0.16 to 0.32
Contrast effect 0.21 0.06 0.10 to 0.34 0.36 0.04 0.28 to 0.45
*Based on bootstrapped analysis of parallel multiple mediator model with multicategorical predictor (patient condom use) adjusting for participant race, gender identity, sexual orientation, and year in training. Indirect, direct, and total effects reported are relative effects because patient condom use is multicategorical.
P < 0.05.
Contrasts effects represent the difference in absolute magnitude of 2 specific indirect effects (mediational pathways). The patient condom use multicategorical variable was recoded into dichotomous variables to generate contrast effects.

Next, we evaluated moderating effects of heterosexism (MHS) on mediation relationships of condom use on willingness to prescribe PrEP (Fig. 1B-1). No significant moderation effects were identified based on the IMMs (all Ps > 0.05). Finally, we evaluated the moderating effects of attitudes toward nonmonogamy (Figs. 1B-2), which moderated a single indirect effect of condom use (continued nonuse vs. continued use) on willingness to prescribe PrEP (IMM = −0.05, 95% CI: −0.10 to −0.001). Specifically, the indirect effect of condom use (continued-nonuse vs. sustained use) on adherence, whereby the continued-nonuse patient was judged as being less likely to adhere to PrEP, which was subsequently associated with lower willingness to prescribe PrEP, was weaker among participants expressing less judgmental attitudes toward nonmonogamy [low (effect = −0.19, 95% CI: −0.30 to −0.10, P < 0.05), medium (effect = −0.27, 95% CI: −0.37 to −0.18, P < 0.05), and high (effect = −0.31, 95% CI: −0.44 to −0.20, P < 0.05)].

We also investigated the moderating roles of patient race and implicit racism on the relationship between condom use and willingness to prescribe PrEP (Fig. 1C). We used the White MSM as the reference condition. No statistically significant moderated moderation effects of implicit racism or patient race were identified for the mediation relationships between condom use and willingness to prescribe PrEP via assumed HIV risk or adherence (all IMMs P > 0.05).

Replication of Analyses With Full Sample

All analyses were repeated with the full sample, including those participants who failed one or both manipulation checks (N = 718). A similar pattern of results emerged (see Tables S4–S6 and Figure S2, Supplemental Digital Content,


For PrEP to fully contribute to ending the HIV epidemic, prescription must reach all patients at risk for HIV, requiring barriers to PrEP prescription to be surmounted. We found participants were least willing to prescribe PrEP to an MSM using condoms with intention to discontinue, consistent with data collected in 2015 that identified the so-called “prevention paradox.” 28 In contrast to this previous study, which found participants were most willing to prescribe PrEP to a patient who planned to continue using condoms if prescribed PrEP, we found participants were most willing to prescribe PrEP to a patient with continued-nonuse of condoms.28 This is aligned with previous work with practicing clinicians, which found that condom nonuse was an indication for PrEP prescription.42

Consistent with previous investigations of stigma surrounding condom use and PrEP, the prevention paradox may be explained by the long-standing belief that condom use for protection during sex confers some degree of moral superiority and conscientiousness.18 The strong sentiment of personal responsibility in US culture and public health messaging may help to explain the “prevention paradox” and the finding in the present study that participants were least willing to prescribe PrEP to the planned discontinuation patient group even though their risks would have been identical to the patients who never used condoms.20

It may also be possible that growing awareness of PrEP in the United States and in health professions education because this previous study has contributed to prescription decisions that are more based on patients' need and clinical guidelines rather than judgments about the patients' sexual activity.43,44 It should also be noted that the previous study used a within-subjects design, whereas the present study was between-subjects, which also may contribute to differences in results.28

Interestingly, we found that continued-nonuse patients were also viewed as the least likely to adhere to PrEP, which was subsequently associated with lower willingness to prescribe PrEP in the multiple mediation analyses. Lower assumed adherence among continued-nonuse patients and higher assumed HIV risk among continued-nonuse patients upheld the hypotheses. However, this finding is counterintuitive because the continued-nonuse patients correctly recognized their HIV risk and sought effective prevention. Despite the patient's proactive effort to protect his health, participants viewed this patient as less likely to adhere to PrEP, and thus were less willing to prescribe. Previous work has found that many people seeking PrEP are not using condoms and do not intend to begin once prescribed PrEP, making prescription of PrEP an important harm-reduction step.45 These findings may also be explained by stigmatizing attitudes toward PrEP users, specifically of promiscuity and irresponsibility.16,18,19 MSM patients seeking PrEP for HIV prevention with stated, continued-nonuse of condoms may be viewed as irresponsible and unable to adhere to protective measures, and thus, this may have influenced participants' judgment of the patient's likelihood of adhering to PrEP if prescribed.

Hesitancy to prescribe PrEP to patients planning to discontinue condom use reinforces the clinicians' role as a gatekeeper to PrEP. Our vignette presented a patient requesting PrEP from the clinician. Even in the setting of this request, participants were less likely to prescribe PrEP to the patient seeking protection from HIV. There are several potential reasons for this finding, one of which may be existing concerns about “risk compensation” that have been implicated as a reason for withholding PrEP prescription by physicians.9,13 Evidence has not supported risk compensation as a ubiquitous outcome of PrEP use.10,11

Moreover, concerns about patients' behaviors if prescribed PrEP are not the reason to withhold PrEP prescription. Even patients who state they will continue using condoms may not be perfectly adherent to condom use. Although it is true that, compared with the other experimental conditions, the continued-use condition was theoretically at lower HIV risk, this should not preclude PrEP prescription. PrEP is more effective than condoms for HIV prevention and can offer an additional layer of protection even for people who are using condoms consistently, which could alleviate anxiety surrounding HIV transmission.46

Additionally, planned condom use and actual use are not the same. Condoms may break, require an additional step from the user, agreement from partners, and access to a condom anytime sexual activity is occurring. Moreover, consistent with the growing emphasis on patient-centered care in most medical professions, the provider should consider the patient's values and priorities.47,48 This requires clinicians to be mindful of their own potential for bias in judgments about patients seeking PrEP and to ensure biases do not affect clinical encounters. The most recent CDC guidelines, released in December 2021, include condomless sex, multiple sexual partners, or partners of unknown HIV status as indications for PrEP prescription.44 Thus, a routinized approach to PrEP, whereby PrEP is discussed with all patients, would align with current guidance and mitigate provider biases.49

Assumed HIV risk also affected willingness to prescribe, whereby both the planned-discontinuation and continued-nonuse patient conditions were correctly viewed as being at higher HIV-risk without PrEP, which was associated with greater willingness to prescribe PrEP relative to the continued-use patient. However, it should be noted that the differences in assumed HIV risk between the planned-discontinuation and continued-nonuse groups are misaligned because the former group would remain at lower HIV risk without PrEP because condom use would continue, whereas the latter group would be at higher risk because condom use would not continue. However, prescription of PrEP to the planned-discontinuation group would result in identical risk to the continued-nonuse group because both patients would be using PrEP only, without condoms, for HIV prevention.

Effects of Social Biases

We found lower willingness to prescribe PrEP to continued-nonuse patients because assumed nonadherence to PrEP was stronger among participants with more negative attitudes toward nonmonogamy (Fig. 2B). The well-documented stigma toward nonmonogamous relationships may partially explain these findings.22,23 Judgments like this can be detrimental in the clinical setting. A patient seeking PrEP is exhibiting a degree of responsibility and agency over their sexual health and is attempting to access one of the most effective HIV prevention methods available. Failure to prescribe PrEP in response to patient request opens the opportunity for HIV transmission within sexual networks. Clinicians' personal views about nonmonogamy should not impede access to PrEP. Furthermore, evidence suggests that people in nonmonogamous relationships are actually more likely to be regularly screened for STIs and to use condoms.23 The lack of significant findings in the present study, with respect to patient race and implicit racism, and lack of moderating effects of heterosexism are encouraging, given that previous studies have found that explicit heterosexism was associated with lower willingness to prescribe PrEP.27


The findings of this study have several implications for public health and medical education going forward. First, medical education must include training about social biases and how these biases may manifest in clinical encounters, specifically for patients from minoritized groups.50–52 Innovation is also needed in this realm because previous trainings related to implicit biases in medical education have shown minimal effectiveness.50,52 Although there is a growing body of literature investigating PrEP education in medical school curricula, few resources exist for improving PrEP education.32,53–56 As future physicians, training medical students with the tools to address the need for improved access to biomedical HIV prevention strategies, like PrEP, in a patient-centered way will be necessary in the ongoing efforts to end the HIV epidemic in the United States.

Beyond medical education, the results of the study reinforce the need for a standardized approach to PrEP within clinical practice. Implementing a protocol in which all patients presenting are screened for HIV risk factors and subsequently offered PrEP if interested would limit the opportunities for biases to impede access.49 Algorithms to identify patients at risk for HIV or who are candidates for PrEP are also another possible method for limiting biases in clinical encounters or missed opportunities to prescribe PrEP.57,58 Additional investigation and innovation is needed to investigate the health system barriers that impede access to PrEP, including opportunities to leverage existing health technology to close these barriers by limiting points at which human bias may interfere.


The findings of the present study should be interpreted in the context of several limitations that invite future research. First, participants were medical students and thus were not able to independently practice, including PrEP prescription. Additional studies are needed to determine whether the current findings are generalizable to practicing physicians and other clinicians. Although this is a limitation in affecting practice in the short term, investigating PrEP prescription decisions and assumptions about patients seeking PrEP among physicians in training is also important to ensure that PrEP scale up continues to reach all patients once these trainees begin their own practice.

Relatedly, further studies are needed to determine whether assumptions and judgments about a fictional patient are connected to differences in PrEP prescription outside of the experimental context explored here. Previous studies have suggested that clinicians' assumptions and social biases can impede access to care; however, studies of the connections between these biases and outcomes have shown mixed results.59–67 Importantly, racial minority and MSM patients report experiencing discrimination during encounters with the healthcare system, suggesting that social biases may manifest in actual clinical encounters even if they do not always do so in controlled, experimental settings.17,59,60,68–71 An important aspect of these future studies will be expansion of the experimental paradigm and other methodological approaches to include other patient groups (eg, women, transgender people, and people from other minoritized racial and ethnic groups).

Finally, the explicit request for PrEP in our vignette may also limit the generalizability of the findings to other clinical encounters. We also specifically indicated the intentions of the patient regarding condom use if prescribed PrEP. In clinical practice, these may be less obvious or may not be explicitly stated by the patient. Additionally, we did not provide any information about PrEP before the study, so participants may have had varying degrees of knowledge about PrEP, which may have affected the results.


Eliminating HIV is a major public health priority for the United States, and PrEP prescription is essential to achieve this ambitious goal. Prescribers' biases may impede access to PrEP, specifically with regards to the opinion that PrEP encourages sexual risk-taking. We found that a patient who was planning to discontinue condom use was least likely to receive a prescription. Negative attitudes toward nonmonogamous relationships exacerbated the effect of condom discontinuation because of greater assumptions of nonadherence to PrEP. These findings suggest that clinicians' biases about patients' sexual behavior may impede access to one of the most effective available HIV prevention methods.


The authors would like to thank all the University administrators who assisted with distribution of study information and all the students who took the time to participate in this study.


1. Centers for Disease Control and Prevention. HIV Surveillance Report, 2019. 2021. Available at:
2. Riddell J, Amico KR, Mayer KH. HIV preexposure prophylaxis: a review. JAMA. 2018;319:1261–1268.
3. Owens DK, Davidson KW, Barry MJ, et al. Preexposure prophylaxis for the prevention of HIV infection: US Preventive Services Task Force recommendation statement. JAMA. 2019;321:2203–2213.
4. Mayer KH, Molina JM, Thompson MA, et al. Emtricitabine and tenofovir alafenamide vs emtricitabine and tenofovir disoproxil fumarate for HIV pre-exposure prophylaxis (DISCOVER): primary results from a randomised, double-blind, multicentre, active-controlled, phase 3, non-inferiority trial. Lancet (London, England). 2020;396:239–254.
5. Finlayson T, Cha S, Xia M, et al. Changes in HIV preexposure prophylaxis awareness and use among men who have sex with men - 20 urban areas, 2014 and 2017. MMWR Morb Mortal Wkly Rep. 2019;68:597–603.
6. Mayer KH, Agwu A, Malebranche D. Barriers to the wider use of pre-exposure prophylaxis in the United States: a narrative review. Adv Ther. 2020;37:1778–1811.
7. Bates L, Honeycutt A, Bass S, et al. Updated estimates of the number of men who have sex with men (MSM) with indications for HIV pre-exposure prophylaxis. J Acquir Immune Defic Syndr. 2021;88:e28–e30.
8. Kamitani E, Johnson WD, Wichser ME, et al. Growth in proportion and disparities of HIV PrEP use among key populations identified in the United States national goals: systematic review and meta-analysis of published surveys. J Acquir Immune Defic Syndr. 2020;84:379–386.
9. Blumenthal J, Haubrich RH. Will risk compensation accompany pre-exposure prophylaxis for HIV? Virtual Mentor. 2014;16:909–915.
10. Kumar S, Haderxhanaj LT, Spicknall IH. Reviewing PrEP's effect on STI incidence among men who have sex with men-balancing increased STI screening and potential behavioral sexual risk compensation. AIDS Behav. 2021;25:1810–1818.
11. Traeger MW, Schroeder SE, Wright EJ, et al. Effects of pre-exposure prophylaxis for the prevention of Human Immunodeficiency Virus infection on sexual risk behavior in men who have sex with men: a systematic review and meta-analysis. Clin Infect Dis. 2018;67:676–686.
12. Gibas KM, van den Berg P, Powell VE, et al. Drug resistance during HIV pre-exposure prophylaxis. Drugs. 2019;79:609–619.
13. Pleuhs B, Quinn KG, Walsh JL, et al. Health care provider barriers to HIV pre-exposure prophylaxis in the United States: a systematic review. AIDS Patient Care STDS. 2020;34:111–123.
14. Calabrese SK. Understanding, contextualizing, and addressing PrEP stigma to enhance PrEP implementation. Curr HIV AIDS Rep. 2020;17:579–588.
15. Skolnik AA, Bokhour BG, Gifford AL, et al. Roadblocks to PrEP: what medical records reveal about access to HIV pre-exposure prophylaxis. J Gen Intern Med. 2020;35:832–838.
16. Calabrese SK, Underhill K. How stigma surrounding the use of HIV preexposure prophylaxis undermines prevention and pleasure: a call to destigmatize “Truvada Whores”. Am J Public Health. 2015;105:1960–1964.
17. Quinn K, Bowleg L, Dickson-Gomez J. “The fear of being Black plus the fear of being gay”: the effects of intersectional stigma on PrEP use among young Black gay, bisexual, and other men who have sex with men. Soc Sci Med. 2019;232:86–93.
18. Dubov A, Galbo P Jr, Altice FL, et al. Stigma and shame experiences by MSM who take PrEP for HIV prevention: a qualitative study. Am J Mens Health. 2018;12:1843–1854.
19. Golub SA, Gamarel KE, Surace A. Demographic differences in PrEP-related stereotypes: implications for implementation. AIDS Behav. 2017;21:1229–1235.
20. Guttman N, Ressler WH. On being responsible: ethical issues in appeals to personal responsibility in health campaigns. J Health Commun. 2001;6:117–136.
21. Balzarini RN, Shumlich EJ, Kohut T, et al. Dimming the “Halo” around monogamy: Re-assessing stigma surrounding consensually non-monogamous romantic relationships as a function of personal relationship orientation. Front Psychol. 2018;9:894.
22. Conley TD, Moors AC, Matsick JL, et al. The fewer the merrier: assessing stigma surrounding consensually non-monogamous romantic relationships. Analyses Socia Issues Public Policy. 2013;13:1–30.
23. Lehmiller JJ. A comparison of sexual health history and practices among monogamous and consensually nonmonogamous sexual partners. J Sex Med. 2015;12:2022–2028.
24. Bowleg L. Intersectionality: an underutilized but essential theoretical framework for social psychology. In: Gough B, ed. The Palgrave Handbook of Critical Social Psychology. London, UK: Palgrave Macmillan; 2017:507–529.
25. Cooper-Patrick L, Gallo JJ, Gonzales JJ, et al. Race, gender, and partnership in the patient-physician relationship. JAMA. 1999;282:583–589.
26. Calabrese SK, Earnshaw VA, Underhill K, et al. The impact of patient race on clinical decisions related to prescribing HIV pre-exposure prophylaxis (PrEP): assumptions about sexual risk compensation and implications for access. AIDS Behav. 2014;18:226–240.
27. Calabrese SK, Earnshaw VA, Krakower DS, et al. A closer look at racism and heterosexism in medical students' clinical decision-making related to HIV pre-exposure prophylaxis (PrEP): implications for PrEP education. AIDS Behav. 2018;22:1122–1138.
28. Calabrese SK, Earnshaw VA, Underhill K, et al. Prevention paradox: medical students are less inclined to prescribe HIV pre-exposure prophylaxis for patients in highest need. J Int AIDS Soc. 2018;21:e25147.
29. Smith DK, Mendoza MC, Stryker JE, et al. PrEP awareness and attitudes in a national survey of primary care clinicians in the United States, 2009-2015. PLoS One. 2016;11:e0156592.
30. Adams LM, Balderson BH. HIV providers' likelihood to prescribe pre-exposure prophylaxis (PrEP) for HIV prevention differs by patient type: a short report. AIDS Care. 2016;28:1154–1158.
31. Bunting SR, Feinstein BA, Calabrese SK, et al. Assumptions about patients seeking PrEP: exploring the effects of patient and sexual partner race and gender identity and the moderating role of implicit racism. PLoS One. 2022;17:e0270861.
32. Bunting SR, Feinstein BA, Hazra A, et al. Knowledge of HIV and HIV pre-exposure prophylaxis among medical and pharmacy students: a national, multi-site, cross-sectional study. Prev Med Rep. 2021;24:101590.
33. Bunting SR, Feinstein BA, Hazra A, et al. Effects of patient sexual orientation and gender identity on medical students' decision-making regarding pre-exposure prophylaxis (PrEP) for HIV prevention: a vignette-based study. Sex Transm Dis. 2021; 48:959–966.
34. Ma DS, Correll J, Wittenbrink B. The Chicago face database: a free stimulus set of faces and norming data. Behav Res Methods 2015;47:1122–1135.
35. Moors AC, Conley TD, Edelstein RS, et al. Attached to monogamy? Avoidance predicts willingness to engage (but not actual engagement) in consensual non-monogamy. J Soc Pers Relat. 2014;32:222–240.
36. Morrison MA, Morrison TG. Development and validation of a scale measuring modern prejudice toward gay men and lesbian women. J Homosex. 2002;43:15–37.
37. Bar-Anan Y. Running Project Implicit's IAT from Qulatrics. 2020. Available at: Accessed July 8, 2020.
38. Bar-Anan Y. minnoJS File Library. 2021. Available at: Accessed July 12, 2020.
39. Hayes AF. Introduction to Mediation, Moderation, and Conditional Process Analysis : A Regression-Based Approach. 2nd ed. New York, NY: Guilford Press. 2018.
40. Hayes AF, Rockwood NJ. Regression-based statistical mediation and moderation analysis in clinical research: observations, recommendations, and implementation. Behav Res Ther. 2017;98:39–57.
41. Hayes AF. An index and test of linear moderated mediation. Multivariate Behav Res. 2015;50:1–22.
42. Mullins TLK, Zimet G, Lally M, et al. HIV Care providers' intentions to prescribe and actual prescription of pre-exposure prophylaxis to at-risk adolescents and adults. AIDS Patient Care STDS. 2017;31:504–516.
43. Bunting SR, Garber SS, Goldstein RH, et al. Health professions students' awareness, knowledge, and confidence regarding pre-exposure prophylaxis (PrEP): results of a national, multidisciplinary survey. Sex Transm Dis. 2020;48:25–31.
44. Centers for Disease Control and Prevention. Preexposure prophylaxis for the prevention of HIV infection in the United States—2021 update: a clinical practice guideline. 2021. Available at:
45. Parker S, Chan PA, Oldenburg CE, et al. Patient experiences of men who have sex with men using pre-exposure prophylaxis to prevent HIV infection. AIDS Patient Care STDS. 2015;29:639–642.
46. Crosby RA. State of condom use in HIV prevention science and practice. Curr HIV AIDS Rep. 2013;10:59–64.
47. Millenson ML, Shapiro E, Greenhouse PK, et al. Patient- and family-centered care: a systematic approach to better ethics and care. AMA J Ethics. 2016;18:49–55.
48. American Nurses Association. Care Coordination and Registered Nurses' Essential Role. 2021. Available at: Accessed February 16, 2022.
49. Calabrese SK, Krakower DS, Mayer KH. Integrating HIV preexposure prophylaxis (PrEP) into routine preventive health care to avoid exacerbating disparities. Am J Public Health. 2017;107:1883–1889.
50. FitzGerald C, Martin A, Berner D, et al. Interventions designed to reduce implicit prejudices and implicit stereotypes in real world contexts: a systematic review. BMC Psychol. 2019;7:29.
51. Morris M, Cooper RL, Ramesh A, et al. Training to reduce LGBTQ-related bias among medical, nursing, and dental students and providers: a systematic review. BMC Med Educ. 2019;19:325.
52. Morris MC, Cooper RL, Ramesh A, et al. Preparing medical students to address the needs of vulnerable patient populations: implicit bias training in US medical schools. Med Sci Educ. 2020;30:123–127.
53. Bunting SR, Calabrese SK, Garber SS, et al. Where do health professions students learn about pre-exposure prophylaxis (PrEP) for HIV prevention? Med Sci Educ. 2021;31:423–427.
54. Bunting SR, Garber SS, Goldstein RH, et al. Student education about pre-exposure prophylaxis (PrEP) varies between regions of the United States. J Gen Intern Med. 2020;35:2873–2881.
55. Bunting SR, Saqueton R, Batteson TJ. A guide for designing student-led, interprofessional community education initiatives about HIV risk and pre-exposure prophylaxis. MedEdPORTAL. 2019;15:10818.
56. Perucho J, Alzate-Duque L, Bhuiyan A, et al. PrEP (pre-exposure prophylaxis) education for clinicians: caring for an MSM patient. MedEdPORTAL. 2020;16:10908.
57. Marcus JL, Hurley LB, Krakower DS, et al. Use of electronic health record data and machine learning to identify candidates for HIV pre-exposure prophylaxis: a modelling study. Lancet HIV. 2019;6:e688–e695.
58. Krakower DS, Gruber S, Hsu K, et al. Development and validation of an automated HIV prediction algorithm to identify candidates for pre-exposure prophylaxis: a modelling study. Lancet HIV. 2019;6:e696–e704.
59. Maina IW, Belton TD, Ginzberg S, et al. A decade of studying implicit racial/ethnic bias in healthcare providers using the implicit association test. Soc Sci Med. 2018;199:219–229.
60. Hall WJ, Chapman MV, Lee KM, et al. Implicit racial/ethnic bias among health care professionals and its influence on health care outcomes: a systematic review. Am J Public Health. 2015;105:e60–e76.
61. Chapman EN, Kaatz A, Carnes M. Physicians and implicit bias: how doctors may unwittingly perpetuate health care disparities. J Gen Intern Med. 2013;28:1504–1510.
62. Sabin JA, Riskind RG, Nosek BA. Health care providers' implicit and explicit attitudes toward lesbian women and gay men. Am J Public Health. 2015;105:1831–1841.
63. Dehon E, Weiss N, Jones J, et al. A systematic review of the impact of physician implicit racial bias on clinical decision making. Acad Emerg Med. 2017;24:895–904.
64. FitzGerald C, Hurst S. Implicit bias in healthcare professionals: a systematic review. BMC Med Ethics. 2017;18:19.
65. Sabin JA, Rivara FP, Greenwald AG. Physician implicit attitudes and stereotypes about race and quality of medical care. Med Care 2008;46:678–685.
66. Sabin JA, Greenwald AG. The influence of implicit bias on treatment recommendations for 4 common pediatric conditions: pain, urinary tract infection, attention deficit hyperactivity disorder, and asthma. Am J Public Health. 2012;102:988–995.
67. Green AR, Carney DR, Pallin DJ, et al. Implicit bias among physicians and its prediction of thrombolysis decisions for black and white patients. J Gen Intern Med. 2007;22:1231–1238.
68. Thomann M, Grosso A, Zapata R, et al. “WTF is PrEP?”: attitudes towards pre-exposure prophylaxis among men who have sex with men and transgender women in New York City. Cult Health Sex. 2018;20:772–786.
69. Brooks RA, Nieto O, Landrian A, et al. Experiences of pre-exposure prophylaxis (PrEP)-related stigma among black MSM PrEP users in Los Angeles. J Urban Health. 2020;97:679–691.
70. Owens C, Hubach RD, Williams D, et al. Exploring the pre-exposure prophylaxis (PrEP) health care experiences among men who have sex with men (MSM) who live in rural areas of the Midwest. AIDS Educ Prev. 2020;32:51–66.
71. Quinn K, Dickson-Gomez J, Zarwell M, et al. “A gay man and a doctor are just like, a recipe for destruction”: how racism and homonegativity in healthcare settings influence PrEP uptake among young Black MSM. AIDS Behav. 2019;23:1951–1963.

HIV; pre-exposure prophylaxis; disparity; social bias; prevention

Supplemental Digital Content

Copyright © 2022 The Author(s). Published by Wolters Kluwer Health, Inc.